Cardiol Clin 24 (2006) ix
Foreword
Emergency Cardiac Care: From ED to CCU
Michael H. Crawford, MD
Consulting Editor
Each year I spend considerable time on the in-
patient cardiology service and on cardiology con-
sultations. During these activities, my team
interacts daily with the emergency department
physicians. I am also involved administratively
in putting together plans of care for various acute
cardiac diseases that often arrive by way of the
emergency department. I am sure that many of
you share all these activities and more with your
emergency department colleagues. Thus, I was de-
lighted when Drs. Amal Mattu and Mark Kele-
men of the University of Maryland agreed to
edit this issue of the Cardiology Clinics. Repre-
senting emergency medicine and cardiology, re-
spectively, they have assembled a group of
outstanding experts from both disciplines for this
issue, which is dedicated to the interface between
the two fields.
Today, physician interactions concerning car-
diac patients are becoming more complex, with
new cardiac imaging techniques being installed in
many emergency departments. Furthermore,
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.10.002
heart failure specialists and electrophysiologists
are increasingly interacting with patients in the
emergency department first. This issue features
articles that are dedicated to these emerging
interface areas in addition to several articles on
acute coronary syndromes, which represent the
majority of cardiac emergency department visits.
Finally, this issue will be an excellent companion
to the November 2005 issue on chest pain units,
which explores this aspect of emergent cardiac
care.
Michael H. Crawford, MD
Division of Cardiology
Department of Medicine
University of California
San Francisco Medical Center
505 Parnassus Avenue, Box 0124
San Francisco, CA 94143-0124, USA
E-mail address:
michael.crawford@ucsfmedctr.org
cardiology.theclinics.com
 Cardiol Clin 24 (2006) xi–xii

Preface
Emergency Cardiac Care: From ED to CCU

Amal Mattu, MD Mark D. Kelemen, MD

Guest Editors

Cardiac disease is the leading cause of death in
the United States. Tremendous resources in health
care are devoted toward preventing and treating
chronic cardiac disease. Despite this resource allo-
cation, emergency department (ED) visits and hos-
pital admissions for emergency cardiac conditions
continue to rise. Two specialties clearly bear the
greatest burden of dealing with this rising health
care problem: emergency medicine and cardiology.
These two specialties, though functionally and phil-
osophically different in many ways, have discovered
the importance of working together and coordinat-
ing efforts in their shared battle against the rising
epidemic of cardiac disease. In this issue of the Car-
diology Clinics, we have attempted to bridge the gap
between emergency medicine and cardiology in
terms of their respective approaches to treatment
of various emergency cardiac conditions.
Emergency physicians are usually the first
physicians to care for patients who have emer-
gency cardiac conditions. They must initiate
evidence-based therapies in a timely manner, but
they must be certain to plan their initial care in
conjunction with the cardiologists who continue
the appropriate therapies and often provide the
final definitive care. When there is poor coordi-
nation between the treatments provided by the
two separate specialties, patient care suffers.
The first article in this issue provides an
appropriate lead-in to the following articles on
acute coronary syndrome. The authors represent
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.10.001
both emergency medicine and cardiology and
therefore are able to provide a coordinated ap-
proach to the evaluation and initial management
of these patients in the ED. The subsequent article
discusses some of the new concepts in atherogen-
esis. An understanding of the pathophysiology of
atherosclerotic coronary artery disease is impor-
tant if physicians are to understand the newer
treatment of patients with acute coronary syn-
drome (ACS). This article and the following
articles then discuss the evidence-based work-up
and therapies for patients who have ACS. With
the understanding that optimal management of
patients who have ACS requires a multidisciplin-
ary health care team that includes prehospital
providers, ED providers, and in-hospital pro-
viders, this issue also features an article that
focuses on methods of improving community
systems of care to optimize treatment of patients
with ACS.
The use of cocaine has increased in recent
years, both in inner-city and suburban areas.
Consequently, ED visits for cocaine-related chest
pain are on the rise. The recent literature has
demonstrated that rapid rule-out protocols are
safe and effective for these patients. This issue
features an article that discusses some of the
pathophysiology of cocaine-related chest pain
and ACAD. The authors also describe protocols
for the evaluation and management of these
patients.
cardiology.theclinics.com
xii
Congestive heart failure (CHF) is common in
the United States, and the prevalence of CHF is
expected to rise as the population ages. ED visits
for decompensated CHF are on the rise as well.
The use of traditional medications that had been
used for many years is now being questioned, and
some of the newer therapies are equally contro-
versial. A chapter addresses these controversies
and provides suggestions for reasonable emer-
gency management of patients who have decom-
pensated CHF.
The final two articles address the emergency
management of cardiac arrhythmias and severe
hypertension, respectively. Many advances in phar-
macology have resulted in improvements in our
ability to treat acute arrhythmias and hyperten-
sion. The authors of these two articles address some
of the new medications available for treatment, and
they suggest optimal management strategies.
In overseeing the development of this issue of
Cardiology Clinics, we have tried to maintain both
an emergency medicine as well as a cardiology
perspective on the content of the articles. We be-
lieve that physicians from both specialties will
benefit from the topics and the content. Tradition-
ally there has been a gap between the specialties of
emergency medicine and cardiology with regard
to their respective approaches to caring for pa-
tients who have emergency cardiac conditions.
PREFACE
However, there is no doubt that both specialties
must overcome that gap if they are to optimize
the management of these patients. We hope that
this issue helps to bridge the gap.
We would like to thank the dedicated authors
of this issue of the Cardiology Clinics, all of whom
contributed significant time researching and writ-
ing the articles. We would also like to thank Ka-
ren Sorensen and Elsevier for their support of
this issue. Finally, we thank our families for their
patience, support, and encouragement throughout
this process.
Amal Mattu, MD
Division of Emergency Medicine
Department of Surgery
University of Maryland School of Medicine
110 S. Paca Street, Sixth Floor, Suite 200
Baltimore, MD 21201, USA
E-mail address: amattu@smail.umaryland.edu
Mark D. Kelemen, MD
Division of Cardiology
University of Maryland School of Medicine
22 South Greene Street
Baltimore, MD 21202, USA
E-mail address:
mkelemen@medicine.umaryland.edu
 Cardiol Clin 24 (2006) 1–17

Initial Approach to the Patient who has Chest Pain

Luis H. Haro, MDa,b,*, Wyatt W. Decker, MDa,b,
Eric T. Boie, MDa,b, R. Scott Wright, MDa,c
a
Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905, USA
b
Department of Emergency Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA
c
Division of Cardiology and Cardiac Coronary Unit, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA

Scope of the problem tamponade physiology, penetrating ulcer, and

According to Center for Disease Control 2001–
2002 National Hospital Ambulatory Medical
Care Survey, an estimated 107 to 110 million
visits were made to hospital emergency depart-
ments. Of these, approximately 3.5 to 5.4 million
visits (3.4% to 5.3%) were patients who presented
with chest pain as their chief complaint [1]. In
2001, first-listed and secondary hospital discharge
data from the National Registry of Myocardial
Infarction-4 (NRMI-4) indicate there were
1,680,000 unique discharges for acute coronary
syndrome (ACS) [2].
In evaluating acute chest pain, the immediate
goal is to determine the accurate diagnosis and to
initiate the appropriate life-saving therapies as
quickly as possible. It is particularly important to
identify as quickly as possible those patients
presenting with ST-segment elevation myocardial
infarction (STEMI) so that the appropriate re-
perfusion therapies can be initiated with as little
delay as possible. Recent work estimates that at
least 500,000 patients each year qualify for acute
reperfusion therapy for STEMI [3].
The particular challenge facing today’s practi-
tioners of emergency medicine is to evaluate every
patient who presents with acute chest pain for
a variety of life-threatening causes of chest pain,
such as ACS, acute aortic dissection (AD), pul-
monary embolism (PE), pericardial disease with
* Corresponding author. Department of Emergency
Medicine, Mayo Clinic, 200 First Street SW, Rochester,
MN 55905.
E-mail address: haro.luis@mayo.edu (L.H. Haro).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.007
tension pneumothorax (Box 1). Once these entities
are excluded, other benign causes of chest pain are
considered. Most of the cases presenting with
acute chest pain are of benign origin.
This article focuses on assessment; diagnosis,
and management within the first 2 to 3 hours of
emergency department presentation of patients
who have a chief complain of chest pain and
whose clinical status or diagnosis merits admis-
sion to the coronary care unit or medical ICU.
Prehospital evaluation and interventions
A patient complaining of chest pain who is at
risk for ACS should be transported from home or
the outpatient clinic to the emergency department
by an ambulance with advanced life-support
(ALS) capabilities. Only ALS ambulance person-
nel can obtain intravenous access, provide sub-
lingual nitroglycerin and morphine, and provide
advanced cardiac life support if the patient’s
condition deteriorates in route.
Advanced emergency medical services (EMS)
can also perform and transmit prehospital ECGs
(PH-ECGs), stabilize a compromised airway in-
cluding endotracheal intubation and initiation of
mechanical ventilation, and initiate pharmacologic
support in situations of hemodynamic compromise
before arrival at the emergency department.
Many patients who have acute myocardial
infarction (AMI) suffer cardiac arrest in the first
few hours of the event. Many of these patients die
at home suddenly. The use of an ALS-based EMS
offers the best option for early, rapid management
of cardiac arrhythmias and sudden cardiac death.
Lives are saved by having excellent prehospital
cardiology.theclinics.com
2 HARO
Box 1. Differential diagnosis of chest
pain
Life-threatening causes
Acute coronary syndrome
Aortic dissection
Pulmonary embolus
Tension pneumothorax
Other cardiovascular and nonischemic
causes
Pericarditis
Atypical angina
Hypertrophic cardiomyopathy
Vasospastic angina
Other noncardiac causes
Boerhaave’s syndrome (esophageal
rupture with mediastinitis)
Gastroesophageal reflux and spasm
Chest-wall pain
Pleurisy
Peptic ulcer disease
Panic attack
Biliary or pancreatic pain
Cervical disc or neuropathic pain
Somatization and psychogenic pain
disorder
care for patients who are in ventricular fibrillation
arrest, for whom survival rates to hospital dis-
charge with acceptable neurologic function can
reach 40% [4].
Unfortunately, ambulance services are not
always requested. Despite many national educa-
tion campaigns, patients continue to bypass the
EMS systems and arrive by other means. In
NRMI-2, 53% of patients who had STEMI
arrived by private means [5]. In other studies,
the average percentage of patients who had a con-
firmed coronary event and used EMS was 23%,
with a range of 10% to 48%. It is a matter of con-
cern that 16% drove themselves to the emergency
department [6], especially considering that ap-
proximately 1 in every 300 patients transported
to the emergency department by private vehicle
goes into cardiac arrest in route [7]. When they
do call an ambulance, the average patient who
has STEMI does not seek medical care for ap-
proximately 2 hours after symptom onset, and
this pattern seems to be unchanged during the
last decade [8–10]. Average and median delays in
et al
obtaining treatment for patients who had STEMI
were 4.7 and 2.3 hours, respectively, from the 14-
country Global Registry of Acute Coronary
Events project [11].
The reasons given by patients in the United
States for delay in seeking care have been studied
in the REACT project. The investigators con-
ducted focus groups (N ¼ 34,207 participants) in
major regions of the United States. Target groups
included adults who had had previous heart at-
tacks, those at higher risk for heart attack, and
bystanders to heart attacks. Reasons given by
the target groups for delaying seeking help were
(1) they expected heart attack symptoms to be
more dramatic; (2) they unrealistically judged
their personal risk as low; (3) they understood lit-
tle about the benefits of rapid interventions; (4)
they were unaware of the benefits of using EMS
instead of alternative transport; and (5) they
seemed to need the ‘‘permission’’ or advice of
health care providers or family to act [9,12,13].
Prehospital ECGs
PH-ECGs are an underused component in
modern ACS care. In most places in the United
States they can be obtained easily by advanced
EMS personnel and transmitted en route to an
emergency department control center. If trans-
mission is a problem, delays in the emergency
department can be avoided, because the computer
readout is highly accurate and can be called in to
the receiving emergency department. Doing so
allows emergency department personnel to alert
the coronary care unit and be ready for fibrino-
lysis or primary percutaneous intervention
(PPCI). The use of PH-ECGs reduces door-to-
needle time for in-hospital fibrinolysis by a mean
of 10 minutes; NRMI-2 found the use of PH-
ECGs reduced the door-to-balloon- time for
primary PCI by a mean of 23 minutes [5].
Prehospital triage
The use of PH-ECGs can also help EMS triage
more efficiently. In general, patients who have
STEMI should be transported to the nearest
facility that best handles the situation. For exam-
ple, a patient who has an uncomplicated STEMI
can be transported to the nearest facility that
offers acute reperfusion therapy. The patient who
has STEMI and who is also in cardiogenic shock
should be transported preferentially to a facility
capable of performing PPCI or coronary artery
bypass graft surgery rather than being transported
 INITIAL EVALUATION OF CHEST PAIN 3

to a facility that has only intravenous fibrinolysis reperfusion therapies has been provided by the
available, if the transport times are not signifi- French nationwide Unité de Soins Intensifs
cantly different. The Should We Emergently Coronaires 2000 registry [18]. In November 2000
Revascularize Occluded Arteries in Cardiogenic in France use of PHF was used for 9% of the pa-
Shock (SHOCK) trial demonstrated that emer- tients who had STEMI (range, 7%–26%; N ¼
gency revascularization improved 1-year survival, 180). This therapy seemed to offer the best survival
achieving an absolute 12.8% reduction in 6- benefit (Fig. 2). One-year survival was 94% for pa-
month mortality in patients treated with early tients treated with PHF, 89% for patients treated
revascularization (P ¼ .027) [14]. Because much of with in-hospital fibrinolysis or PPCI, and 79% for
the mortality in patients in cardiogenic shock oc- patients who did not receive reperfusion therapy.
curs early in the time course, it is prudent to trans- The American College of Cardiology (ACC)
port patients rapidly to the nearest center of and the American Heart Association (AHA)
excellence that is best equipped to provide the op- recommend establishment of a PHF protocol in
timal cardiovascular care. systems where the prehospital transport times are
more than 20 minutes and ALS-EMS units have
Initiation of prehospital fibrinolysis high volume (O25,000) runs per year [3]. It also
Recent work has suggested that time to reper- recommended that the PHF team include full-
fusion remains one of the most important deter- time paramedics, have available PH-ECG tech-
minants of the degree of salvaged myocardium nology with physician support, and be overseen
[15,16]. Fig. 1 suggests that the earlier reperfusion by a medical director committed to developing
is initiated, the more myocardium one can sal- and maintaining a quality PHF program. Unfor-
vage. Fig. 1 also reveals that a strategy that signif- tunately, no single study in the United States
icantly delays primary reperfusion therapy may has demonstrated a reduction in short-term mor-
result in little myocardial salvage. Such data tality risks compared with hospital-based fibrino-
have inspired clinical trails to perform prehospital lysis. (An in-depth discussion of options for
fibrinolysis (PHF). A meta-analysis of PHF trials reperfusion is given elsewhere in this issue.)
showed time to treatment was reduced by a mean
of 58 minutes, with a range of 33 minutes (Seattle,
WA) to 130 minutes (rural Scotland). The pooled Emergency department evaluation
data demonstrated a 17% relative risk (RR) re-
Once a patient arrives at the emergency de-
duction in early mortality [17]. A more real-world
partment, the initial nursing triage of patients who
contemporary analysis of PHF versus other
have chest pain and are at risk for ACS must be to
a telemetry bed staffed with nurses and physicians
100 capable of performing an immediate assessment
Mortality Reduction, %

80 D Shifts in Potential Outcornes and delivering advanced cardiac life support.

With Different Treatment Strategies
A to B No Benefit
60 A to C Benefit
C B to C Benefit
40 D to B Harm
D to C Harm
20 B A
Extent of
Myocardial Salvage
0
0 4 8 12 16 20 24
Time From Symptom Onset to
Reperfusion Therapy, h
Critical Time-Dependent Period Time-Independent Period
Goal: Myocardial Salvage Goal: Open Infarct-Related Artery

Fig. 1. Hypothetical construct of the relationship

among the duration of symptoms of acute myocardial
infarction before reperfusion therapy, mortality reduc- Fig. 2. Age-adjusted Kaplan-Meier 1-year survival ac-
tion and extent of myocardial salvage. (From Gersh cording to reperfusion strategy. PCI, percutaneous pri-
JB, Stone WG, White DH, et al. Pharmacological mary intervention. (From Danchin N, Blanchard D,
facilitation of primary percutaneous coronary inter- Steg GP, et al. Impact of prehospital thrombolysis for
vention for acute myocardial infarction. JAMA acute myocardial infarction on 1-yr outcome. Circula-
2005;293(8):980; with permission.) tion 2004;110:1913; with permission.)
4 HARO
Placement on a monitor, intravenous access,
oxygen administration, and administration of
aspirin (ASA) or clopidogrel (if the patient is
allergic to ASA) should be done within 5 minutes
of patient arrival. These actions should be driven
by nursing care to minimize the time to initiation
of therapy. In the critically ill patient whose vitals
signs are compromised (ie, cardiac arrest, tachyar-
rhythmias, severe bradycardia, shock, or hypo-
tension), the advanced cardiac life support
guidelines developed by the AHA should be
followed. (Detailed management of particular
arrhythmias is discussed elsewhere in this issue.)
If the patient is stable, and if no PH-ECG is
available, an ECG should be obtained within 10
minutes of arrival at the emergency department,
according to the ACC/AHA guidelines [3]. No
study to date has shown that this can be done con-
sistently, however. Barriers to optimizing early
care for patients who have chest pain are relatively
trivial but are common in busy emergency depart-
ments. Such barriers include large or demanding
clinical volumes and time spent undressing the pa-
tient (especially the elderly), in monitor place-
ment, in obtaining intravenous access, and in
administering ASA and oxygen. These are impor-
tant and necessary tasks, but they should not de-
lay the acquisition of an ECG. The authors
encourage all emergency department personnel
to create a systems-based approach that intention-
ally works to minimize door-to-ECG-acquisition
time in a way that facilitates clinical decision mak-
ing and improves patient outcomes. To evaluate
better the patient presenting with chest pain,
they advocate using a standard 12-lead ECG
and additional ECG techniques including right-
sided leads (V3R through V6R), posterior leads
(V7, V8, and V9), and continuous ST-segment
monitoring in selected patients who have ongoing
chest pain and high pretest probability of ACS.
Approximately 20% of patients who present to
the emergency department with chest pain have
a completely normal 12-lead ECG. The use of
these additional techniques helps uncover a signif-
icant number of patients who have AMI but
whose 12-lead ECG is not diagnostic. The rate
of AMI in patients who have chest pain and a to-
tally normal ECG remains around 1% to 4% [19].
Patient history and examination
The initial history and physical examination
optimally should be performed within 10 minutes
of patient arrival. The initial encounter should be
et al
focused and is done to identify best those who
have life-threatening cardiac and noncardiac
situations.
The pain characteristics in ACS are frequently
substernal and are characterized as crushing,
aching, vise-like, or pressure. The pain commonly
radiates to the neck, jaw, and left arm. Associated
symptoms include dyspnea, nausea, vomiting, di-
aphoresis, and presyncope. Pain often begins
abruptly, lasting 15 minutes or longer and taking
several minutes to reach maximal intensity. The
pain is worse with activity and improves with rest.
Although sharp, stabbing, or fleeting pain is
regarded as atypical for ischemic pain, such pain
is seen in 5% of patients who have AMI [20].
Elderly patients who have ACS can present with
a range of complaints including generalized weak-
ness, altered mental status, syncope, atypical chest
pain, and dyspnea. Dyspnea is the single most com-
mon presenting symptom of angina in patients who
are more than 85 years old [21]. Women presenting
with ACS tend to be older on average than their
male counterparts, to have more comorbid dis-
ease (eg, diabetes and hypertension), and to
have a longer delay from symptom onset to pre-
sentation in the emergency department [22].
Assessment of cardiac risk factors is tradition-
ally considered a routine element of the patient
history, but its value in the emergency department
has been disputed. Patient age, sex, body habitus,
family history of coronary artery disease, and
comorbid illness including diabetes mellitus, hy-
pertension, hypercholesterolemia, and tobacco
abuse are all classic coronary risk factors. Aside
from age, sex, and family history of premature
coronary artery disease, the actual role of risk
factors in predicting ACS or AMI in patients
presenting to the emergency department with
chest pain seems minimal [23]. Risk factors are
based on population studies and thus are more
predictive of development of coronary artery dis-
ease over a lifetime, not of whether a patient expe-
riencing chest pain in the emergency department is
likely to have ACS [21]. One should not dwell on
this component of the history.
The initial physical examination should focus
on the cardiovascular system. The clinician should
evaluate the jugular venous pressure, looking for
elevation, the presence of a Kussmaul’s sign, and
the presence of hepato-jugular reflux. The lung
fields should be examined for signs of congestion
and wheezing. The heart and peripheral vascula-
ture should be examined for abnormalities. The
abdomen should also be examined for signs
 INITIAL EVALUATION OF CHEST PAIN 5

of hepatic congestion and abdominal aortic the most expeditious reperfusion management.
pathology. The authors recommend the use of intravenous
fibrinolytic therapy (IFT) in hospitals without on-
Cardiac biomarker and laboratory assessment site and experienced catheterization laboratories.
The use of IFT should be restricted to patients
The use of cardiac biomarkers is well estab-
who present within 6 hours of symptom onset and
lished in patients who have ACS. These bio-
have clear-cut ST segment elevation or new left
markers provide the most accurate diagnosis of
bundle branch block. IFT should not be used in
acute myocardial injury and are considered the
asymptomatic patients whose symptoms began
reference standard for the diagnosis of AMI
more than 24 hours previously [3]. This therapy
[3,24,25]. In addition to diagnostic accuracy, the
is most useful in the first 3 hours after symptom
troponins provide prognostic data [26].
onset and restores coronary flow completely in
Additional cardiac biomarkers that provide
about half the patients treated with it. The success
prognostic data in STEMI and non-STEMI
rate, defined as Thrombolysis in Myocardial In-
patients include B-type natriuretic peptide (BNP)
farction (TIMI)-3 flow at 90 minutes after the
and C-reactive protein [27] Patients who had ele-
start of treatment, varies between 32% and
vations of several of these biomarkers faced the
55%. The great risk of IFT is hemorrhage, includ-
greatest risks. In the Orbofiban in Patients with
ing a 0.5% to 1.2% risk of intracranial hemor-
Unstable Coronary Syndromes Trial 16 study,
rhage and a 3% to 6% risk of other major
baseline measurements of troponin I, C-reactive
bleeding complications (gastrointestinal bleeding,
protein, and BNP were performed in 450 patients.
retroperitoneal bleeding, a R 4-g drop in hemo-
Elevations in troponin I, C-reactive protein, and
globin) [29]. PPCI can be performed in nearly all
BNP were independent predictors of the composite
patients who have STEMI and is more successful
of death, myocardial infarction (MI), or congestive
than IFT at restoring TIMI-3 flow [30,31]. PPCI
heart failure. When patients were categorized on
has a lower risk of intracranial hemorrhage and
the basis of the number of elevated biomarkers at
reinfarction compared with IFT [3,18]. PPCI
presentation, there was a near doubling of the mor-
must be done promptly. Several groups have
tality risk for each additional biomarker that was
reported increased mortality with increased
elevated (P ¼ .01) [28].
door-to-balloon time [32,33,34]. De Luca and col-
Recent work suggests a prognostic role for
leagues [35,36], in a study population of 1791 pa-
soluble CD-40 ligand and myeloperoxidase in
tients who had STEMI treated with emergent
patients who have AMI, but additional work is
PCI, demonstrated that for every 30-minute delay
needed to confirm the long-term prognostic role
in reperfusion therapy with PPCI, there was an in-
of these newer markers and what value they add
creased 1-year mortality of 7.5%. ACC/AHA
when combined with troponin, BNP, and C-
guidelines stipulate, that after adjusting for base-
reactive protein in patients who have AMI.
line characteristics, time from symptom onset to
The authors recommend additional laboratory
balloon inflation is significantly correlated with
testing in patients presenting with chest pain.
1-year mortality (RR, 1.08 for each 30-minute de-
Glucose, creatinine, and electrolyte levels and
lay; P ¼ .04). The ACC recommends a door-to-
a complete blood cell count should be obtained.
balloon time of 60 to 120 minutes [3]. This goal
In patients in whom acute pulmonary embolism is
can be achieved only if specific intradepartmental
suspected, a D-dimer should be obtained. All
time goals are reached for the critical emergency
these laboratory measures should be obtained as
department actions:
soon as possible.
1. The diagnosis: door-to-ECG (preferably less
Reperfusion strategies in ST-segment elevation
than 10 minutes)
myocardial infarction
2. The decision to treat: door-to-catheterization
A detailed discussion of reperfusion strategies in team activation (preferably within 15–25 mi-
STEMI is given elsewhere in this issue. Following is nutes and preferably performed by the emer-
a brief discussion of initial considerations. gency physician on call to minimize delays
When faced with a patient whose ECG dem- in consultation)
onstrates STEMI or new left bundle branch block, 3. The transition in care: door-to-emergency de-
it is important to have pre-established multidisci- partment departure (preferably within 45–60
plinary guidelines in place to indicate the best and minutes)
6 HARO
This flow allows a 30-minute response time for
the catheterization team members to respond after
hours. Most invasive cardiologists are able to
perform the first balloon inflation within 45
minutes after the patient’s arrival at the catheter-
ization laboratory. This flow would allow a door-
to-balloon time of 90 to 120 minutes. Although it
is impossible to achieve this flow for 100% of
patients, maintaining these goals and making
efforts to achieve such timelines are essential steps
in achieving consistency in care and making
health care more reliable. Future quality-assur-
ance efforts in emergency medicine and cardiology
will be to study, publish, and advise on best
practice models that can serve as blueprints to
achieve such goals.
High-risk ST-segment elevation myocardial
infarction
In certain situations, slightly delayed PPCI is
preferable to immediate IFT. Patients who pres-
ent in cardiogenic shock have a high mortality
risk, and the data from the SHOCK trial sug-
gested that immediate revascularization is superi-
or to delayed revascularization [14]. In practice,
for such patients it is preferable to initiate reperfu-
sion therapy with PPCI, even delayed PPCI, rather
than immediate IFT if one can activate a team for
PPCI reasonably quickly. Additionally, obser-
vational data from the NRMI registry have sug-
gested that patients who have more advanced
CHF (Killip class R II) have better outcomes
with PPCI than with IFT [5].
Risk of bleeding
A rare but important clinical conundrum is the
patient who does not have a high risk of STEMI
and who has a perceived high risk for bleeding
from IFT. What should one do? Delay treatment
with IFT and transfer the patient to a facility with
PPCI, accept the risk and administer IFT anyway,
or withhold all reperfusion therapies? One poten-
tial decision is to withhold IFT in any patient who
has a greater than 4% risk for life-threatening
bleeding and to transfer the patient to another
facility for PPCI. To assess risk of bleeding, the
ACC/AHA has defined contraindications and
cautions for fibrinolysis use [3]. Risk scores (based
on points for bleeding risks) are better predictors,
and among these the best are those derived from
observational studies [37]. In centers where imme-
diate PPCI is not available, the use of such risk
scores is highly recommended.
et al
Delayed primary percutaneous coronary
revascularization
In the United States the balance of risk–benefit
between the expedited transfer of patients for
PPCI and more immediate treatment with IFT
remains an uncertain science. The decision re-
garding transfer must be based on multiple
factors, and transfer should be made only when
there is a clear-cut benefit for the patient. Al-
though there is little consensus in the United
States regarding a role for delayed PPCI in
patients presenting to community hospitals with-
out PPCI capability, the data from Danish Acute
Myocardial Infarction-2 trial are provocative. The
investigators demonstrated that patients trans-
ferred for PPCI within 2 hours of presentation
had a better composite outcome (death, stoke,
and recurrent nonfatal MI) than if treated with
fibrinolysis at the local hospital [33,35,38]. No
consensus has emerged in the United States re-
garding this issue, and ongoing randomized clini-
cal trials are testing this strategy along with
a strategy of facilitated PPCI using upstream ad-
ministration of IFT before PPCI.
From the current literature, it is not possible to
state that a particular reperfusion strategy is
applicable to all STEMI, in all clinical settings,
and in all hours of the day. It is most important to
choose the appropriate reperfusion strategy based
on the patient’s clinical presentation and symp-
tom onset and to provide the therapy in a timely
fashion.
Early therapy for acute cardiac syndromes
Oxygen
Supplemental oxygen administration has be-
come routine for all patients presenting with chest
pain. Experimental results indicate that breathing
oxygen might limit ischemic myocardial injury,
and there is evidence that it reduces ST-segment
elevation. Therefore, the use of oxygen is recom-
mended for all AMI patients during in the first 6
hours and longer if the AMI is complicated by
congestive heart failure, PE, or other significant
underlying disease causing hypoxemia.
Aspirin
Aspirin (162–325 mg) should be chewed imme-
diately at arrival if no ASA allergy exists. ASA
produces a rapid antithrombotic effect by near-
total inhibition of thromboxane A2. The second
International Study of Infarct Survival Co-
llaborative (ISIS-2) demonstrated an absolute
risk difference in 35-day mortality of 2.4% (RR,
 INITIAL EVALUATION OF CHEST PAIN
23%) [39]. When combined with IFT, the absolute
difference in mortality was 5.2% (RR, 25%) [40].
In patients who are allergic to ASA, clopidogrel
should be substituted.
Unfractionated heparin
The authors recommend the routine use of
unfractionated heparin (UFH) in all patients who
have AMI, and it is essential for those undergoing
IFT. UFH administration should precede IFT
in all circumstances. The authors recommend
weight-adjusted UFH administration including
a bolus of 60 U/kg (maximum of 4000 U)
followed by a 12-U/kg/hour infusion (maximum
of 1000 U/hour) adjusted to a partial thrombo-
plastin time at 1.5 to 2.0 times control. If a nonse-
lective fibrinolytic (streptokinase, urokinase, or
anistreplase) is used, UFH can be given only to
patients who have a high risk of systemic emboli,
such as large or anterior MI, atrial fibrillation, or
known left ventricular thrombus. For patients
who will receive PPCI, the authors recommend
a weight-adjusted bolus dose of UFH of 50 to
70 U/kg accompanied by a 12-U/kg/hour infusion
(maximum of 1000 U/hour).
Low molecular weight heparin
Low molecular weight heparin is an acceptable
alterative to UFH for patients younger than 75
years, provided that serum creatinine is not
greater than 2.5 mg/dL in men or 2.0 mg/dL in
women.
The authors recommend the use of enoxaparin
as a 30-mg intravenous bolus followed by 1.0 mg/
kg injected subcutaneously every 12 hours for 48
to 72 hours. In the United States, the Food and
Drug Administration has not yet approved enox-
aparin for treatment of IFT, but ongoing studies
are testing the efficacy of this combination.
Nitroglycerin
Patients who have ongoing chest discomfort
should receive sublingual nitroglycerin (0.4 mg)
every 5 minutes for a total of three doses, after
which the need for intravenous infusion is as-
sessed. Nitrates reduce preload and afterload
through peripheral arterial and venous dilatation,
relax the epicardial coronary arteries to improve
coronary flow, and dilate collateral vessels, poten-
tially creating a favorable subendocardial-to-
epicardial flow ratio. Nitrates are harmful in
patients who have hypotension, bradycardia, or
a suspected right ventricular infarction and in
those who have taken a phosphodiesterase
7
inhibitor-5 for erectile dysfunction within the
last 24 hours.
Morphine sulfate
Pain increases sympathetic activity, and surges
in catecholamine levels have been implicated as
having a role in plaque fissuring and thrombus
propagation in AMI, as well as reducing the
threshold for ventricular fibrillation. Morphine is
useful in controlling the pain of AMI but should
be used judiciously. When necessary, morphine
sulfate should be given in 2 to 4 mg doses
intravenously with increments 2 to 8 mg at 5- to
15-minute intervals. Recent data from Can Rapid
Risk Stratification Of Unstable Angina Patients
Suppress Adverse Outcomes With Early Imple-
mentation Of The ACC/AHA Guidelines study
[41] suggest that the use of morphine, alone or
in combination with nitroglycerin for patients pre-
senting with non-STEMI, is associated with ad-
verse outcomes. The rate of AMI increased from
3.0% in the group of patients not receiving mor-
phine (n ¼ 40,036) to 3.8% in the group of pa-
tients receiving morphine (n ¼ 17,003). Death
increased from 4.7% to 5.5%, respectively, and
the composite endpoint of death and AMI in-
creased from 7.1% to 8.5%, respectively. There
might be a selection bias, because the morphine
group had higher incidence of ST-segment depres-
sion, transient ST-segment elevation, and positive
cardiac markers and was more likely to receive an
ECG within 10 minutes of arrival and to be cared
for by a cardiologist. The authors address this po-
tential by providing a risk adjustment and a pro-
pensity score matched-pair analysis for 33,972 of
the patients. They conclude that the use of mor-
phine is associated with a higher mortality and
raise concerns regarding the safety of using mor-
phine for this selected population. They hypothe-
size that common side effects such as hypotension,
bradycardia, and respiratory depression result in
decreased myocardial oxygen delivery, increased
arterial carbon dioxide, and perhaps even de-
creased cerebral perfusion. Unfortunately, these
parameters were not evaluated in this observa-
tional study. The authors’ observation and con-
clusions are interesting and deserve future study.
To date the ACC/AHA recommendations sup-
port the use of morphine as a class I indication
(conditions for which there is evidence or general
agreement that a given procedure or treatment is
beneficial, useful, and effective), level of evidence
C (only consensus opinion of experts, case studies,
or standard of care exists). It is possible that this
8 HARO
classification will change to a class IIB (useful-
ness/efficacy less well established by evidence/
opinion).
Beta-blockers
Intravenous or oral beta-blockers should be
given promptly to patients who have AMI and
who do not have a contraindication. Immediate
beta-blocker therapy seems to reduce the magni-
tude of infarction and the incidence of associated
complications in patients not receiving fibrinoly-
sis, to reduce the rate of reinfarction in those
receiving fibrinolysis, and to reduce the frequency
of life-threatening arrhythmias. During the first
few hours of STEMI, beta-blockers diminish
myocardial oxygen demand by reducing heart
rate, systemic arterial pressure, and myocardial
contractility. In addition, prolongation of diastole
may augment perfusion to ischemic myocardium,
particularly the subendocardium. In the ISIS-1
trial, immediate oral atenolol, 5 to 10 mg,
followed by atenolol, 100 mg daily, reduced 7-
day mortality from 4.3% to 3.7% (P ! .02; 6 lives
saved per 1000 treated). In the Metoprolol in
Acute Myocardial Infarction trial [42], metopro-
lol, 15 mg administered intravenously in three di-
vided doses followed by 50 mg orally every 6
hours for 48 hours and then 100 mg daily, reduced
15-day mortality from 4.9% to 4.3% as compared
with placebo. The benefits of routine early intra-
venous use of beta-blockers in the fibrinolytic
era have been challenged by two later randomized
trials and by a post hoc analysis of the use of ate-
nolol in the Global Utilization of Streptokinase
and TPA (alteplase) for Occluded Coronary Ar-
teries (GUSTO-1) trial [43,45].
Beta-blocker therapy is contraindicated in
patients who have STEMI and moderate left
ventricular failure until compensated and in
patients who have bradycardia, hypotension,
shock, a PR interval greater than 0.24 second,
second- or third-degree atrioventricular block,
active asthma, or reactive airway disease. Beta-
blockers are also contraindicated in cocaine-in-
duced chest pain because of the risk of inducing
coronary spasm.
Glycoprotein IIb/IIIa inhibitors
Antagonism of glycoprotein (Gp) IIb/IIIa re-
ceptor blocks the final common pathway of
platelet aggregation. Three such agents are avail-
able in the United States: abciximab, tirofiban,
and eptifibatide. The use of these agents with IFT
is not proven despite two large trials (GUSTO-V
et al
and Assessment of the Safety and Efficacy of
a New Thrombolytic-3), which tested the efficacy
of combined therapy. It is reasonable to start an
intravenous Gp IIB/IIIA antagonist before initi-
ation of PPCI in selected patients. In STEMI, the
evidence favors abciximab for patients receiving
immediate PPCI, but there are no direct compar-
isons of abciximab and eptifibatide. Decisions
regarding upstream use of a Gp IIb/IIIa agent
should be made in consultation with the consul-
ting interventional cardiologist.
The use of Gp IIb/IIIa inhibitors in unstable
angina and non-STEMI was best evaluated by
Boersma and colleagues [45] who published a com-
prehensive meta-analysis that included six investi-
gations (PARAGON A, Platelet Receptor
Inhibition in Ischemic Syndrome Management,
Platelet Receptor Inhibition in Ischemic Syn-
drome Management in Patients Limited By Un-
stable Signs And Symptoms, GUSTO IV-ACS,
Platelet IIb/IIIa Antagonist for the Reduction of
Acute Coronary Syndrome Events in a Global
Organization Network B (PARAGON B), and
Platelet Glycoprotein IIb/IIIa in Unstable An-
gina: Receptor Suppression Using Integrilin Ther-
apy (PURSUIT) The primary endpoint was death
or nonfatal MI. The meta-analysis included
31,402 patients. Patients who received the Gp in-
hibitor had a 1.2% risk reduction in the odds of
death or MI after randomization (5.7% versus
6.9%; P ¼ .0003). Recently, the Treat Angina
with Aggrastat and Determine Cost of Therapy
with an Invasive or Conservative Strategy-TIMI 18
trial combined early PCI with a Gp IIb/IIIa antag-
onist and demonstrated a benefit from the use of the
combination [46]. The authors believe that initia-
tion of these agents in the emergency department
is reasonable and may facilitate successful early
PCI. (A detailed discussion of early management
of non-ST-segment elevation ACS appears later
in this issue).
Undifferentiated chest pain: the threats to life
If the ECG obtained has no significant ST-
segment abnormalities, the evaluation of acute
chest pain continues with an in-depth clinical
history taking that focuses on the characteristics
of pain, the time of onset, and the duration of
symptoms, associated symptoms, risk assessment
for ACS, PE, AD, and pericardial disease with
tamponade physiology, and an examination that
emphasizes vital signs and cardiovascular, pulmo-
nary, and neurologic status. Most physicians do
 INITIAL EVALUATION OF CHEST PAIN
a mental exercise; others use an algorithmic
approach; but the focus of the evaluation is not
on determining the most likely cause of chest
pain. Instead, the question is: What life threaten-
ing entity could cause this patients chest pain
(even if the possibility is less than 5%), and how
will I make sure that I exclude it? As with missed
AMI, the medical community, patients, and cer-
tainly the judicial system have no tolerance for
missed life-threatening entities. The discussion
that follows is based on the two most important
threats to life, AD and PE.
Acute aortic dissection
Epidemiology
AD is the most common and most lethal aortic
emergency [45,47,48]. The true incidence has been
difficult to determine, because many incorrectly
diagnosed cases escape notice. The occurrence of
AD was reported to be between 5 and 20 per mil-
lion population [49]. The incidence of hospital ad-
mission for AD ranged between 1 in 5335 to 1 in
16,550 [50]. Among the life-threatening causes of
chest pain, AD has the highest mortalitydan esti-
mated 1% to 2% per hour for the first 48 hours
[44]. Unfortunately, when initially evaluating pa-
tients who have AD, physicians correctly suspect
the diagnosis in as few as 15% to 43% of cases
[51–53]. Diagnostic delays of more than 24 hours
after hospitalization are common and occur in up
to 39% of the cases (31% for proximal AD and
53% for distal AD) [54]. Finally, when the diagno-
sis is made, it is often an incidental discovery
made during an advanced imaging procedures re-
quested to assess for other diagnoses. Several fac-
tors drive this poor performance. The most
frustrating combination for a physician to face
when evaluating a patient is a chief complaint
(ie, chest pain) that has no typical presentation
and is life threatening. Unfortunately, that combi-
nation is the rule in AD: classic findings such as
ripping interscapular back pain, diastolic mur-
mur, and a wide mediastinum are present less
than one third of the cases [51,53,55].
During the last decade a substantial number of
studies have evaluated the predictive value of
several historical clues and physical examination
findings in AD. Particularly useful is Klompas’
[53] comprehensive review of the literature and the
International Registry of Aortic Dissection (the
IRAD database) whose inception and structure
is based on 18 large referral centers in six coun-
tries [55]. Following are some of the most recent
9
advances in understanding the clinical presenta-
tion of AD, based on these and other studies.
Clinical manifestations of aortic dissection
Traditionally, AD occurs in patients in the
later decades of life: 95% of these patients are
older than 40 years, and the mean age at pre-
sentation is 65 years [51,53,55]. Risk factors for
AD include hypertension, male sex, non-white
race, connective tissue disease (ie, Ehlers-Danlos
syndrome or Marfan’s syndrome), bicuspid aortic
valve, coarctation of the aorta, and drug use in-
cluding methamphetamine and cocaine. Januzzi
and colleagues [56] evaluated patients younger
than 40 years who had AD. Of 1078 patients in
IRAD who had AD, 69 (6.4%) were younger
than 40 years old. In these 69 patients, traditional
risk factors such as hypertension and atheroscle-
rosis were significantly lower than in the overall
population of patients who had AD. The inci-
dence of Marfan’s syndrome, bicuspid aortic
valve, and prior aortic valve surgery was signifi-
cantly higher in these patients (P ! .0001).
Clinical manifestations of AD are often dom-
inated by the pathoanatomic characteristics of
a malperfusion syndrome from a dissection-re-
lated side branch obstruction [57]. Severe pain is
the most common presenting symptom; 74% to
84% of the patients recall an abrupt onset
[51,53,55]. This symptom alone should trigger sus-
picion of an AD. Anterior chest pain is more fre-
quent in patients who have AD involving the
aortic arch, whereas patients who have AD distal
to the left subclavian more often experience back
pain and abdominal pain (29% of all patients
who have AD; 42% of the patients who have dis-
tal AD) [55]. Contrary to common belief, pain is
described as sharp more often than tearing or rip-
ping. Pain that migrates throughout the chest is
present in only 28% of the patients, with no differ-
ence between type A or type B.
Less frequent presentations of AD are stroke
(carotid occlusion), heart failure (aortic valve
insufficiency), syncope (tamponade, central ner-
vous system ischemia), buttock and leg pain with
or without lower extremity weakness (femoral
artery occlusion), back and flank pain (renal
artery occlusion), abdominal pain (mesenteric
ischemia or celiac trunk involvement), and of
course the infamous painless AD.
Of particular importance is the presence of
syncope as a symptom of AD. In 2001, IRAD had
identified 728 patients who had AD. Syncope was
found in 96 patients (13%), including 24 (3%)
10
who described it in isolation without any symp-
toms of chest or back pain [56]. Syncope is more
frequent in proximal AD than in distal AD
(19% versus 3%; P O .001). Those who had syn-
cope were more likely to have cardiac tamponade
(28% versus 8%; P O .001), stroke (18% versus
4%; P O .001), and other neurologic deficits
such as decreased level of consciousness, coma,
and spinal cord ischemia (25% versus 14%; P O
.005). In 46% of patients, a cause was not found.
Other symptoms that require particular atten-
tion are transient ischemic attack and other focal
neurologic complaints. These findings are docu-
mented in 4.7% to 17% of the cases [51,53,55]. A
complaint of focal deficits paired with chest pain
has one of the highest likelihood ratios of being
AD (positive likelihood ratio, 6.6–33) [53].
Findings on physical examination
Physical examination findings associated with
AD are typically present in less than half the cases
[53]. Among the most useful signs able to predict
AD is a pulse deficit. A pulse deficit between the
carotid, radial, or femoral arteries is relatively in-
frequent (25%–30%) but when present is strongly
suggestive of AD in the setting of chest or back
pain (positive likelihood ratio, 5.7; 95% CI, 1.4–
23.0) [53]. This finding can predict in-hospital
complications and mortality. Bossone and col-
leagues [58] noted that in-hospital mortality is
higher when pulses absent (41% versus 25%).
The more absent pulses the higher mortality.
Blood pressure on presentation does not seem to
be helpful in predicting who might have AD. Al-
though approximately half the patients present
with elevated blood pressure, an equal proportion
are either hypertensive or normotensive [52,53,55].
A history of chronic hypertension as a risk factor is
helpful, however, because it is the most frequent
risk factor. Other physical findings are murmur
of aortic insufficiency (detected in one third of
the cases) and muffled heart tones and jugular ve-
nous distention that point toward cardiac
tamponade.
Biomarkers in the workup of aortic dissection
The lack of symptoms or signs that have a good
negative predictive value has forced investigators
to look at serologic means to diagnose AD. This
concept is particularly attractive because a serum
test with a good negative predictive value would
obviate the need for imaging. In the last decade
several efforts have been made toward achieving
HARO et al
this goal. Investigators have looked at soluble
elastic compounds, D-dimer, and smooth muscle
myosin heavy chain (SMMHC).
SMMHC is a major component of smooth
muscle. Katoh and Suzuki [59] first described the
use of SMMHC in AD in Japan in 1995.
SMMHC was tested in serum of healthy subjects
with levels of 0.9 G 0.9 ng/mL and in four pa-
tients who had AD confirmed by surgery, all
four of whom demonstrated elevated levels at pre-
sentation (O 7 ng/mL) that dropped to normal
values after 24 hours. The immunoassay showed
a sensitivity of 90% in the first 12 hours after on-
set of symptoms with a specificity of 97%. Most
recently, Suzuki and colleagues [60] documented
25 patients enrolled in the IRAD whose measured
SMMHC showed elevated levels of 19.6 G 56.6
ng/mL (normal ! 2.5 ng/mL) with a presentation
time of 6.1 hours G 4.5 hours. This study showed
superior diagnostic performance in the early
hours after onset (O 90% sensitivity in the first
3 hours after onset). The sensitivity decreased to
44% after 3 hours, however, and decreased most
significantly after 6 hours [60].
Elastin is one of the major components of the
arterial wall. An ELISA to measure soluble elastin
fragments, a product of human aortic elastin, was
developed by Shinohara and colleagues [61]. They
reported that 16 of 25 patients who had AD had
elevated soluble elastin fragments; unfortunately,
all patients who did not have a patent false lumen
had normal levels. Although the authors conclude
that the test might be helpful in the diagnosis and
screening of AD, one patient who had AMI had
elevated levels, the study was retrospective, and
the negative predictive value was poor and there-
fore not useful to exclude AD.
D-dimer is frequently used to exclude throm-
boembolic disease in low-risk patients. It is
a cross-linked fiber degradation product formed
by plasmin, which serves as a marker of clot lysis.
Weber and colleagues [62] prospectively tested D-
dimer levels in 10 patients suspected of having
AD. In addition, they retrospectively reviewed
14 patients who had proven AD; 35 patients
served as a control group. D-dimer was elevated
in all patients who had AD. No patients who
did not have AD had elevated D-dimer. The au-
thors concluded that the presence of AD is unlikely
in the setting of a negative D-dimer [62]. This
conclusion is thought provoking and currently
is being evaluated in a multicenter study along
with soluble elastin fragments and SMMHC
(L.H. Haro, personal communication, 2004).
 INITIAL EVALUATION OF CHEST PAIN
ECG findings suggesting ST-segment elevation
myocardial infarction
Emergency physicians and cardiologists are
frequently challenged by a patient who presents
with chest pain that radiates to the back and an
ECG with ST-segment elevation suggestive of an
AMI. In these cases, therapy is frequently de-
layed, because patients often go to CT or transfer
without fibrinolytic therapy to rule out dissection.
The frequency of ST-segment elevation and Q
waves suggestive of an AMI in AD are well
documented in the literature. In the initial
IRAD publication, the finding was documented
in 4.6% of type A and in 0.7% of type B [55]. In
Komplas’ review [53], new MI on ECG was pres-
ent in 7% of the cases (95% CI, 4–14). Coronary
artery involvement (CAI) in AD and ST abnor-
malities in the ECG do not go hand-in-hand, how-
ever. Bossone and colleagues [63] evaluated the
clinical characteristics and outcomes of 475 pa-
tients who had AD, of whom 64 (13.5%) had
CAI. When they reviewed the ECGs, patients
who had CAI were more likely to show new Q
wave or ST-segment elevation (16.7% versus
4.3%; P ¼ .000l). Thus, these ECG findings
were present in only one of six patients who had
CAI. Therefore, ST-segment elevation in AD
seems to be uncommon, and the ECG often is
not diagnostic even when AD with CAI is present.
The low frequency of AD with ST-segment eleva-
tion compared with actual STEMI would argue
for selected and infrequent need for imaging.
(The US Census Bureau projected a 296,042,501
population for May 2005. The best estimate of
the occurrence of AD is between 5 and 20 per mil-
lion population, or approximately 1400 to 6000
new dissections per year, 62% of which would
be type A. Five percent to 7% of those type A dis-
sections would have ST-segment abnormalities
suggestive of a STEMI, representing less than
300 cases in the United States each year. Approx-
imately 500,000 patients in the United States have
STEMI each year.)
In general, when a patient presents with chest
pain and ST-segment elevation in the ECG, one
should assume STEMI and treat accordingly. The
authors believe that reperfusion therapy with
PPCI is the safest way to proceed. If no culprit
artery is found, arteriography or intraoperative
transesophageal echocardiography (TEE) can be
performed. If PPCI is not available, and symptom
onset occurred more than 3 hours previously, the
benefit of fibrinolysis is low. Obtaining a CT
emergently or transferring the patient for PPCI
11
is the best option. Heparin can be withheld. In
a patient who demonstrates ST-segment elevation
in the ECG, an anterior distribution, and more
than 3 hours of pain with no findings highly
suggestive of AD (severe abrupt pain, focal deficit,
pulse deficit), and a completely normal chest
radiograph, the likelihood of AD is low (likeli-
hood ratio, 0.07%; 95% CI, 0.03–0.17) [18], and
the morbidity and mortality of STEMI are high.
The authors recommend fibrinolysis while await-
ing transfer or further imaging such as CT or
TEE.
Imaging
The choice of imaging modality is usually
based on availability and the patient’s stability.
CT and TEE are frequent choices [55]. CT is the
most readily available, widely used, noninvasive
technique for the diagnosis of AD. The sensitivity
and specificity of CT approaches 100% for the di-
agnosis of AD [64]. Aortography was the tradi-
tional method for making the diagnosis of AD,
with an accuracy of 95% to 99% [65]. Aortogra-
phy, however, is highly invasive, requires the pa-
tient be out of the emergency department for an
extended period of time, and exposes the patient
to a significant contrast load [47]. TEE is being
used with increasing frequency and has been
shown to be safe, even in critically ill patients
[58,66,67]. The sensitivity of TEE for detecting
both proximal and distal dissections is 100%
[59]. The main limitations to the use of the TEE
is a lack of widespread, 24-hour availability. Fi-
nally, MRI is the newest imaging method for the
diagnosis of AD. It is highly sensitive and specific
and does not require exposing the patient to con-
trast material. It is not ideal in ventilated or mon-
itored patients and is not widely available.
Treatment of aortic dissection
Treatment of AD is aimed at eliminating the
forces that favor progression of the dissection.
Prompt production of blood pressures can be
accomplished through use of sodium nitroprus-
side with a rate adjusted to achieve a systolic
blood pressure between 100 and 120 mm Hg
[44,47,57]. Concomitant use of a beta-blocker to
avoid reflex tachycardia secondary to the nitro-
prusside use is desirable to decrease further the
shear forces that promote propagation. A target
heart rate of 60 to 80 beats per minute is desirable
[44,57].
Patients who have acute dissection involving
the ascending aorta should receive rapid surgical
12 HARO
consultation. An exception might be isolated arch
dissection; most physicians consider this a surgical
entity. Richartz and colleagues [68] published the
international experience with isolated arch dissec-
tions. Of 989 patients, 92 (9%) had isolated arch
dissection. Of these, 39 (42%) were treated surgi-
cally, and 53 (58%) were treated medically. Thir-
ty-day mortality was 17% (23% with surgery and
13% with medical management) and therefore
was much lower than typical for type A dissec-
tions (35% with surgery and 65% with medical
management). Complications were the same. The
conclusion was that isolated arch dissections are
best managed conservatively. Distal ADs are in
general traditionally treated medically. Surgery is
indicated in distal dissections when there is evi-
dence of lower extremity or visceral ischemia, re-
nal failure, or paraplegia [47].
Percutaneous management of aortic dissection
Complicated type B dissections have been
subject to novel percutaneous therapies such as
percutaneous fenestration (restoring flow to the
true lumen by creating a tear in the dissection flap
between the true and false lumen) and percutane-
ous stent–graft placement. With these techniques
compromised flow can be restored in 90% of the
cases (range, 70%–100%). If a patient survives the
intervention, postoperative average 30-day mor-
tality is 10% (range, 0%–25%), and additional
surgical intervention is rarely needed [69].
Pulmonary embolism
Missed PE is a major source of malpractice
litigation in emergency medicine [70]. It is esti-
mated that the diagnosis of PE is missed 400,000
times annually, leading to 100,000 preventable
deaths [71]. Other studies have estimated that only
30% of PE is diagnosed ante mortem [72]. The mor-
tality rate for untreated PE is 18.4%dseven times
greater than that of appropriately treated PE
[73]. Certainly, failure to be diagnosed is the great-
est threat to the patient who has PE [74]. The chal-
lenges faced in the diagnosis of PE are similar to
those discussed for ACS and AD.
In general, the presentation of PE is non-
specific. Young patients who have excellent car-
diac reserve tend to have mild, transient, or no
symptoms at all [74]. Patients may complain of
chest pain which is typically sudden in onset and
pleuritic in nature. Dyspnea, palpitations, presyn-
cope, or syncope may also be presenting com-
plaints. Accurate diagnosis is clouded when
clinical presentations coexist with underlying
et al
obstructive lung disease, pneumonia, or underly-
ing congestive heart failure. In such cases, PE
can present with the symptoms of any of these
entities [75].
The clinical likelihood of a patient’s having PE
has long been estimated by implicit means.
Physician judgment is based on the patient’s
clinical presentation, history and physical evalua-
tion, and risk factor assessment. Studies, however,
have shown poor agreement among physicians
in estimating pretest probability of disease [76].
Physician experience affects pretest probability as-
sessment, with less experienced clinicians demon-
strating less ability to assign pretest probability
accurately [77]. Implicit assessment results in
a large group of patients being placed in the mod-
erate-risk category. Thus, clinical judgment has
yielded disappointing results in accurately deter-
mining the pretest probability of disease.
Therefore, clinical scoring systems have been
developed. Wells and colleagues [78] and Ander-
son and others [79] have created a seven-feature
bedside assessment tool to categorize patients as
having low, moderate, or high pretest probability
of PE. Even with Wells’ criteria, studies have
shown poor agreement among physicians on the
assignment of pretest probability and poor accu-
racy for the same [80]. Until a reliable, validated
clinical scoring system can be developed, the as-
signment of pretest probability of PE will remain
a challenge.
History and physical examination
As in the entities previously described, history
taking in a patient presenting with chest pain
suspicious for PE should focus on features of the
pain and associated symptoms. Particularly im-
portant in patients suspected of having PE is risk
factor assessment. Box 2 lists the many of risk fac-
tors that need to be considered in a patient who
has possible PE [74]. The physical examination
in a patient suspected of having PE should focus
on vital signs and pulmonary findings. Tachycar-
dia and tachypnea are classically described, but
the former is often absent in younger patients,
and the later is absent in up to 13% of patients
who have documented PE [74]. The physical ex-
amination may show pleural rub, rales, or findings
consistent with pulmonary consolidation. In sum-
mary, no physical finding is sensitive or specific
for the diagnosis of PE. Physical examination of-
fers no clues to the diagnosis in 28% to 58% of
cases [74].
 INITIAL EVALUATION OF CHEST PAIN 13

D-dimer assays
Box 2. Risk factors for pulmonary The advent of quantitative ELISA has improved
embolism the sensitivity of D-dimer assays tremendously [81].
D-dimer assays as a whole have poor specificity,
Inherited disorders (thrombophilias) with numerous conditions resulting in false-positive
Elevated individual clotting factor levels results. They are most useful when combined with
(VIII, IX, XI) other noninvasive imaging or clinical probability
Factor VLeiden mutation assessment scoring systems. In this way, a negative
Hyperhomocystinemia D-dimer can be incorporated into diagnostic algo-
Protein C, protein S, or antithrombin III rithms to withhold anticoagulation safely in pa-
deficiency tients who have a low pretest probability for PE
Prothrombin G20210A mutation [81]. Recent studies report the negative likelihood
ratios for a negative result on a quantitative rapid
Acquireddpersistent ELISA D-dimer make them as predictive as a nor-
Age mal V/Q scan or negative duplex ultrasound [82].
Antiphospholipid antibodies (lupus Eleven prospective clinical studies have evalu-
anticoagulant, anticardiolipin ated the role of D-dimer in excluding venous
antibody) thromboembolic disease. In patients who have
History of pulmonary embolism/deep a high clinical pretest probability of PE but
venous thrombosis a negative D-dimer, there is not enough evidence
History of superficial thrombophlebitis to support stopping an investigation for PE [73].
Hyperviscosity syndrome (polycythemia In contrast, it has become increasingly accepted
vera, malignant melanoma) that the diagnosis of PE is effectively ruled out
Immobilization (bedridden, paresis, or in patients who have a low clinical pretest proba-
paralysis) bility of disease and a negative quantitative rapid
Malignancy ELISA D-dimer [82]. Controversy still exists over
Medical conditions the extent of workup necessary in patients who
have a moderate pretest clinical probability and
Congestive heart failure a negative D-dimer assay.
Obesity One significant limitation affecting widespread
Nephrotic syndrome use of D-dimer is the numerous commercially
available assays that are not interchangeable,
Tobacco abuse differing significantly in sensitivity and negative
Acute myocardial infarction likelihood ratios [81]. Latex glutination assays and
Varicose veins whole-blood qualitative assays do not demon-
strate the same negative predictive value as quan-
Acquireddtransient titative ELISA assays and should not be
Central venous catheter/pacemaker incorporated in diagnostic algorithms similarly.
placement
Hormone replacement therapy V/Q scanning
Immobilizationdisolated extremity The Prospective Investigation of Pulmonary
Long distance travel/air travel Embolism Diagnosis investigators employed V/Q
Oral contraceptive pills scanning as the primary advanced imaging di-
Pregnancy and puerperium agnostic modality for patients suspected of having
Surgery PE [83]. Ventilation/perfusion scans are most
Trauma helpful when they are read as either normal or
high probability. Results of V/Q scans, however,
From Sadosty AT, Boie ET, Stead LG. Pul- fail to provide definitive indications for withhold-
monary embolism. Emerg Med Clin North ing or administering anticoagulation in up to 70%
Am 2003;21(2):363–84; with permission. of patients on whom the test is performed [74].

CT scanning
CT is widely available, noninvasive, increasingly
sensitive, and has the advantage of revealing
14 HARO
alternative diagnoses when PE is not found [84]. CT
can miss subsegmental PE. Some authors have ar-
gued that this finding is insignificant and that out-
come studies of rate of subsequent or recurrent
PE and death would be more a appropriate refer-
ence standard than comparisons to the standard
of angiographic subsegmental PE detection rates
[85]. Eleven such studies exist, both prospective
and retrospective, which demonstrate patient out-
come is not adversely affected when anticoagula-
tion is withheld based on a negative spiral CT [86].
Diagnostic evaluation summary
A review by Fedullo and Tapson [87] provides
rational guidance to the approach to the patient
suspected of having PE, using a combination of
pretest probability assessment, D-dimer assays,
and advanced imaging modalities to rule in or
rule out effectively the diagnosis of PE.
Patients who have a low pretest probability of
PE account for 25% to 65% of all patients
evaluated for PE, with subsequent diagnosis of
PE in 5% to 10% of cases [78,88–90]. For these
patients at low clinical probability, a negative
quantitative ELISA D-dimer effectively rules out
the diagnosis of PE.
Patients who have a high pretest probability
for PE comprise 10% to 30% of all patients
evaluated for PE, with subsequent diagnosis of PE
in 70% to 90% [78,88–90]. D-dimer has no signif-
icant role in this patient population, because a neg-
ative result does not rule out the presence of PE.
Spiral CT should be performed in these patients.
If CT is negative, duplex ultrasound or CT venog-
raphy of the lower extremities may be indicated. If
lower extremity studies are also negative in this
high-risk patient population, pulmonary angio-
gram is indicated to rule out the presence of PE.
Intermediate-risk patients comprise 25% to
65% of all patients examined for PE, with sub-
sequent diagnosis of PE in 25% to 45% [78,88–90].
References
[1] McCaig LF, Burt CW. National hospital ambula-
tory medical care survey: 2002 emergency depart-
ment summary. Adv Data 2004;(340):1–34.
[2] American Heart Association. Heart disease and
stroke statisticsd2004 update. Available at: http://
www.americanheart.org/presenter.jhtml?identifier¼
3000090. Accessed November 15, 2003.
[3] Antman EM, Anbe DT, Armstrong PW, et al. ACC/
AHA guidelines for the management of patients
with ST-elevation myocardial infarction; a report
of the American College of Cardiology/American
et al
Heart Association Task Force on Practice Guide-
lines (Committee to Revise the 1999 Guidelines for
the Management of Patients with Acute Myocardial
Infarction). J Am Coll Cardiol 2004;44(3):671–719.
[4] Bunch TJ, White RD, Gersh BJ, et al. Outcomes and
in-hospital treatment of out-of-hospital cardiac ar-
rest patients resuscitated from ventricular fibrilla-
tion by early defibrillation. Mayo Clin Proc 2004;
79(5):613–9.
[5] Canto JG, Zalenski RJ, Ornato JP, et al, for the Na-
tional Registry of Myocardial Infarction 2 Investiga-
tors. Use of emergency medical services in acute
myocardial infarction and subsequent quality of
care: observations from the National Registry of
Myocardial Infarction 2. Circulation 2002;106:
3018–23.
[6] Brown AL, Mann NC, Daya M, et al. Demographic,
belief, and situational factors influencing the deci-
sion to utilize emergency medical services among
chest pain patients: Rapid Early Action for Coro-
nary Treatment (REACT) study. Circulation 2000;
102:173–8.
[7] Becker L, Larsen MP, Eisenberg MS. Incidence of
cardiac arrest during self-transport for chest pain.
Ann Emerg Med 1996;28:612–6.
[8] Rogers WJ, Canto JG, Lambrew CT, et al. Tempo-
ral trends in the treatment of over 1.5 million
patients with myocardial infarction in the US from
1990 through 1999: the National Registry of Myo-
cardial Infarction 1, 2 and 3. J Am Coll Cardiol
2000;36:2056–63.
[9] Goff DC, Feldman HA, McGovern PG, et al, for the
Rapid Early Action for Coronary Treatment (RE-
ACT) Study Group. Prehospital delay in patients
hospitalized with heart attack symptoms in the
United States: the REACT trial. Am Heart J 1999;
138:1046–57.
[10] Welsh RC, Ornato J, Armstrong PW. Prehospital
management of acute ST-elevation myocardial in-
farction: a time for reappraisal in North America.
Am Heart J 2003;145:1–8.
[11] Steg PG, Goldberg RJ. Baseline characteristics,
management practices and in-hospital outcomes
of patients hospitalized with acute coronary syn-
dromes in the Global Registry of Acute Coronary
Events (GRACE). Am J Cardiol 2002;90(4):
358–63.
[12] Finnegan JR Jr, Hendrika Meischke H, Zapka JG,
et al. Patient delay in seeking care for heart attack
symptoms: findings from focus groups conducted
in five US regions. Prev Med 2000;31:205–13.
[13] Goff DC Jr, Mitchell P, Finnegan J, et al, for the RE-
ACT Study Group. Knowledge of heart attack
symptoms in 20 US communities. Results from the
Rapid Early Action for Coronary Treatment Com-
munity Trial. Prev Med 2004;38(1):85–93.
[14] Hochman JS, Sleeper LA, White HD, et al, for the
Should We Emergently Revascularize Occluded
Coronaries for Cardiogenic Shock (SHOCK)
 INITIAL EVALUATION OF CHEST PAIN
Investigators. One-year survival following early re-
vascularization for cardiogenic shock. JAMA 2001;
285:190–2.
[15] Gersh JB, Stone WG, White DH, et al. Pharmaco-
logical facilitation of primary percutaneous coro-
nary intervention for acute myocardial infarction.
JAMA 2005;293(8):979–86.
[16] Reimer KA, Lowe JE, Rasmussen MM, et al. The
wavefront phenomenon of ischemic cell death: 1.
Myocardial infarct size vs duration of coronary oc-
clusion in dogs. Circulation 1977;56:786–94.
[17] The European Myocardial Infarction Project
Group. Prehospital thrombolytic therapy in patients
with suspected acute myocardial infarction. N Engl J
Med 1993;329:383–9.
[18] Danchin N, Blanchard D, Steg GP, et al. Impact of
prehospital thrombolysis for acute myocardial in-
farction on 1-yr outcome. Circulation 2004;110:
1909–15.
[19] Brady WJ, Roberts D, Morris F. The nondiagnostic
ECG in the chest pain patient: normal and nonspe-
cific initial ECG presentations of acute MI. Am J
Emerg Med 1999;17(4):394–7.
[20] Lim SH, Sayre MR, Gibler WB. 2-D echocardiogra-
phy prediction of adverse events in ED patients with
chest pain. Am J Emerg Med 2003;21(2):106–10.
[21] Jones ID, Slovis CM. Emergency department evalu-
ation of the chest pain patient. Emerg Med Clin
North Am 2001;19(2):269–82.
[22] Boccardi L, Verde M. Gender differences in the clin-
ical presentation to the emergency department for
chest pain. Ital Heart J 2003;4(6):371–3.
[23] Kontos MC. Evaluation of the emergency depart-
ment chest pain patient. Cardiol Rev 2001;9(5):
266–75.
[24] Alpert JS, Thygesen K, Antman E, et al. Myocardial
infarction redefined: a consensus document of the
Joint European Society of Cardiology/American
College of Cardiology Committee for the redefini-
tion of myocardial infarction. J Am Coll Cardiol
2000;36:959–69.
[25] Wu AH, Apple FS, Gibler WB, et al. National Acad-
emy of Clinical Biochemistry Standards of Labora-
tory Practice: recommendations for the use of
cardiac markers in coronary artery diseases. Clin
Chem 1999;45:1104–21.
[26] Antman EM, Tanasijevic MJ, Thompson B, et al.
Cardiac specific troponin I levels to predict the risk
of mortality in patients with acute coronary syn-
dromes. N Engl J Med 1996;335:1342–9.
[27] Lindahl B, Toss H, Siegbahn A, et al, for the FRISC
Study Group. Markers of myocardial damage and
inflammation in relation to long-term mortality in
unstable coronary artery disease. N Engl J Med
2000;16(343):1139–47.
[28] Sabatine MS, Morrow DA, de Lemos JA, et al. Mul-
timarker approach to risk stratification in non-ST el-
evation acute coronary syndromes simultaneous
assessment of troponin I, C-reactive protein, and
15
B-type natriuretic peptide. Circulation 2002;105:
1760.
[29] Weaver WD, Simes RJ, Betriu A, et al. The most ef-
fective therapy for MI reperfusion in primary percu-
taneous coronary revascularization (PPCI). It is
superior to IFT (comparison of primary coronary
angioplasty and intravenous thrombolytic therapy
for acute myocardial infarction: a quantitative re-
view). JAMA 17;278(23):2093–8.
[30] Brodie BR, Stuckey TD, Wall TC, et al. Importance
of time to reperfusion for 30-day and late survival
and recovery of left ventricular function after pri-
mary angioplasty for acute myocardial infarction.
J Am Coll Cardiol 1998;32:1312–9.
[31] Brodie BR, Stone GW, Morice MC, et al, for the
Stent Primary Angioplasty in Myocardial Infarction
Study Group. Importance of time to reperfusion on
outcomes with primary coronary angioplasty for
acute myocardial infarction (results from the Stent
Primary Angioplasty in Myocardial Infarction Tri-
al). Am J Cardiol 2001;88:1085–90.
[32] Berger PB, Ellis SG, Holmes DR, et al. Relationship
between delay in performing direct coronary angio-
plasty and early clinical outcome in patients with
acute myocardial infarction: results from the Global
Use of Strategies to Open Occluded Arteries in
Acute Coronary Syndromes (GUSTO-IIb) trial. Cir-
culation 1999;100:14–20.
[33] Cannon CP, Gibson CM, Lambrew CT, et al. Rela-
tionship of symptom-onset-to-balloon time and
door-to-balloon time with mortality in patients un-
dergoing angioplasty for acute myocardial infarc-
tion. JAMA 2000;283:2941–7.
[34] Widimsky P, Groch L, Zelizko M, et al, on behalf of
the PRAGUE Study Group Investigators. Multi-
centre randomized trial comparing transport to pri-
mary angioplasty vs immediate thrombolysis vs
combined strategy for patients with acute myocar-
dial infarction presenting to a community hospital
without a catheterization laboratory The PRAGUE
Study. Eur Heart J 2000;21:823–31.
[35] De Luca G, Suryapranata H, Ottervanger JP, et al.
Time delay to treatment and mortality in primary an-
gioplasty for acute myocardial infarction: every min-
ute of delay counts. Circulation 2004;109:1223–5.
[36] Williams DO. Treatment delayed is treatment de-
nied. Circulation 2004;109:1806–8.
[37] Brass LM, Lichtman JH, Wang Y, et al. Intracranial
hemorrhage associated with thrombolytic therapy
for elderly patients with acute myocardial infarction:
results from the Cooperative Cardiovascular Pro-
ject. Stroke 2000;31:1802–11.
[38] Andersen HR, Nielsen TT, Rasmussen K, et al, for
the DANAMI- 2 Investigators. A comparison of
coronary angioplasty with fibrinolytic therapy in
acute myocardial infarction. N Engl J Med 2003;
349:733–42.
[39] Second International Study of Infarct Survival Col-
laborative Group (ISIS-2). Randomised trial of
16 HARO
intravenous streptokinase, oral aspirin, both, or nei-
ther among 17,187 cases of suspected acute myocar-
dial infarction: ISIS-2. Lancet 1988;2:349–60.
[40] Roux S, Christeller S, Lüdin E. Effects of aspirin on
coronary reocclusion and recurrent ischemia after
thrombolysis: a metaanalysis. J Am Coll Cardiol
1992;19:671–7.
[41] Meine TJ, Roe MT, Chen AY, et al, for the CRU-
SADE Investigators. Association of intravenous
morphine use and outcomes in acute coronary syn-
dromes: results from the CRUSADE Quality Im-
provement Initiative. Am Heart J 2005; 149.
[42] The MIAMI Trial Research Group. Metoprolol in
acute myocardial infarction: patient population.
Am J Cardiol 1985;56:10G–4G.
[43] Pfisterer M, Cox JL, Granger CB, et al. Atenolol use
and clinical outcomes after thrombolysis for acute
myocardial infarction: the GUSTO-I experience.
Global Utilization of Streptokinase and TPA (alte-
plase) for Occluded Coronary Arteries. J Am Coll
Cardiol 1998;32:634–40.
[44] Hals G. Acute thoracic aortic dissection: Current
evaluation and management. Emerg Med Rep
2000;21:1.
[45] Boersma E, Harrington RA, Moliterno DJ, et al.
Platelet glycoprotein IIb/IIIa inhibitors in acute cor-
onary syndromes: a meta-analysis of all major clini-
cal randomized trials. Lancet 2002;359:189–98.
[46] Cannon PC, Weintraub WS, Demopoulos LA, et al.
Comparison of early invasive and conservative strat-
egies in patients with unstable coronary syndromes
treated with the glycoprotein IIb/IIIa inhibitor tiro-
fiban. N Engl J Med 2000;25(344):1879–87.
[47] Salkin MS. Thoracic aortic dissection: avoiding fail-
ure to diagnose. ED Legal Letter 1997;8(11):107–18.
[48] Thoracic and abdominal aortic aneurysms. In:
Tintinalli JE, Krome RL, et al, editors. Emergency
medicineda comprehensive study guide. 3rd edi-
tion. New York: McGraw Hill; 1992. p. 1384.
[49] Pate JW, Richardson RL, Eastridge CE. Acute aor-
tic dissection. Am J Surg 1976;42:395–404.
[50] Hirst AE, Johns VJ, Kime SW, et al. Dissecting an-
eurysm of the aorta. A review of 505 cases. Medicine
1958;37:217–22.
[51] Spitell PC, Spitell JA, Joyce JW, et al. Clinical fea-
tures and differential diagnosis of aortic dissection:
experience with 236 cases (1980–1990). Mayo Clin
Proc 1993;68:642–51.
[52] Meszaros I, Morocz J, Szlavi J, et al. Epidemiology
and clinicopathology of aortic dissection: a popula-
tion-based longitudinal study over 27 years. Chest
2000;117:1271–8.
[53] Klompas M. Does this patient have an acute thoracic
aortic dissection? JAMA 2002;287(17):2262–72.
[54] Viljanen T. Diagnostic difficulties in aortic dissec-
tion: retrospective study of 89 surgically treated
patients. Ann Chir Gynaecol 1986;75:328–32.
[55] Hagan PG, Nienaber CA, Isselbacher EM, et al. The
international registry of acute aortic dissection
et al
(IRAD): new insights into an old disease. JAMA
2000;283:897–903.
[56] Januzzi JL, Isselbacher EM, Fatorri R, et al. Char-
acterizing the young patient with aortic dissection:
results from the international registry of aortic dis-
section (IRAD) [abstract 1081–154]. J Am Coll Car-
diol 2003;41:253A.
[57] Nienamber CA, Eagle KA. Aortic dissection: new
frontiers in diagnosis and management. Circulation
2003;108:628–35.
[58] Bossone E, Vincenzo R, Nienamber CA, et al. Use-
fulness of pulse deficit to predict in-hospital compli-
cations and mortality in patients with type A aortic
dissection. Am J Cardiol 2002;89:851–5.
[59] Katoh H, Suzuki T. A novel immunoassay of
smooth muscle myosin heavy chain in serum.
J Immanuol Methods 1995;185:57–63.
[60] Suzuki T, Trimarchi S, Smith D, et al. Early diagno-
sis of acute aortic dissection: identification of clinical
variables associated with early diagnosis and deter-
mination of the usefulness of biochemical diagnosis
as shown by the international registry of acute aortic
dissection (IRAD) database. Circulation 2004;
110(17):370.
[61] Shinohara T, Suzuki K, Okada M, et al. Soluble
elastin fragments in serum are elevated in acute aor-
tic dissection. Thromb Vasc Biol 2003;23:1839–44.
[62] Weber T, Hogler S, Auer J, et al. D-dimer in acute
aortic dissection. Chest 2003;123(5):1375–8.
[63] Bossone E, Mehta RH, Trimarchi S, et al. Coronary
involvement in patients with acute type A aortic dis-
section. J Am Coll Cardiol 2003;235A:1034–41.
[64] Yoshida S, Akiba H, Tamakawa M, et al. Thoracic
involvement of type A aortic dissection and intramu-
ral hematoma: diagnostic accuracydcomparison of
emergency helical CT and surgical findings. Radiol-
ogy 2003;228(2):430–5.
[65] Eagle KA, Quertermous T, Kritzer GA, et al. Spec-
trum of conditions initially suggesting acute aortic
dissection but with negative aortograms. Am J Car-
diol 1986;57(4):322–6.
[66] Kouchoukos NT, Dougenis DD. Surgery of the tho-
racic aorta. N Engl J Med 1997;336:1876–86.
[67] Nienaber CA, Spielmann RP, von Kodolitsch Y,
et al. Diagnosis of thoracic aortic dissection. Mag-
netic resonance imaging versus transesophageal
echocardiography. Circulation 1992;85(2):434–47.
[68] Richartz B, Schiller F, Bossone E, et al. Aortic arch
dissection as a distinct entity: lessons learned from
IRAD. Circulation 2002;106:473.
[69] Nienaber CA, Eagle KA. Aortic dissection: new
frontiers in the diagnosis and management. Part II:
therapeutic management and follow-up. Circulation
2003;108:772–8.
[70] Laack TA, Goyal DG. Pulmonary embolism: an un-
suspected killer. Emerg Med Clin North Am 2004;
22(4):961–83.
[71] Feied C. Pulmonary embolism. In: Rosen P,
Barkin R, Daniel DF, editors. Emergency medicine:
 INITIAL EVALUATION OF CHEST PAIN
concepts and clinical practice. St. Louis (MO): Mos-
by-Year Book; 1998. p. 1770–2.
[72] Morgenthaler TI, Ryu JH. Clinical characteristics of
fatal pulmonary embolism in a referral hospital.
Mayo Clin Proc 1995;70:417–24.
[73] Carson JL, Kelley MA, Duff A, et al. The clinical
course of pulmonary embolism. N Engl J Med
1992;326(19):1240–5.
[74] Sadosty AT, Boie ET, Stead LG. Pulmonary embo-
lism. Emerg Med Clin North Am 2003;21(2):363–84.
[75] Goldhaber SZ. Pulmonary embolism. Lancet 2004;
363(7417):1295–305.
[76] Jackson RE, Rudoni RR, Pascual R. Emergency
physician (EP) assessment of the pre-test probability
of pulmonary embolism (PE). Acad Emerg Med
1999;6:437.
[77] Kline JA, Wells PS. Methodology for a rapid proto-
col to rule out pulmonary embolism in the emergency
department. Ann Emerg Med 2003;42(2):266–75.
[78] Wells PS, Ginsberg JS, Anderson DR, et al. Use of
a clinical model for safe management of patients
with suspected pulmonary embolism. Ann Intern
Med 1998;129(12):997–1005.
[79] Anderson DR, Wells PS, Stiell I, et al. Thrombosis in
the emergency department: use of a clinical diagnosis
model to safely avoid the need for urgent radiological
investigation. Arch Intern Med 1999;159(5):477–82.
[80] Sanson BJ, Lijmer JG, MacGillavry MR, et al. Com-
parison of a clinical probability estimate and two
clinical models in patients with suspected pulmo-
nary embolism. ANTELOPE-Study Group. Thromb
Haemost 2000;83(2):199–203.
[81] Frost SD, Brotman DJ, Michota FA. Rational use
of D-dimer measurement to exclude acute venous
thromboembolic disease. Mayo Clin Proc 2003;
78(11):1385–91.
17
[82] Stein PD, Hull RD, Patel KC, et al. D-dimer for the
exclusion of acute venous thrombosis and pulmo-
nary embolism: a systematic review. Ann Intern
Med 2004;140(8):589–602.
[83] PIOPED Investigators. Value of the ventilation/per-
fusion scan in acute pulmonary embolism. Results of
the Prospective Investigation of Pulmonary Embo-
lism Diagnosis (PIOPED). JAMA 1990;263(20):
2753–9.
[84] Perrier A, Howarth N, Didier D, et al. Performance
of helical computed tomography in unselected out-
patients with suspected pulmonary embolism. Ann
Intern Med 2001;135(2):88–97.
[85] Wolfe TR, Hartsell SC. Pulmonary embolism: mak-
ing sense of the diagnostic evaluation. Ann Emerg
Med 2001;37(5):504–14.
[86] Schoepf UJ, Goldhaber SZ, Costello P. Spiral com-
puted tomography for acute pulmonary embolism.
Circulation 2004;109(18):2160–7.
[87] Fedullo PF, Tapson VF. Clinical practice. The eval-
uation of suspected pulmonary embolism. N Engl J
Med 2003;349(13):1247–56.
[88] Khorasani R, Gudas TF, Nikpoor N, et al. Treat-
ment of patients with suspected pulmonary embo-
lism and intermediate probability lung scans: is
diagnostic imaging underused? AJR 1997;169(5):
1355–7.
[89] Miniati M, Pistolesi M, Marini C, et al. Value of per-
fusion lung scan in the diagnosis of pulmonary
embolism: results of the Prospective Investigative
Study of Acute Pulmonary Embolism Diagnosis
(PISA-PED). Am J Respir Crit Care Med 1996;
154(5):1387–93.
[90] Miniati M, Monti S, Bottai M. A structed clinical
model for predicting the probability of pulmonary
embolism. Am J Med 2003;114(3):173–9.
 Cardiol Clin 24 (2006) 19–35

Pathogenesis and Early Management

of Non–ST-segment Elevation Acute
Coronary Syndromes
Tomas H. Ayala, MDa,*, Steven P. Schulman, MDb,c
a
Division of Cardiology, University of Maryland School of Medicine, 22 South Greene Street,
Baltimore, MD 21201, USA
b
Johns Hopkins University School of Medicine, 600 North Wolfe Street, Baltimore, MD 21287, USA
c
Coronary Care Unit, the Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287, USA

In 2001, there were more than 5.6 million visits
to United States emergency departments because
of chest pain or related symptoms [1]. Of these,
nearly 1.4 million patients were admitted with acute
coronary syndromes (ACS) [2]. The term ‘‘ACS’’
has evolved into a useful descriptor of a collection
of symptoms associated with acute myocardial is-
chemia caused by sudden restriction of coronary
blood flow. It is used to describe collectively acute
myocardial infarction (AMI), both with and with-
out ST-segment elevation, and unstable angina.
By definition, then, it is a deliberately broad term
that encompasses a range of related clinical condi-
tions with varying degrees of host response, injury,
prognosis, and, therefore, treatment.
In its very lack of specificity, the term ‘‘ACS’’
reflects the difficulty often encountered in differ-
entiating between myocardial infarction (MI)
(especially without ST segment elevation) and
unstable angina at the time of presentation. As
reflected by its name, non–ST-segment elevation
MI (NSTEMI) is defined as presence of myocar-
dial damage (typically detected by biochemical
markers of myocardial necrosis) in the absence of
ST segment elevation, true posterior MI, or new
left bundle branch block on 12-lead ECG [3]. Its
pathogenesis, clinical presentation, even angio-
graphic appearance are virtually indistinguishable
from those of unstable angina, save for evidence
* Corresponding author.
E-mail address: tayala@medicine.umaryland.edu
(T.H. Ayala).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.006
of myocardial necrosis, which is usually not ap-
parent at initial presentation [4,5]. Therefore, the
combined entity of unstable angina/NSTEMI is
often referred to as non–ST-segment elevation
ACS (NSTE ACS). There are subtle but impor-
tant differences in the pathogenesis of NSTE
ACS and ST-segment elevation MI (STEMI);
more crucial are differences in treatment, specifi-
cally in urgency of reperfusion therapy.
This article focuses on pathogenesis and early
management of NSTE ACS. Because atheroscle-
rotic coronary artery disease represents the typical
substrate for NSTE ACS, the discussion would be
incomplete without an overview of the develop-
ment of the atherosclerotic plaque. The article then
reviews the specific pathologic pathways leading
to NSTE ACS, with particular attention to dis-
ruption of the vulnerable plaque and ensu-
ing nonocclusive thrombosis. Finally, it addresses
appropriate early management based on the cur-
rent understanding of the underlying pathophysi-
ology and on a review of available clinical data.
Pathogenesis of coronary artery disease
In 1986, Ross [6] proposed the ‘‘response to in-
jury’’ hypothesis of atherogenesis implicating en-
dothelial injury and subsequent intimal smooth
muscle cell proliferation and inflammation as the
initiating events in atherosclerosis. This article
presents an overview of the proposed mechanisms
of endothelial injury and the molecular and cellu-
lar responses to such injury that lead to plaque
cardiology.theclinics.com
20 AYALA & SCHULMAN

formation, progression, and ultimately rupture. It
is useful, first, to review the histologic characteris-
tics of the developing plaque at various stages.
Histology of atherosclerotic lesions
Stary and colleagues [7] have proposed a classi-
fication scheme of atherosclerotic lesions based on
distinct morphologic characteristics. They de-
scribe two types of precursor or early lesions,
two types of advanced lesions, and a linking stage
between the two groups. The first three lesion
types are characterized by being focal and lacking
evidence of intimal disruption. More advanced le-
sions demonstrate intimal disruption and even
modification of underlying media and adventitia.
Initial or type I lesions are microscopic collec-
tions of isolated macrophage foam cells (MFC)d
macrophages containing lipid dropletsdwithin
arterial intima; these lesions have been found in
as many as 45% of infants aged 8 months [8]. By
the time of puberty, type II lesions are present [9].
These display a more organized pattern of MFC,
arranged in adjacent layers and accompanied by
smooth muscle cells that also contain lipid drop-
lets, as well as small quantities of dispersed lipid
droplets in the extracellular matrix (ECM) [7].
The preatheroma or type III lesion represents
a transitional stage between early lesions and ad-
vanced atheroma. Its morphologic composition
is marked by the presence of extracellular lipid
pools between layers of smooth muscle cells.
This lesion type still contains layers of MFC and
non–lipid-containing macrophages [7]. The lipid
pools eventually coalesce to form the lipid core
that is characteristic of atheroma or type IV lesion
[10]. In this lesion, the lipid core disrupts intimal
structure but generally does not narrow the arte-
rial lumen; instead there is compensatory eccentric
expansion of the external boundary of the vessel,
a process known as positive arterial remodeling
(Fig. 1) [11,12]. Type V lesions also have a lipid
core, now surrounded by layers of fibrous tissue
[10]. These lesions are quite heterogeneous but
generally are composed of varying proportions
of collagen-rich ECM (also with proteoglycans
and elastic fibers), a lipid core, smooth muscle
cells, inflammatory cells (MFC and T lympho-
cytes), thrombi, and calcium deposits [10,13]. It
is this advanced lesion that is generally prone to
disruption.
Endothelium and endothelial injury
More than being simply a selectively permeable
vascular inner layer, the endothelium is a dynamic
organ with paracrine and autocrine functions that
has a central role in regulating vascular hemody-
namics and vascular hemostasis [14]. Endothelial
regulation of vasomotion is performed by a bal-
ance in synthesis and release of vasodilators such
as nitric oxide and prostacyclin, as well as vaso-
constrictors such as endothelin-1 and platelet acti-
vating factor (PAF). The synthesis and release of
these substances is controlled locally, modulated
by a variety of biologically active molecules and
mediators that allow an appropriate response to
mechanical or chemical injury. In addition to their
vasomotor effects, nitric oxide and prostacyclin
are also central in inhibiting platelet activation,
thereby preventing thrombosis. When disrupted,
however, endothelial cells become procoagulant
in an effort to stop blood flow and restore vascular
integrity [14].

EEM expansion with relative lumen preservation Lumen shrinkage

Normal vessel Mild CAD Moderate CAD Severe CAD

Fig. 1. Positive remodeling. In early atherosclerosis, plaque growth is compensated for by eccentric expansion of the
external elastic membrane (EEM) with relative preservation of luminal area. As atherosclerosis progresses, the EEM
does not expand further, and plaque growth results in luminal shrinkage. CAD, coronary artery disease. (Adapted
from Glagov S, Weisenberg E, Zarins CK, et al. Compensatory enlargement of human atherosclerotic coronary arteries.
N Eng J Med 1987;316:1371; with permission.)
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS 21

A number of mechanisms of endothelial injury
leading to dysfunction have been proposed.
Chronic minimal injury caused by turbulent flow
at bending points, chronic exposure to elevated
low-density lipoprotein (LDL) cholesterol levels,
smoking, diabetes, hypertension, oxidative stress,
and infectious factors may all contribute to alter-
ations in the function of endothelial cells (Fig. 2)
[4,6]. This altered response includes a decrease in
nitric oxide biosynthesis, increase in endothelin-1
production, and activation of the transcription fac-
tor nuclear factor kB [6,15]. This transcription
factor has been implicated in the regulation of in-
flammatory cytokines such as tumor necrosis
factor a (TNF-a), interleukin (IL)-1b, and macro-
phage colony-stimulating factor; chemokines such
as macrophage chemoattractant protein 1 (MCP-
1); and adhesion molecules such as intercellular
adhesion molecule 1 and vascular cell adhesion
molecule-1 [16]. These factors serve to attract
blood monocytes [17], which migrate into the sub-
endothelial space and transform into macro-
phages. In turn, these activated macrophages
internalize oxidized LDL to become MFC and re-
lease mitogens that lead to smooth muscle cell pro-
liferation [6]. They also amplify the effect of
nuclear factor kB by releasing MCP-1, IL-1b,
TNF-a, and various other inflammatory cyto-
kines, perpetuating the inflammatory cell recruit-
ment process [17,18]. Plaque macrophages are
stimulated to produce cytokines and collagen-
degrading enzymes by a variety of factors, includ-
ing angiotensin II [19], oxidized LDL [20], and
the very inflammatory cytokines they express
[21]. TNF-a, IL-1b (both expressed primarily by
macrophages), and interferon gamma (IFN-g)
are overexpressed in atherosclerotic lesions but
not in normal arteries [22]. IFN-g is uniquely ex-
pressed by the activated T-helper type 1 lympho-
cyte, induced primarily by a combination of IL-
18 and IL-12, both produced by macrophages
[18,23]. These inflammatory cells and cytokines
play a central role in the development of athero-
sclerosis and in the pathogenesis of ACS.
Pathogenesis of the acute coronary syndromes
Myocardial ischemia occurs when myocardial
oxygen demand exceeds supply. In ACS this
occurrence is typically because of a sudden re-
duction in coronary blood flow, and therefore of
oxygen supply, relative to oxygen demand. Braun-
wald [24] recognizes five general circumstances in
which this scenario can occur. The most common,

Flow turbulence Hypertension

High LDL Diabetes
Smoking Oxidative stress

Endothelial injury

+
+
NO NF B ET-1
PGI2 PAF

MCP-1 IL1β
ICAMs TNFα
VCAMs Other cytokines

+ +
Blood monocyte Matrix
Tissue monocyte Macrophage
breakdown
Plaque growth, vulnerability, rupture, and thrombosis

Fig. 2. Endothelial injury results in decreased vasodilator production, increase in vasoconstrictors, and increase in
proinflammatory mediators that ultimately result in plaque growth and vulnerability. ET-1, endothelin-1; ICAMs, inter-
cellular adhesion molecules; IL1b, interleukin 1 beta; MCP-1, macrophage chemoattractant protein-1; NO, nitric oxide;
PAF, platelet-activating factor; PGI2, prostacyclin; TNFa, tumor necrosis factor alpha; VCAMs, vascular cell adhesion
molecules.
22 AYALA & SCHULMAN
and most clinically important, involves disruption
of the advanced atherosclerotic plaque with for-
mation of a thrombus that partially or completely
occludes the arterial lumen, thus reducing myocar-
dial perfusion. A second cause is progressive nar-
rowing of the arterial lumen caused by advancing
chronic atherosclerosis. Vasospasm, or sudden
contraction of vascular smooth muscle cells, re-
sults in a sudden narrowing of a focal segment of
coronary artery, producing impaired myocardial
perfusion. Arterial inflammation may lead to arte-
rial narrowing or to plaque destabilization with
ensuing plaque rupture and thrombosis. Finally,
any systemic condition that increases myocardial
oxygen demand (eg, fever, thyrotoxicosis), de-
creases overall coronary blood flow (eg, systemic
hypotension), or decreases oxygen delivery (eg,
profound anemia) can cause myocardial ischemia.
Because of its clinical significance, plaque disrup-
tion and ensuing arterial thrombosis are the focus
of this discussion.
The vulnerable plaque
Traditionally, the vulnerable plaque has been
defined as one that is prone to rupture [25]. Recent
pathologic findings suggest that, although rupture
is the leading cause of arterial thrombosis, other
processes can also result in thrombus formation.
The comprehensive term of ‘‘high-risk plaque’’
provides a more complete description of the vul-
nerable plaquedone that may result in thrombus
through rupture, cap erosion, or calcified nodules
[26]. A significant limitation in the understanding
of the pathology behind the ACS in humans is
that histologic sections are available only after
death. Most patients who have ACS do not die,
and it is unclear if the pathologic process is iden-
tical in survivors.
Plaque rupture
Under appropriate circumstances the protec-
tive fibrous cap covering an atheromatous lesion
can rupture, exposing the thrombogenic lipid core
to coagulation factors and platelets in the blood-
stream, which in turn leads to thrombosis [25,27].
Autopsy studies have shown that plaque rupture
is associated with 60% to 80% of coronary
thrombi in patients who have died of AMI
[25,28]. Plaques prone to rupture seem to share
common morphologic characteristics: a thin fi-
brous cap (!65 mm thick), an abundant popula-
tion of macrophages, a relative scarcity of
smooth muscle cells, and a soft extracellular lipid
core [22,29–31]. Plaques of this type are refer-
red to as ‘‘thin-cap fibroatheromas.’’ The clinical
significance of thin-cap fibroatheromas was con-
firmed by an angioscopic study of coronary ar-
teries in 63 patients demonstrating a relationship
between yellow, lipid-rich plaques with thin fi-
brous caps and acute ischemic events [32]. Cir-
cumstances that lead to rupture are complex and
are not completely understood; however, a combi-
nation of inflammatory factors and mechanical
stress seems to be central in the development of
plaque disruption.
Why does the fibrous cap fracture? Rather
than being static, the fibrous cap is a metabolically
active structure, constantly undergoing remodel-
ing by means of synthesis and degradation of
collagen and other ECM components (Fig. 3) [33].
Fibrillar interstitial collagen, synthesized by
smooth muscle cells, is the principal determinant
of the fibrous cap’s biomechanical strength [27].
Inflammatory cytokines and inflammatory cells,
specifically, macrophages and T lymphocytes,
seem to exert significant control over the level of
collagen and thus the cap’s mechanical integrity.
Dysregulation of normal ECM remodeling, medi-
ated by inflammation, represents an attractive (al-
though as yet unproven) hypothesis to explain
aspects of fibrous cap rupture: an imbalance fa-
voring breakdown of collagen over its synthesis
results in structurally compromised areas of the fi-
brous cap.
Given that they play a central role in the
progression of atherosclerosis, it is no surprise
that macrophages and activated T lymphocytes
have been found in abundance at the sites of
plaque rupture [28,34]. These cells and the cyto-
kines and enzymes they produce are responsible
for detrimental effects to ECM composition.
When appropriately activated by cytokines and
other signals, macrophages degrade ECM through
expression of a variety of matrix metalloprotei-
nases (MMPs) that catalyze the critical initial steps
of collagen breakdown. Certain MMPs, notably
MMP-2 (also known as 72kD gelatinase), are pres-
ent constitutively in nonatherosclerotic vessels and
function in normal ECM metabolism; addition-
ally, MMP-2 is secreted mainly in an inactive
complex with its endogenous inhibitors (tissue in-
hibitor of metalloproteinase -1 and -2, TIMP-1
and -2) [35]. In contrast, atherosclerotic plaques
demonstrate overexpression of activated MMP-1
(interstitial collagenase-1), MMP-3 (stromelysin),
MMP-9 (92kD gelatinase), and MMP-13 (intersti-
tial collagenase-3), especially at the shoulder
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
A
Matrix
Smooth muscle cells
synthesis
B Mat
synt rix
hesis
Ma
degr trix
adat
ion
MMP-1
MMP-3
IL1β
TNFα TIMP
+
CD-40 lig
T lymph
IFN-γ
+ +
Fig. 3. (A). Balanced atherosclerotic cap homeostasis. (B) Activated macrophage foam cells (MFC) stimulated by T lym-
phocytes, inflammatory cytokines, oxidized LDL (ox-LDL), and circumferential hemodynamic stress (HS) overexpress
matrix metalloproteinases (MMP) which leads to matrix degradation and cap thinning See text for details. IL1b, inter-
leukin-1 beta; IFN-g, interferon gamma; TIMP, tissue inhibitor of metalloproteinase.
region of the fibrous cap (the junction of the pla-
que with normal vessel) and surrounding the lipid
core, where ruptures most often occur [22,35].
Macrophage expression of MMPs is induced by
a variety of factors, including IL-1b, TNF-a, and
IFN-g [36], by activation by T lymphocytes
through CD40 ligand [37], and by circumferential
hemodynamic stress [38]. Induction of MMP is
also accomplished by neutrophils and oxidized
LDL. As is the case with macrophages and acti-
vated T lymphocytes, neutrophil infiltration has
been noted in ruptured plaques [39]. Neutrophils
express MMP-8, or neutrophil collagenase, which
preferentially degrades type I collagen [40]. Oxi-
dized LDL, present within the lipid core, stimu-
lates the expression of MMP-9 by macrophages
and inhibits the production of the MMP inhibitor
TIMP-1 [41]. Even smooth muscle cells, when
stimulated by IL-1b and IFN-g, can degrade
ECM by releasing activated MMPs [35] and the
elastolytic proteases cathepsin S and K [42].
Protease-mediated degradation is comple-
mented by reduced ECM production in the thin-
cap fibroatheromas. IFN-g has been shown to
inhibit collagen gene expression in smooth muscle
cells [27]. Additionally, the combination of IL-1b,
23
Matrix Constitutive matrix
metalloproteinases
degradation
(MMP-2)
Thinned fibrous cap
MMP-9 + Ox-LDL
MMP-13
Lipid Core
MFC
HS
+
TNF-a, and IFN-g promotes smooth muscle cell
apoptosis, which results in decreased collagen
synthesis [43].
Inflammation in ACS may not be limited to
a strictly local phenomenon. Evidence of neutro-
phil activation by reduced blood levels of neutro-
phil myeloperoxidase was found in samples from
the aortic and great coronary veins of patients
who had unstable angina [44]. This activation was
found whether the culprit lesion was in the left an-
terior descending artery (which is drained selec-
tively by the great coronary vein) or in the right
coronary artery (which is not drained by the great
coronary vein). Additionally, evidence of neutro-
phil activation was absent in patients who had
chronic stable angina or variant angina. The pres-
ence of generalized coronary artery inflammation
in the setting of ACS and the contribution of in-
flammation to the pathogenesis of ACS have led
some to postulate that multiple unstable coronary
lesions should be evident during an acute coro-
nary event. In an intravascular ultrasound study
of 24 patients who had initial ACS, there was ev-
idence of multiple plaque ruptures at sites other
than the culprit lesion, suggesting a generalized
level of plaque instability [45]. Other evidence
24 AYALA & SCHULMAN
suggests that the extent of inflammation is not just
regional but systemic. In a study of 36 patients ad-
mitted with AMI, the group that had preinfarc-
tion unstable angina had higher levels of the
systemic inflammatory markers C-reactive pro-
tein, serum amyloid protein A, and IL-6, than
did those with unheralded infarctions [46].
Mechanical factors can affect plaque’s suscep-
tibility to inflammatory assault. High circumfer-
ential stress has been associated with increased
expression of MMP-1 [38]. A significant determi-
nant of this stress is the eccentric, soft, lipid core,
the presence of which has been shown to concen-
trate high circumferential stresses at the shoulder
regions of the plaque [47]. Most plaques rupture
at these regions of elevated tensile stress [48]. Shear
stresses can also have an impact on plaque disrup-
tion. A study of carotid plaques at autopsy demon-
strated that areas of plaque exposed to high flow
and high shear forces had an abundance of plaque
macrophages relative to the density of smooth
muscle cells compared with areas of plaque ex-
posed to low shear forces and low flow [49]. There
seems to be increased susceptibility to rupture in
areas with a higher concentration of inflammatory
cells [50]. The thinness of the fibrous cap itself can
have negative mechanical repercussions. A mathe-
matical model of arterial vessels suggests that for
a given luminal area, decreasing thickness of the fi-
brous cap is associated with increased circumferen-
tial stress [51]. Furthermore, repetitive exposure to
pulsatile flow at these areas of high circumferential
stress may result in plaque fatigue [51]. Another
important aspect of the influence of mechanical
stress on plaque rupture is positive arterial remod-
eling. Intravascular ultrasound studies in human
coronaries have shown that positive remodeling
is more common in patients who have unstable an-
gina, whereas negative remodeling (in which the
plaque encroaches on the lumen) is more common
in patients who have chronic stable angina [52].
Positive remodeling also correlates with a larger
lipid core and increased macrophage content,
both of which promote plaque vulnerability [53].
Additionally, positive remodeling may influence
plaque distensibility, that is, the softness of the pla-
que, which in turns makes it more prone to rupture
[54]. Finally, positive arterial remodeling has been
correlated with decreased density of collagen fibers
[55]. The finding that positive arterial remodeling
can increase plaque vulnerability while preserving
luminal area may account for the observation
that, in a majority of patients presenting with
ACS, recent prior coronary angiography fails to
demonstrate culprit lesion stenoses exceeding
70% [56,57].
The thin cap fibroatheroma is uniquely suscep-
tible to rupture. In the setting of inflammation and
matrix degradation, a structurally unstable fibrous
cap exposed to both transient and steady mechan-
ical disturbances leads to increases in tensile stress
and ultimately to fracture. A variety of triggers of
plaque rupture have been proposed, including air
pollution [58], marijuana use [59], cocaine use [60],
sexual activity, and emotional or physical stress
[61], among others. The exact mechanism by which
these activities prompt rupture in the susceptible
plaque is not well understood.
Other mechanisms of plaque vulnerability
Although plaque rupture is the most important
precipitating event for thrombosis in ACS, other
mechanisms of thrombosis without rupture have
been proposed. In one series of 125 cases of sudden
coronary death caused by confirmed acute epicar-
dial thrombosis, only 74 (59%) had histologic
evidence of plaque rupture; 45 (36%) showed
thrombus overlying a smooth muscle cell–rich,
proteoglycan-rich plaque denuded of endothelial
cells, and 6 (5%) showed a calcified nodule at the
site of thrombus [26]. Additionally, compared with
thrombosis associated with plaque rupture, cases
with thrombosis caused by erosion were more
likely to be younger than 50 years, women, and
smokers [26,62]. These findings were confirmed
in an autopsy series of 291 patients who died of
AMI and had confirmed coronary thrombosis. In
25% of cases, thrombosis was caused by plaque
erosion [63].
Although some have proposed repeated focal
vasospasm as the underlying cause for the endo-
thelial denudation seen in plaque erosion, this
process has not been confirmed [26]. The observed
absence of endothelial cells in these lesions has
generated the attractive hypothesis that severe en-
dothelial apoptosis as a response to mechanical or
chemical injury leads to thrombosis. Cell mem-
brane microparticles derived from apoptotic cells
in atherosclerotic plaque have been shown to in-
crease local levels of thrombogenic (TF) [64].
The same study showed, however, that most of
these microparticles were derived from apoptotic
macrophages and T lymphocytes, not from endo-
thelial cells; it is unclear if endothelial apoptosis is
always associated with thrombosis.
The most unusual lesion associated with
thrombosis is the calcified nodule, a superficial
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
fibrous-rich calcified lesion that erupts from the
intima into the lumen [26]. The processes leading
up to this lesion, or the way in which it induces
thrombosis, are not known.
Arterial thrombosis
Plaque rupture or erosion both result in
thrombus formation. This process is dynamic,
and it is the balance of systemic and local
procoagulant and fibrinolytic factors that ulti-
mately determines the extent of thrombus forma-
tion and thus the clinical outcome [4].
Thrombus formation
The mechanisms behind thrombus formation
associated with plaque disruption are better un-
derstood than those associated with plaque
erosion. Davies [65] describes three stages of
thrombosis. The first is characterized by platelet
adhesion within the thrombogenic lipid core, the
second by extension of the thrombus into the lu-
men with an increase in fibrin content and distal
microembolization of activated platelets, and the
third, by a growing thrombus composed of loose
fibrin networks and entrapped red cells. This pro-
gression of stages is not deterministic and can be
affected by host factors and clinical intervention.
Once plaque rupture occurs, the disrupted
edges of the fibrous cap as well as the lipid core
are exposed to blood-borne coagulation factors
and platelets. Disruption of the normal endothe-
lial layer transforms the endothelium from
its constitutive antithrombotic state to a procoa-
gulant state. It does so mainly by increased
production of PAF and von Willebrand factor
(vWF), a large multimeric protein with binding
sites for subendothelial collagen [14]. Potentiated
by arterial shear stress, platelets adhere to exposed
subendothelium by platelet membrane glycopro-
tein Ib/IX-V complexed with vWF [66]. Adherent
platelets then become activated under the influ-
ence of various mediators. Activation involves
a shape change in the platelet, as well as release
of alpha granules and dense granules containing
potent platelet activation agents, including adeno-
sine diphosphate and serotonin. Activation is also
enhanced by circulating epinephrine and throm-
bin and by subendothelial collagen and vWF
[67]. Additionally, activated platelets synthesize
and release thromboxane A2, a potent vasocon-
strictor and platelet activator, which is increased
in patients who have unstable angina [68]. The
25
final step of activation is aggregation: expression
of the platelet membrane Gp IIb/IIIa receptor al-
lows platelet–platelet interactions whereby adher-
ent platelets recruit circulating platelets to form
an occlusive thrombus. Platelet aggregation is me-
diated by fibrinogen or vWF, either of which can
bind to the Gp IIb/IIIa receptor, thus linking the
platelets. The platelet thrombus is then stabilized
by a fibrin mesh, the result of the simultaneously
ongoing coagulation cascade [67].
Fernandez-Ortiz and colleagues [69] demon-
strated that, at least in vitro, it is the lipid core
that acts as the most significant substrate for the
formation of the platelet thrombus. The exact
source of the thrombogenic properties of the lipid
core is not known. It has been suggested that lip-
ids themselves and acellular debris found in the
core may activate hemostasis [70]. This suggestion
is an incomplete explanation and overlooks the
significant role of TF. TF is a transmembrane gly-
coprotein that initiates the extrinsic clotting cas-
cade. Its relationship to ACS was demonstrated
in a study showing that TF is present in signifi-
cantly greater amounts in the tissue of patients
who have unstable angina than in those who
have stable angina [71]. Despite its thrombogenic-
ity, however, the lipid core is the source of surpris-
ingly little TF. In fact, in patients who have
unstable angina, TF is located predominantly in
the cellular component of the plaque [71]. How
can this apparent discrepancy be explained?
Inflammatory cytokines, activated T cells, and
oxidized LDL in the lipid core induce TF expres-
sion by plaque macrophages, smooth muscle cells,
and endothelial cells [37,71,72]. At the time of pla-
que rupture, most TF seems to be associated with
apoptotic cell fragments (mainly from macro-
phages and endothelial cells) that form part of
the necrotic debris in the lipid core [64]. Thus, al-
though there is no intrinsic TF expression within
the lipid core, TF activity is found there in the set-
ting of plaque rupture. Using an in vitro perfusion
model, Toschi and colleagues [72] demonstrated
the presence of active TF-VIIa complex within
the lipid core, as well as a correlation between
this complex and platelet adhesion.
TF, of course, is also responsible for initiating
the plasma coagulation system that leads to the
structural fibrin network of the thrombus. TF
forms a high-affinity complex with coagulation
factors VII and VIIa; this latter complex activates
factors IX and X, and this activation in turn leads
to thrombin generation and thrombosis [73]. Not
only does thrombin serve to activate platelets, it
26 AYALA & SCHULMAN
converts fibrinogen to fibrin and activates factor
XIII, which in turn stabilizes the fibrin clot [67].
Factors that mediate the extent of thrombosis
Not all thrombi or all plaque ruptures result in
MI; they may, however, contribute to the pro-
gression of atherosclerotic disease. In his series of
25 cases of sudden cardiac death caused by acute
coronary thrombosis, Falk [74] found that 81% of
the thrombi had a layered appearance, suggesting
alternating episodes of thrombus formation and
fragmentation preceding the fatal infarction. Other
investigators have suggested that healing of re-
peated plaque ruptures leads to progression of ath-
erosclerosis: the thrombus organizes and heals by
migration and proliferation of smooth muscle
cells. This healing fibromuscular response incor-
porates the mural thrombus and worsens luminal
stenosis [75]. Finally, even though fibrotic lesions
lack a thrombogenic lipid core, erosion of these
plaques has been associated with thrombus forma-
tion [26]. A number of factors, both local and sys-
temic, have been implicated in determining the
extent to which an arterial thrombus develops at
the site of endothelial disruption.
Increased shear stress produces a conforma-
tional change in the platelet Gp Ib/IX-V receptor
and vWF resulting in more avid platelet adhesion
[66]. The clinical importance of this observation
was demonstrated by a study that demonstrated
increased platelet adhesion at the apex of a dis-
rupted lesion; furthermore, it showed that the ex-
tent of platelet adhesion corresponded directly
with the extent of stenosis following plaque rup-
ture [76]. The same group also showed that in-
creasing the extent of plaque disruption resulted
in a larger, more stable, platelet thrombus [77].
This finding is hardly surprising, because more ex-
tensive plaque injury leads to greater exposure of
blood to thrombogenic TF within the lipid core
and to endothelial platelet activators such as
PAF and vWF and thus results in more wide-
spread activation of the coagulation cascade. Fi-
nally, the presence of a mural thrombus from
prior plaque ruptures, such as those described by
Falk [74], seems to contribute to extensive throm-
bosis. Not only does a mural thrombus increase
the degree of luminal stenosis, it provides for
a rich source of potent platelet activation in the
form of thrombin. This scenario was observed in
a pig model of vessel injury with mural thrombus
in which the direct thrombin inhibitor hirudin
prevented extension of thrombus, whereas aspirin
and heparin did not [78]. In summary, the extent
of stenosis, plaque disruption, size of the lipid
core, and the presence of a mural thrombus are lo-
cal factors that seem to increase thrombogenicity
in the disrupted plaque.
Several systemic factors, including elevated
cholesterol levels, increased catecholamine levels,
smoking, and hyperglycemia, have been associ-
ated with increased thrombogenicity [4]. LDL-
lowering therapy with 3-hydroxy-3-methylglutaryl
coenzyme A (HMG-CoA) reductase inhibitors
(statins) has been shown to decrease blood throm-
bogenicity and thrombus growth [79]. Although
some of this effect may be caused by pleiotropic
effects of statins, it is clear that LDL, and in par-
ticular oxidized LDL, plays a role in increased
thrombotic activity. LDL promotes TF expres-
sion in plaque macrophages and expression of
prothrombotic plasminogen activator inhibitor-1
in endothelial cells [14,71]. Elevated catechol-
amines are associated with increased platelet
activation; activities that increase circulating cate-
cholamines, such as cigarette smoking, [80] co-
caine use [60], and increased physical activity
[61], have been associated with triggering acute
cardiovascular events. Smoking seems to be of
particular importance in young patients who
have underlying plaque erosion, reflecting the sig-
nificance of increased systemic thrombogenicity in
these patients whose lesions lack a lipid core [62].
Hyperglycemia associated with poor diabetic con-
trol results in increased blood thrombogenicity,
caused in part by increased platelet activa-
tion and also by activation of the TF pathway
[81,82]. Other metabolic abnormalities may also
increase blood thrombogenicity. Activation of the
renin-angiotensin-aldosterone system increases
activity of plasminogen activator inhibitor-1, me-
diated by angiotensin II [27]. Lipoprotein (a) is
structurally similar to plasminogen and may com-
petitively inhibit its fibrinolytic activity [4]. It is
also likely that several proinflammatory cytokines
promote both atherosclerosis and plaque rupture
and thrombogenesis, suggesting a role for systemic
inflammation in the development of ACS [27].
Early management of the non–ST-segment
elevation acute cardiac syndromes
The most recent update of the American
College of Cardiology/American Heart Associa-
tion guidelines for NSTE ACS highlights several
areas of management, including early risk
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
stratification, early use of aspirin and clopidogrel,
indications for early invasive strategy, and use of
Gp IIb/IIIa inhibitors (abciximab, tirofiban, and
eptifibatide) based on risk. Additionally, the
guidelines recommend an algorithm for use of
specialized chest pain units to manage patients
who have possible and low-risk ACS while
avoiding unnecessary hospitalizations [3]. This
section discusses the early management of NSTE
ACS, that is, management during the patient’s ini-
tial stay in the chest pain unit.
Risk assessment
Patient management in ACS begins with risk
stratification based on clinical history, physical
examination, and ancillary data such as 12-lead
ECG and biomarkers of cardiac injury. Based on
this initial assessment, patients can be categorized
into one of four groups: noncardiac chest pain,
chronic stable angina, possible ACS, and definite
ACS. Those who have possible or definite ACS are
observed in a chest pain unit until the diagnosis is
confirmed or discounted [3].
Given the wide range of presentations sub-
sumed within ACS, risk assessment is essential in
determining the choice of therapy, specifically the
use of Gp IIb/IIIa inhibitors and of early invasive
therapy. Gp IIb/IIIa inhibitors bind the platelet
receptor and prevent platelet aggregation [83].
Their use has been shown to reduce the frequency
and duration of ischemia [83] and the rate of the
combined endpoint of death, reinfarction, or re-
fractory angina in patients who have NSTE ACS
[84]. Early invasive therapy consists of coronary
angiography and angiographically guided revascu-
larization within 48 hours of symptom onset. This
approach, compared with conservative therapy of
medical management and noninvasive evaluation
of ischemia, has also been shown to reduce rates
of the combined endpoint of death, reinfarction,
or 6-month rehospitalization [85]. Although these
trials suggest an overall benefit for aggressive ther-
apy in NSTE ACS, there is such a wide range of
clinical presentations that the risk–benefit ratio
for these therapies may also vary greatly, depend-
ing on individual patient’s risk.
Obviously, the presence of hemodynamic in-
stability, pulmonary edema, prolonged rest chest
pain, or sustained ventricular tachycardia por-
tends a poor short-term outcome and should be
addressed with early institution of aggressive
therapy [3]. Many patients who do not have these
27
clear high-risk features upon presentation may
benefit from aggressive therapy, however. To pre-
dict risk in these patients a variety of schema has
been proposed. Among the easiest to use and best-
validated is the Thrombolysis in Myocardial
Infarction (TIMI) risk score (Box 1) [86]. A score
exceeding 4 in this equally weighted seven-item
scale allows identification of the patients who ben-
efit most from early invasive and Gp IIb/IIIa in-
hibitor therapies based on short- and long-term
risks [87]. Many investigators, citing the direct re-
lationship between rising cardiac troponin levels
and poor outcomes, have suggested that elevated
troponin levels alone should be considered
markers of high risk [3].
Treatment
Initial treatment for NSTE ACS includes con-
tinuous ECG monitoring for ischemia and possi-
bility of arrhythmia. Additionally, anti-ischemic
therapy should be initiated promptly, tailored to
the level of patient risk. Finally, the mainstay of
initial therapy is directed at addressing arterial
thrombus: antiplatelet and anticoagulant therapy
(Table 1).
Box 1. Thrombolysis in Myocardial
Infarction (TIMI) risk score prediction
variables
Age greater than 65 years
Three or more risk factors for coronary
artery disease
Known coronary artery stenosis greater
than 50%
ST-segment deviation on presenting
ECG
Two or more episodes of angina within
the preceding 24 hours
Use of aspirin within the preceding 7
days
Elevated serum cardiac biomarker levels
Adapted from Gluckman TJ, Sachdev M,
Schulman SP, et al. A simplified approach to
the management of non-ST elevation acute
coronary syndromes. JAMA 2005;293(3):349;
with permission.
28 AYALA & SCHULMAN

Table 1
Summary of cardiovascular disease management in NSTE ACS
Intervention Agent(s)/Treatment Modalities
Antiplatelet therapy Aspirin
ADP receptor antagonist
(clopidogrel)
GP IIb/IIIa inhibitor (abciximab,
eptifibatide, tirofiban)
Anticoagulation Unfractionated heparin
Low molecular weight heparin
ACE inhibition No clear preferred agent
Angiotensin receptor No clear preferred agent
blockade
Beta blockade No clear preferred agent
Blood pressure control ACE inhibitors and beta
blockers first line
Abbreviations: ACE, angiotensin-converting enzyme; ADP, adenosine diphosphate; BP, blood pressure; CABG,
coronary artery bypass graft; LMWH, low molecular weight heparin; PCI, percutaenous coronary interventions.
Adapted from Gluckman TJ, Sachdev M, Schulman SP, et al. A simplified approach to the management of non-ST
elevation acute coronary syndromes. JAMA 2005;293(3):349; with permission.
Comments
All patients indefinitely
Initially with 162–325 mg followed by
75–160 mg daily thereafter
All patients, unless anticipated need for
urgent CABG surgery or within
5 days of electively scheduled CABG surgery
Duration up to 1 year
All patients with continuing ischemia, an
elevated troponin level, a TIMI risk
score O 4, or anticipated PCI
Avoid abciximab if PCI is not planned
Alternative to LMWH for patients
managed with early invasive strategy
Preferred anticoagulant if managed conservatively
Alternative to unfractioned heparin for patients
managed with an early invasive strategy
Avoid if creatinine clearance
! 60 mL/min (unless anti-Xa levels
are to be followed) or CABG
surgery within 24 hours
All patients with left ventricular
systolic dysfunction (ejection
fraction ! 40%), heart failure, hypertension,
or other high-risk features
All patients intolerant to ACE inhibitors
Avoid combination therapy with
ACE inhibitors acutely, but consider
in patients with chronic left
ventricular systolic dysfunction
(ejection fraction ! 40%) and heart failure
All patients
Goal BP at least ! 130/85 mm Hg

Anti-ischemic therapy
Beta blockade reduces cardiac workload and
oxygen demand, and the use of beta-blockers in the
acute setting has been associated with a 13%
relative risk reduction in progression to AMI [87].
The current recommendation is for prompt initia-
tion of beta-blocker therapy unless contraindicated
(eg, by marked first-degree atrioventricular block,
second- or third-degree atrioventricular block, his-
tory of asthma, left ventricular dysfunction with
congestive heart failure) [3]. A recommended initial
regimen for high-risk patients is intravenous meto-
prolol, 5 mg every 5 minutes to a total of 15 mg; if
tolerated, 25 to 50 mg oral metoprolol should be
administered 15 minutes after the last intravenous
dose and given every 6 hours for the first 48 hours.
Patients at lower risk may be managed with oral
therapy [3]. For patients in whom beta-blockers
are contraindicated, the nondihydropyridine cal-
cium antagonists verapamil and diltiazem may be
used. One trial using intravenous verapamil in
3447 patients who had suspected MI showed
a trend toward lower risk of death or nonfatal
MI [88]. Guidelines recommend the use of long-act-
ing nondihydropyridine calcium antagonists either
in the setting of beta-blocker intolerance or if ische-
mic pain persists despite full use of beta blockade
and nitrate administration [3].
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
Nitroglycerin is an endothelium-independent
arterial vasodilator that increases myocardial
blood flow through coronary vasodilation and
decreases myocardial oxygen demand through
venodilation and reduced preload [89]. In the ab-
sence of contraindications such as hypotension or
concurrent use of sildenafil, tadafil, or vardenafil
within 24 hours, intravenous nitroglycerin is rec-
ommended for the patient who has ongoing chest
pain despite three doses of sublingual nitroglyc-
erin and administration of beta-blockers [3]. The
purpose of nitroglycerin is to relieve ischemic
pain, because it has not been demonstrated to im-
prove survival [90,91]. American College of Cardi-
ology guidelines currently recommend the use of
intravenous morphine sulfate at doses of 1 to
5 mg if nitrate and beta-blocker therapy fail to
relieve pain and in patients who have acute pul-
monary edema, with care taken to avoid hypoten-
sion [3]. Recent registry data regarding 57,039
patients treated for NSTE ACS at 443 United
States hospitals, however, showed that the
17,003 patients who received intravenous mor-
phine within 24 hours of presentation had a higher
adjusted risk of death (odds ratio, 1.48; 95% CI,
1.33–1.64), even though patients receiving mor-
phine were also more likely to receive acute evi-
dence-based therapies and to be treated by
a cardiologist. These data suggest that morphine
sulfate in acute NSTE ACS may not be as safe
as previously thought and that randomized trials
are needed [92].
HMG-CoA reductase inhibitors have been
shown to decrease blood thrombogenicity [79],
and early aggressive therapy with these agents in
the setting of AMI reduced the risk of a combined
endpoint of death, MI, unstable angina rehospi-
talization, revascularization, and stroke [93].
This therapy was instituted 10 days after presenta-
tion, however. It is unclear if administration of
statins in the acute setting is of additional benefit.
Antiplatelet therapy
Given the central role of platelet activation and
aggregation in thrombus formation, it is reason-
able to expect that antiplatelet therapy would be
at the core of NSTE ACS management. Recom-
mended agents include aspirin, which inhibits
platelet aggregation and activation primarily by
inhibiting thromboxane A2 synthesis, clopidogrel,
which inhibits platelet activation by blocking the
adenosine diphosphate receptor, and the Gp IIb/
IIIa inhibitors.
29
Aspirin has been shown to reduce the rates of
death or nonfatal MI by 51% (5% versus 10.1%;
P ¼ .0005) in patients who have unstable angina,
without an increase in gastrointestinal symptoms
[94]. Current guidelines recommend an initial
dose of 162 to 325 mg followed by a daily dose
of 75 to 160 mg [3].
In the Clopidogrel in Unstable Angina to
Prevent Recurrent Ischemic Events (CURE)
study, 12,562 patients who had NSTE ACS were
randomly assigned to clopidogrel (300 mg loading
dose followed by 75 mg daily for 9 months) and
aspirin (75–325 mg daily) versus aspirin alone [95].
The group receiving clopidogrel had a relative risk
reduction of 20% in the combined endpoint of
cardiovascular death, nonfatal MI, or stroke
(9.3% versus 11.4%; P ! .001). The benefit was
driven primarily by a reduction in risk of subse-
quent MI (5.2% versus 6.7%; relative risk, 0.77;
95% CI, 0.67–0.89). The benefit of clopidogrel
also holds for patients undergoing percutaneous
coronary interventions (PCI). The PCI-CURE
study randomly assigned 2658 patients who had
NSTE ACS and who were scheduled for PCI to
pretreatment with clopidogrel plus aspirin versus
aspirin alone; treatment was initiated a median
of 6 days before PCI and continued for 8 months
[96]. Patients who received clopidogrel experi-
enced a 30% relative risk reduction in the com-
bined endpoint of cardiovascular death, nonfatal
MI, or urgent target vessel revascularization at
30 days and a 31% relative risk reduction of car-
diovascular death or nonfatal MI at study end.
Although clopidogrel is clearly beneficial in
NSTE ACS, there are caveats to its use. In the
CURE trial, major bleeding was significantly
more common in the clopidogrel arm (3.7% ver-
sus 2.7%; P ¼ .001), although there was no signif-
icant increase in life-threatening bleeding [95].
Additionally, use of clopidogrel within 5 days of
surgery can increase bleeding risk [97]. Current
guidelines recommend prompt initiation of clopi-
dogrel, with use for at least 1 month and up to 9
months, withheld 5 to 7 days before coronary ar-
tery bypass grafting. Additionally, patients who
have intolerance or hypersensitivity to aspirin
should be given clopidogrel instead [3].
Benefit from Gp IIb/IIIa inhibitors is closely
tied to patient risk and therefore to concomitant
use of an early invasive strategy. A meta-analysis of
six trials including 31,402 patients who had NSTE
ACS and who were managed with conservative
strategy found that benefit of Gp IIb/IIIa inhibitor
use was limited to those who had elevated troponin
30 AYALA & SCHULMAN
levels or who needed early revascularization, that
is, higher-risk patients [98]. Furthermore, in pa-
tients who had NSTE ACS and who were treated
conservatively, the Gp IIb/IIIa inhibitor abciximab
demonstrates no benefit and may be harmful [99].
In patients undergoing PCI, however, there is clear
benefit to Gp IIb/IIIa inhibitor use. A meta-analy-
sis of eight trials (14,644 patients) of any Gp IIb/
IIIa inhibitor versus placebo in patients undergo-
ing PCI showed a reduction of both MI rate (5%
versus 7.8%; P ! .001) and need for urgent revas-
cularization (3.4% versus 5.9%; P ! .001) [100].
Subgroup analysis in study showed that only abcix-
imab resulted in a significant reduction in AMI af-
ter PCI. This effect may reflect the timing of PCI
relative to drug administration. Data suggest that
abciximab is the Gp IIb/IIIa inhibitor of choice
for PCI within 4 hours of presentation, whereas tir-
ofiban and eptifibatide should be used if PCI is de-
ferred [87]. Based on these data, current guidelines
recommend administration of eptifibatide or tirofi-
ban (but not abciximab) to high-risk patients (ie,
those who have elevated troponin levels, a TIMI
risk score greater than 4, continuing ischemia)
who have NSTE ACS and are undergoing conser-
vative therapy [3,87]. Because the decision to pro-
ceed to PCI and the timing thereof are not always
clear at the time of initial presentation, the choice
of agent in this setting may be deferred until the
time of PCI [3].
In summary, all patients who have NSTE ACS
should receive aspirin and clopidogrel on pre-
sentation, unless there is anticipated need for
urgent surgical revascularization. Gp IIb/IIIa in-
hibitors are clearly recommended in high-risk
patients and those referred to PCI (ie, those with
high-risk clinical features, TIMI score greater than
4, or elevated troponin levels). Unfortunately,
there are no large-scale studies that evaluate the
incremental benefit of combined use of all three
proposed antiplatelet therapies versus strategies
using only two agents.
Anticoagulation
Heparin potentiates thrombin inactivation by
circulating antithrombin III, thus preventing
thrombus extension [101]. In a meta-analysis of six
trials of patients who had unstable angina treated
with unfractionated heparin (UFH) for 2 to
5 days plus aspirin versus aspirin alone, there was
a 33% relative risk reduction in the incidence of
death or MI (7.9% versus 10.4%; P ¼ .06) [102].
Unfortunately, UFH requires frequent monitoring
of the activated partial thromboplastin time (aPTT)
and has been associated with delays in achieving ap-
propriate anticoagulation, an aPTT 1.5 to 2.5 times
control [103]. Low molecular weight heparin
(LMWH) has more predictable pharmacokinetics,
is more readily bioavailable, and is administered
subcutaneously [103]. A recent meta-analysis of
six trials encompassing 21,946 patients sought to
compare the efficacy and safety of the LMWH
enoxaparin at 1 mg/kg every 12 hours and UFH
in patients who had NSTE ACS. At 30 days, there
was no statistically significant difference in mortal-
ity and a small difference in the combined endpoint
of death or MI favoring enoxaparin (10.1% versus
11.0%; 95% CI, 0.83–0.99), without significant dif-
ference in major bleeding [104]. The analysis
showed benefit of LMWH over UFH in patients
treated conservatively. Current guidelines recom-
mend the use of LMWH in addition to aspirin in pa-
tients who have NSTE ACS unless coronary artery
bypass grafting is planned within 24 hours [3]. For
patients who have renal insufficiency or in whom
early invasive strategy is planned, UFH is probably
a better choice [87].
Another parenteral anticoagulant, the direct
thrombin inhibitor hirudin, has been studied in
NSTE ACS. In the trial by the Organisation
to Assess Strategies for Ischemic Syndromes
(OASIS-2), 10,141 patients who had NSTE ACS
were randomly assigned to receive a 72-hour
infusion of UFH or hirudin [105]. At 7 days those
in the hirudin group had a lower (but not statisti-
cally significant) rate of the combined endpoint of
death or new MI (3.6% versus 4.2%; relative risk,
0.84; 95% CI, 0.69–1.02, P ¼ .077). Most of the
benefit occurred during the drug infusion period,
suggesting that the outcome of the therapies equal-
izes with time. A meta-analysis of 11 randomized
trials consisting of 35,790 patients compared the
use of any direct thrombin inhibitor (hirudin, bi-
valirudin, argatroban, efegatran, or inogatran)
with UFH in patients who had ACS. At 30 days,
there was a lower rate of death or MI in the group
treated with direct thrombin inhibitors (7.4% ver-
sus 8.2%; relative risk, 0.91, 95% CI, 0.84–0.99,
P ¼ .02) [106]. The use of direct thrombin inhibi-
tors along with aspirin and clopidogrel or Gp
IIb/IIIa inhibitors in NSTE ACS has not been
studied in large-scale trials; thus their incremental
benefit is not known. Furthermore, hirudin has
been shown to increase bleeding risks compared
with heparin [105]. Although more studies are ex-
pected, routine use of direct thrombin inhibitors in
NSTE ACS is not currently recommended [3].
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
Patients in the chest pain unit who have
suspected NSTE ACS should be given UFH if
early PCI is planned or if severe renal insufficiency
(creatinine clearance ! 60 mL/minute) is present.
Otherwise, LMWH should be used in addition to
aspirin and clopidogrel. The role for direct throm-
bin inhibitors is unclear at present.
Summary
NSTE ACS is a clinically significant problem.
Endothelial dysfunction triggered by traditional
cardiovascular risk factors (and perhaps by other
as yet unidentified risks) in the susceptible host
leads to the formation and development of athero-
sclerotic plaque. Inflammatory mediators and
mechanical stresses contribute to plaque rupture
by disrupting the protective fibrous cap. In about
25% of patients who have ACS, typically those
who are younger, female, or smokers, plaque
erosion seems to be the main underlying pathologic
mechanism. Endothelial alteration, inflammation,
or exposure of the lipid core results in the release of
TF, vWF, and PAF. The release of these factors
leads to platelet activation and aggregation as well
as to the formation of a fibrin clot, resulting in
arterial thrombosis that occludes the vessel. A
variety of factors, including circulating catechol-
amines, LDL levels, blood glucose levels, and
systemic thrombogenic factors, can affect the
extent and stability of the thrombus, thereby
determining whether the occlusion is complete
and fixed, labile and nonocclusive (NSTE ACS),
or clinically silent resulting in a mural thrombus
and plaque growth. The acute treatment of NSTE
ACS is directed at interrupting the prothrombotic
environment surrounding the ruptured plaque;
thus, antiplatelet agents such as aspirin, clopidog-
rel, and glycoprotein IIb/IIIa receptor antagonists,
as well as anticoagulants such as heparin, are the
mainstays of early therapy.
References
[1] McCaig LF, Burt CW. National Hospital Ambula-
tory Medical Care Survey: 2001 emergency depart-
ment summary. Advance data from Vital and
Health Statistics # 335. Hyattsville (MD): National
Center for Health Statistics; 2003.
[2] Kozak LJ, Owings MF, Hall MJ. National Hospi-
tal Discharge Survey: 2001 annual summary with
detailed diagnosis and procedure data. National
Center for Health Statistics. Vital Health Stat
2004;13:156.
31
[3] Braunwald E, Antman EM, Beasley JW, et al.
ACC/AHA 2002 guideline update for the manage-
ment of patients with unstable angina and non–ST-
segment elevation myocardial infarction: a report
of the American College of Cardiology/American
Heart Association Task Force on Practice Guide-
lines (Committee on the Management of Pa-
tients With Unstable Angina). 2002. Available
at: http://www.acc.org/clinical/guidelines/unstable/
unstable.pdf. Accessed February 1, 2005.
[4] Fuster V, Badimon L, Badimon JJ, et al. The path-
ogenesis of coronary artery disease and the acute
coronary syndromes. N Engl J Med 1992;326
(4,5):242–50; 310–7.
[5] DeWood MA, Stifter WF, Simpson CS, et al. Cor-
onary arteriographic findings soon after non-Q-
wave myocardial infarction. N Engl J Med 1986;
315(7):417–23.
[6] Ross R. The pathogenesis of atherosclerosisdan
update. N Engl J Med 1986;314(8):488–500.
[7] Stary HC, Chandler AB, Glagov S, et al. A defini-
tion of initial, fatty streak, and intermediate lesions
of atherosclerosis. Circulation 1994;89(5):2462–78.
[8] Stary HC. Macrophages, macrophage foam cells,
and eccentric intimal thickening in the coronary
arteries of young children. Atherosclerosis 1987;
64(2–3):91–108.
[9] Stary HC. Evolution and progression of athero-
sclerotic lesions in coronary arteries of children
and young adults. Arteriosclerosis 1989;9(Suppl I):
I-19–32.
[10] Stary HC, Chandler AB, Dinsmore RE, et al. A
definition of advanced types of atherosclerotic
lesions and a histological classification of athero-
sclerosis. Circulation 1995;92(5):1355–74.
[11] Glagov S, Weisenberg E, Zarins CK, et al. Com-
pensatory enlargement of human atherosclerotic
coronary arteries. N Engl J Med 1987;316(22):
1371–5.
[12] Losordo DW, Rosenfield K, Kaufman J, et al. Fo-
cal compensatory enlargement of human arteries in
response to progressive atherosclerosis. Circulation
1994;89(6):2570–7.
[13] Davies MJ. Stability and instability: two faces of
coronary atherosclerosis. The Paul Dudley White
Lecture 1995. Circulation 1996;94(8):2013–20.
[14] Cines DB, Pollak ES, Buck CA, et al. Endothelial
cells in physiology and in the pathophysiology of
vascular disorders. Blood 1998;91(10):3527–61.
[15] Kinlay S, Libby P, Ganz P. Endothelial function
and coronary artery disease. Curr Opin Lipidol
2001;12(4):383–9.
[16] Barnes PJ, Karin M. Nuclear factor-kappa B: a piv-
otal transcription factor in chronic inflammatory
diseases. N Engl J Med 1997;336(15):1066–71.
[17] Nelken NA, Coughlin SR, Gordon D, et al.
Monocyte chemoattractant protein-1 in human
atheromatous plaques. J Clin Invest 1991;88(4):
1121–7.
32 AYALA & SCHULMAN
[18] Mallat Z, Corbaz A, Scoazec A, et al. Expression of
interleukin-18 in human atherosclerotic plaques
and relation to plaque instability. Circulation
2001;104(14):1598–603.
[19] Schieffer B, Schieffer E, Hilfiker-Kleiner D, et al.
Expression of angiotensin II and interleukin 6 in
human coronary atherosclerotic plaques: potential
implications for inflammation and plaque instabil-
ity. Circulation 2000;101(12):1372–8.
[20] Rosenfeld ME, Yla-Herttuala S, Lipton BA, et al.
Macrophage colony-stimulating factor mRNA
and protein in atherosclerotic lesions of rabbits
and humans. Am J Pathol 1992;140(2):291–300.
[21] Uyemura K, Demer LL, Castle SC, et al. Cross-
regulatory roles of interleukin (IL)-12 and IL-10
in atherosclerosis. J Clin Invest 1996;97(9):2130–8.
[22] Sukhova GK, Schonbeck U, Rabkin E, et al. Evi-
dence for increased collagenolysis by interstitial
collagenases-1 and -3 in vulnerable human athero-
matous plaques. Circulation 1999;99(19):2503–9.
[23] Libby P. The molecular bases of the acute coronary
syndromes. Circulation 1995;91(11):2844–50.
[24] Braunwald E. Unstable angina: an etiologic ap-
proach to management. Circulation 1998;98(21):
2219–22.
[25] Davies MJ, Thomas A. Thrombosis and acute cor-
onary artery lesions in sudden cardiac ischemic
death. N Engl J Med 1984;310(18):1137–40.
[26] Virmani R, Kolodgie FD, Burke AP, et al. Lessons
from sudden coronary death: a comprehensive
morphological classification scheme for atheroscle-
rotic lesions. Arterioscler Thromb Vasc Biol 2000;
20(5):1262–75.
[27] Libby P. Current concepts of the pathogenesis of
the acute coronary syndromes. Circulation 2001;
104(3):365–72.
[28] van der Wal AC, Becker AE, van der Loos CM,
et al. Site of intimal rupture or erosion of throm-
bosed coronary atherosclerotic plaques is charac-
terized by an inflammatory process irrespective of
the dominant plaque morphology. Circulation
1994;89(1):36–44.
[29] Burke AP, Farb A, Malcom GT, et al. Coronary
risk factors and plaque morphology in men with
coronary disease who died suddenly. N Engl J
Med 1997;336(18):1276–82.
[30] Davies MJ, Richardson PD, Woolf N, et al. Risk of
thrombosis in human atherosclerotic plaques: role
of extracellular lipid, macrophage, and smooth
muscle cell content. Br Heart J 1993;69(5):377–81.
[31] Kolodgie FD, Virmani R, Burke AP, et al. Patho-
logic assessment of the vulnerable human coronary
plaque. Heart 2004;90(12):1385–91.
[32] Thieme T, Wernecke KD, Meyer R, et al. Angio-
scopic evaluation of atherosclerotic plaques:
validation by histomorphologic analysis and asso-
ciation with stable and unstable coronary syn-
dromes. J Am Coll Cardiol 1996;28(1):1–6.
[33] Libby P. The molecular bases of the acute coronary
syndromes. Circulation 1995;91(11):2844–50.
[34] Moreno PR, Falk E, Palacios IF, et al. Macro-
phage infiltration in acute coronary syndromes:
implications for plaque rupture. Circulation 1994;
90(2):775–8.
[35] Galis ZS, Sukhova GK, Lark MW, et al. Increased
expression of matrix metalloproteinases and matrix
degrading activity in vulnerable regions of human
atherosclerotic plaques. J Clin Invest 1994;94(6):
2493–503.
[36] Libby P, Sukhova GK, Lee RT, et al. Cytokines
regulate vascular functions related to stability of
the atherosclerotic plaque. J Cardiovasc Pharmacol
1995;25(Suppl 2):S9–12.
[37] Mach F, Schonbeck U, Bonnefoy JY, et al. Activa-
tion of monocyte/macrophage functions related to
acute atheroma complication by ligation of CD40.
Circulation 1997;96(2):396–9.
[38] Lee RT, Schoen FJ, Loree HM, et al. Circumferen-
tial stress and matrix metalloproteinase 1 in human
coronary atherosclerosis. Implications for plaque
rupture. Arterioscler Thromb Vasc Biol 1996;
16(8):1070–3.
[39] Naruko T, Ueda M, Haze K, et al. Neutrophil infil-
tration of culprit lesions in acute coronary syn-
dromes. Circulation 2002;106(23):2894–900.
[40] Herman MP, Sukhova GK, Libby P, et al. Expres-
sion of neutrophil collagenase (matrix meta-
lloproteinase-8) in human atheroma: a novel
collagenolytic pathway suggested by transcriptional
profiling. Circulation 2001;104(16):1899–904.
[41] Xu XP, Meisel SR, Ong JM, et al. Oxidized
low-density lipoprotein regulates matrix metallo-
proteinase-9 and its tissue inhibitor in human
monocyte-derived macrophages. Circulation 1999;
99(8):993–8.
[42] Sukhova GK, Shi GP, Simon DI, et al. Expression
of the elastolytic cathepsins S and K in human
atheroma and regulation of their production in
smooth muscle cells. J Clin Invest 1998;102(3):
576–83.
[43] Geng YJ, Wu Q, Muszynski M, et al. Apoptosis of
vascular smooth muscle cells induced by in vitro
stimulation with interferon-gamma, tumor necrosis
factor-alpha, and interleukin-1 beta. Arterioscler
Thromb Vasc Biol 1996;16(1):19–27.
[44] Buffon A, Biasucci LM, Liuzzo G, et al. Wide-
spread coronary inflammation in unstable angina.
N Engl J Med 2002;347(1):5–12.
[45] Rioufol G, Finet G, Ginon I, et al. Multiple athero-
sclerotic plaque rupture in acute coronary syn-
drome: a three-vessel intravascular ultrasound
study. Circulation 2002;106(7):804–8.
[46] Liuzzo G, Baisucci LM, Gallimore JR, et al. En-
hanced inflammatory response in patients with pre-
infarction unstable angina. J Am Coll Cardiol
1999;34(6):1696–703.
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
[47] Richardson PD, Davies MJ, Born GV. Influence of
plaque configuration and stress distribution on fis-
suring of coronary atherosclerotic plaques. Lancet
1989;2(8669):941–4.
[48] Cheng GC, Loree HM, Kamm RD, et al. Distribu-
tion of circumferential stress in ruptured and stable
atherosclerotic lesions: a structural analysis with
histopathological correlation. Circulation 1993;
87(4):1179–87.
[49] Dirksen MT, van der Wal AC, van den Berg FM,
et al. Distribution of inflammatory cells in athero-
sclerotic plaques relates to the direction of flow.
Circulation 1998;98(19):2000–3.
[50] Lendon CL, Davies MJ, Born GV, et al. Athero-
sclerotic plaque caps are locally weakened when
macrophage density is increased. Atherosclerosis
1991;87(1):87–90.
[51] Loree HM, Kamm RD, Stringfellow RG, et al.
Effects of fibrous cap thickness on peak circumfer-
ential stress in model atherosclerotic vessels. Circ
Res 1992;71(4):850–8.
[52] Schoenhagen P, Ziada KM, Kapadia SR, et al. Ex-
tent and direction of arterial remodeling in stable
versus unstable coronary syndromes: an intravas-
cular ultrasound study. Circulation 2000;101(6):
598–603.
[53] Varnava AM, Mills PG, Davies MJ. Relationship
between coronary artery remodeling and plaque
vulnerability. Circulation 2002;105(8):939–43.
[54] Takano M, Mizuno K, Okamatsu K, et al. Me-
chanical and structural characteristics of vulnera-
ble plaques: analysis by coronary angioscopy and
intravascular ultrasound. J Am Coll Cardiol
2001;38(1):99–104.
[55] Lafont A, Durand E, Samuel JL, et al. Endothelial
dysfunction and collagen accumulation: two inde-
pendent factors for restenosis and constrictive
remodeling after experimental angioplasty. Circu-
lation 1999;100(10):1109–15.
[56] Giroud D, Li JM, Urban P, et al. Relation of the
site of acute myocardial infarction to the most se-
vere coronary arterial stenosis prior to angiogra-
phy. Am J Cardiol 1992;69(8):729–32.
[57] Ambrose JA, Tannenbaum MA, Alexopoulos
D, et al. Angiographic progression of coronary
artery disease and the development of myocar-
dial infarction. J Am Coll Cardiol 1988;12(1):
56–62.
[58] Peters A, Dockery DW, Muller JE, et al. Increased
particulate air pollution and the triggering of
myocardial infarction. Circulation 2001;103(23):
2810–5.
[59] Mittleman MA, Lewis RA, Maclure M, et al. Trig-
gering myocardial infarction by marijuana. Circu-
lation 2001;103(23):2805–9.
[60] Mittleman MA, Mintzer D, Maclure M, et al. Trig-
gering of myocardial infarction by cocaine. Circu-
lation 1999;99(21):2737–41.
33
[61] Muller JE. Circadian variation and triggering of
acute coronary events. Am Heart J 1999;137(4 Pt
2):S1–8.
[62] Farb A, Burke AP, Tang AL, et al. Coronary
plaque erosion without rupture into a lipid core:
a frequent cause of coronary thrombosis in sud-
den coronary death. Circulation 1996;93(7):
1354–63.
[63] Arbustini E, Dal Bello B, Morbini P, et al. Plaque
erosion is a major substrate for coronary thrombo-
sis in acute myocardial infarction. Heart 1999;
82(3):269–72.
[64] Mallat Z, Hugel B, Ohan J, et al. Shed membrane
microparticles with procoagulant potential in hu-
man atherosclerotic plaques: a role for apoptosis
in plaque thrombogenicity. Circulation 1999;
99(3):348–53.
[65] Davies MJ. The pathophysiology of acute coronary
syndromes. Heart 2000;83(3):361–6.
[66] Kroll MH, Hellums JD, McIntyre LV, et al. Plate-
lets and shear stress. Blood 1996;88(5):1525–41.
[67] Shafer AI, Ali NM, Levine GN. Hemostasis,
thrombosis, fibrinolysis, and cardiovascular dis-
ease. In: Braunwald E, Zipes DP, Libby P, editors.
Heart disease: a textbook of cardiovascular medi-
cine. 6th edition. Philadelphia: W.B. Saunders;
2001. p. 2099–107.
[68] Fitzgerald DJ, Roy L, Catella F, et al. Platelet acti-
vation in unstable coronary disease. N Engl J Med
1986;315(16):983–9.
[69] Fernandez-Ortiz A, Badimon JJ, Falk E, et al.
Characterization of relative thrombogenicity of
atherosclerotic plaque components: implications
for consequences of plaque rupture. J Am Coll Car-
diol 1994;23(7):1562–9.
[70] Guyton JR, Klemp KF. The lipid-rich core re-
gion of human atherosclerotic fibrous plaques:
prevalence of small lipid droplets and vesicles
by electron microscopy. Am J Pathol 1989;134(3):
705–17.
[71] Moreno PR, Bernardi VH, Lopez-Cuellar J, et al.
Macrophages, smooth muscle cells, and tissue
factor in unstable angina. Implications for cell-
mediated thrombogenicity in acute coronary syn-
dromes. Circulation 1996;94(12):3090–7.
[72] Toschi V, Gallo G, Lettino M, et al. Tissue factor
modulates thrombogenicity of human atheroscle-
rotic plaques. Circulation 1997;95(3):594–9.
[73] Nemerson Y. Tissue factor and hemostasis. Blood
1988;71(1):1–8.
[74] Falk E. Unstable angina with fatal outcome: dy-
namic coronary thrombosis leading to infarc-
tion and/or sudden death. Circulation 1985;71(4):
699–708.
[75] Burke AP, Kolodgie FD, Farb A, et al. Healed pla-
que ruptures and sudden coronary death: evidence
that subclinical rupture has a role in plaque pro-
gression. Circulation 2001;103(7):934–40.
34 AYALA & SCHULMAN
[76] Lassila R, Badimon JJ, Vallabhajosula S, et al. Dy-
namic monitoring of platelet deposition on severely
damaged vessel wall in flowing blood: effects of dif-
ferent stenosis on thrombus growth. Arteriosclero-
sis 1990;10(2):306–15.
[77] Badimon L, Badimon JJ, Turitto VT, et al. Plate-
let thrombus formation on collagen type I: a model
of deep vessel injury. Circulation 1988;78(6):
1431–42.
[78] Meyer BJ, Badimon JJ, Mahilac A, et al. Inhibition
of growth of thrombus on fresh mural thrombus.
Targeting optimal therapy. Circulation 1994;
90(5):2432–8.
[79] Rauch U, Osende JI, Chesboro JH, et al. Statins
and cardiovascular diseases: the multiple effects of
lipid-lowering therapy by statins. Atherosclerosis
2000;153(1):181–9.
[80] Narkiewicz K, van de Borne PJ, Hausber M, et al.
Cigarette smoking increases sympathetic outflow in
humans. Circulation 1998;98(6):528–34.
[81] Osende JI, Badimon JJ, Fuster V, et al. Thrombo-
genicity in type 2 diabetes mellitus patients is asso-
ciated with glycemic control. J Am Coll Cardiol
2001;38(5):1307–12.
[82] Rao AK, Chouhan V, Chen X, et al. Activation of
the tissue factor pathway of blood coagulation dur-
ing prolonged hyperglycemia in young healthy
men. Diabetes 1999;48(5):1156–61.
[83] Schulman SP, Goldschmidt-Clermont PJ, Topol
EJ, et al. Effects of integrelin, a platelet glycopro-
tein IIb/IIIa receptor antagonist, in unstable
angina. A randomized multicenter trial. Circula-
tion 1996;94(9):2083–9.
[84] The platelet receptor inhibition in ischemic syn-
drome management in patients limited by unstable
signs and symptoms (PRISM-PLUS) study investi-
gators. Inhibition of the platelet glycoprotein IIb/
IIIa receptor with tirofiban in unstable angina
and non-Q wave myocardial infarction. N Engl J
Med 1998;338(21):1488–97.
[85] Cannon CP, Weintraub WS, Demopoulos LA,
et al. Comparison of early invasive and conserva-
tive strategies in patients with unstable coronary
syndromes treated with the glycoprotein IIb/IIIa
inhibitor tirofiban. N Engl J Med 2001;344(25):
1879–87.
[86] Antman EM, Cohen M, Bernink PJ, et al. The
TIMI risk score for unstable angina/non-ST el-
evation MI: a method for prognostication and
therapeutic decision making. JAMA 2000;
284(7):835–42.
[87] Gluckman TJ, Sachdev M, Schulman SP, et al. A
simplified approach to the management of non-
ST elevation acute coronary syndromes. JAMA
2005;293(3):349–57.
[88] The Danish Study Group on Verapamil in Myocar-
dial Infarction. Effect of verapamil on mortality
and major events after acute infarction. Am J Car-
diol 1990;66(10):779–85.
[89] Cannon CP, Braunwald E. Unstable angina. In:
Braunwald E, Zipes DP, Libby P, editors. Heart
disease: a textbook of cardiovascular medicine.
6th edition. Philadelphia: W.B. Saunders; 2001. p.
1241–9.
[90] Fourth International Study of Infarct Survival
Collaborative Group. ISIS-4: a randomised fact-
orial trial assessing early oral captopril, oral mono-
nitrate, and intravenous magnesium sulphate in
58,050 patients with suspected acute myocardial in-
farction. Lancet 1995;345(8951):669–85.
[91] Gruppo Italiano per lo Studio della Sopravvivenza
nell’infarto Miocardico. Effects of lisinopril and
transdermal glyceryl trinitrate singly and together
on 6-week mortality and ventricular function
after acute myocardial infarction. Lancet 1994;
343(8906):1115–22.
[92] Meine TJ, Roe MT, Chen AY, et al. Association of
intravenous morphine use and outcomes in acute
coronary syndromes: results from the CRUSADE
quality improvement initiative. Am Heart J 2005;
149:1043–9.
[93] Cannon CP, Braunwald E, McCabe CH, et al. In-
tensive versus moderate lipid lowering with statins
after acute coronary syndrome. N Engl J Med
2004;350(15):1495–504.
[94] Lewis HD Jr, Davis JW, Archibald DG, et al.
Protective effects of aspirin against acute myocar-
dial infarction and death in patients with unstable
angina: results of a Veterans Administration Co-
operative Study. N Engl J Med 1983;309(7):
396–403.
[95] The CURE investigators. Effects of clopidogrel in
addition to aspirin in patients with acute coronary
syndromes without ST-segment elevation. N Engl J
Med 2001;345(7):494–502.
[96] Mehta SR, Yusuf S, Peters RJ, et al. Effects of pre-
treatment with clopidogrel and aspirin followed by
long-term therapy in patients undergoing percuta-
neous coronary intervention: the PCI-CURE
study. Lancet 2001;358(9281):527–33.
[97] Hongo RH, Ley J, Dick SE, et al. The effect of clo-
pidogrel in combination with aspirin when given
before coronary artery bypass grafting. J Am Coll
Cardiol 2002;29(2):2271–5.
[98] Boersma E, Harrington RA, Moliterno DJ, et al.
Platelet glycoprotein IIb/IIIa inhibitors in acute
coronary syndromes: a meta-analysis of all major
randomized clinical trials. Lancet 2002;359(9302):
189–98.
[99] Global use of strategies to open occluded in acute
coronary syndromes (GUSTO IV-ACS) Investi-
gators. Effect of glycoprotein IIb/IIIa receptor
blocker abciximab on outcome in patients with
acute coronary syndromes without early coronary
revascularisation: the GUSTO IV-ACS rando-
mised trial. Lancet 2001;357(9272):1915–24.
[100] Brown DL, Fann CS, Chang CJ. Meta-analysis of
effectiveness and safety of abciximab versus
 PATHOGENESIS AND EARLY MANAGEMENT OF NSTE ACS
eptifibatide or tirofiban in percutaneous coronary
intervention. Am J Cardiol 2001;87(5):537–41.
[101] Hirsh J. Heparin. N Engl J Med 1991;324(22):
1565–74.
[102] Oler A, Whooley MA, Oler J, et al. Adding heparin
to aspirin reduces the incidence of myocardial in-
farction and death in patients with unstable angina.
A meta-analysis. JAMA 1996;276(10):811–5.
[103] Cohen M, Demers C, Gurfinkel EP, et alfor the Ef-
ficacy and Safety of Subcutaneous Enoxaparin in
Non-Q Wave Coronary Events (ESSENCE) study
group. A comparison of low molecular weight hep-
arin with unfractionated heparin for unstable coro-
nary artery disease. N Engl J Med 1997;337(7):
447–52.
[104] Petersen JL, Mahaffey KW, Hasselblad V, et al. Ef-
ficacy and bleeding complications among patients
35
randomized to enoxaparin or unfractionated hepa-
rin for antithrombin therapy in non-ST-segment el-
evation acute coronary syndromes: a systematic
overview. JAMA 2004;292(1):89–96.
[105] Organisation to Assess Strategies for Ischemic Syn-
dromes. (OASIS-2) Investigators. Effects of
recombinant hirudin (lepirudin) compared with
heparin on death, myocardial infarction, refractory
angina, and revascularisation procedures in
patients with acute myocardial ischaemia without
ST elevation: a randomised trial. Lancet 1999;
353(9151):429–38.
[106] Direct Thrombin Inhibitor Trialists’ Collaborative
Group. Direct thrombin inhibitors in acute coro-
nary syndromes: principal results of a meta-analy-
sis based on individual patients’ data. Lancet
2002;359(9303):294–302.
 Cardiol Clin 24 (2006) 37–51

Early Management of ST-segment Elevation

Myocardial Infarction
Amish C. Sura, MD, Mark D. Kelemen, MD, MSc, FACC*
Division of Cardiology, University of Maryland School of Medicine, 22 South Greene Street,
Baltimore, MD 21202, USA

The primary goal in treating ST-segment ele- Pharmacologic reperfusion

vation myocardial infarction (STEMI), a condition
Thrombolytic therapy with streptokinase was
caused by acute coronary occlusion, is rapid, early,
first attempted in 1958 [5]. Direct intracoronary
complete, and sustained myocardial reperfusion.
administration was used briefly in the mid 1970s
Animal models of coronary occlusion have dem-
and early 1980s but was abandoned when intrave-
onstrated myocardial necrosis after 30 minutes,
nous administration was found to have similar ef-
with 50% myocardial salvage if reperfusion is
ficacy [6–10]. Thrombolytic agents have a high
accomplished within 90 minutes [1]. Reperfusion
specificity for the substrate plasminogen, hydro-
in STEMI limits myocardial damage and reduces
lyzing a peptide bond to yield the active enzyme
mortality by about 25% [2]. Various pharmaco-
plasmin. Free plasmin is rapidly neutralized by
logic and mechanical reperfusion options are
the serine proteinase inhibitor alpha-antiplasmin,
available that differ in efficacy based on time to
whereas fibrin-bound plasmin is protected from
treatment from symptom onset and hemodynamic
rapid inhibition, thereby promoting clot lysis.
status at presentation. Surgical approaches have
Given its relative ease of administration and
limited utility as acute therapies for STEMI [3].
proven mortality benefit, intravenous thrombolytic
More important than the mode of reperfusion is
therapy is the most common form of reperfusion
the rapid initiation of a reperfusion strategy once
therapy worldwide [11]. The benefits of reperfu-
a patient is identified as having a STEMI. The
sion therapy are time related (Fig. 1), with
American College of Cardiology (ACC)/American
decreasing benefits with increasing delays to ther-
Heart Association (AHA) guidelines suggest
apy [2]. The beneficial effect of fibrinolytic ther-
a maximum time from initial medical contact to
apy is substantially higher in patients presenting
initiation of thrombolytic therapy of 30 minutes
within 2 hours after symptom onset than in
and to balloon inflation of no more than 90
those presenting later [12]. Administration of
minutes if percutaneous coronary intervention
thrombolytics primarily occurs upon patient pre-
(PCI) is chosen [3]. Several algorithms for identifi-
sentation to the hospital, but prehospital throm-
cation of patients who have STEMI and choice of
bolysis reduces time to treatment by up to 1
reperfusion therapy have been developed, each
hour and reduces mortality by 17% [13].
taking into account time to presentation, contrain-
A 1994 meta-analysis from the Fibrinolytic
dications to therapy, and the hemodynamic status
Therapy Trialists’ Collaborative Group found
of the patient [4]. A realistic assessment of time re-
that the absolute mortality benefit at 5 weeks
quired to initiate therapy must also factor in the
was 3% for those presenting within 6 hours from
choice of reperfusion strategy.
symptom onset, 2% for those presenting within 7
to 12 hours, and a nonsignificant benefit of 1% for
those presenting within 13 to 18 hours [2]. The net
* Corresponding author.
effect in major thrombolytic trials has been an
E-mail address: mkelemen@medicine.umaryland.edu
approximately 30% relative risk reduction in
(M.D. Kelemen).

0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.011 cardiology.theclinics.com
38
12
1-YEAR MORTALITY (%)
10
8
6 these limitations and enhance efficacy. The ideal
4
2
0 a minimal effect on normal hemostasis [3,26].
0 60 120 180 240 300
ISCHEMIC TIME (minute)
Fig. 1. There is a continuous relationship between time-
to-treatment and mortality at 1 year, with increasing de-
lays resulting in increased mortality. The relationship
between time-to-treatment and 1-year mortality, as a
continuous function, was assessed with a quadratic
regression model. Dotted lines represent 95% CIs of pre-
dicted mortality. (Modified from Rogers WJCJ, Lambrew
CT, et al. Temporal trends in the treatment of over 1.5
million patients who had myocardial infarction in the
US from 1990 through 1999: the National Registry of
Myocardial Infarction 1, 2 and 3. J Am Coll Cardiol
2000;36:2056–63; with permission.)
short-term mortality to an absolute value of 7%
to 10%; the clinical advantage persists for at least
10 years [14,15]. A similar improvement in sur-
vival has been noted in community-based studies
[16]. Several studies have demonstrated that the
morbidity and mortality effects of thrombolytics
are realized only if therapy is given within 12
hours of symptom onset [3,17,18]. Most patients,
however, still do not receive thrombolytics within
the recommended 30 minutes from presentation
to the medical system [19].
Although time from symptom onset to initiation
of therapy is a critical determinant of survival,
many patients who have STEMI and are eligible for
reperfusion receive delayed treatment or no treat-
ment at all. Results from the multinational Global
Registry of Acute Coronary Events (GRACE)
show that approximately one third of patients
who had STEMI and who had no known contra-
indications did not receive reperfusion therapy, and
this finding was independent of geographic region
[20–22]. Fear of bleeding complications from fibri-
nolytic drugs is the primary reason for underuse of
these drugs [22].
Thrombolytic agents do have several limita-
tions. There is failure to achieve patency in 15%
to 20% of patients, failure to achieve normal
(Thrombolysis in Myocardial Infarction trial
[TIMI] grade III) flow in 40 to 50% of patients,
SURA & KELEMEN
and a 10% to 15% rate of reocclusion [23]. There
is also a 0.5% to 1% risk of intracranial hemor-
rhage [24,25]. As a result, there is an ongoing
effort to develop newer agents that may overcome
thrombolytic agent should be fibrin specific and
should be directed at newly formed fibrin (present
within intravascular thrombus), while exerting
360 The onset of fibrinolytic activity should be rapid,
and the circulating half-life should be long enough
to permit abbreviated dosing and wide-scale ac-
ceptance in clinical practice. It should not be anti-
genic, and it should be available at a cost that
fosters its routine use. Although pharmacologic
characteristics of newer thrombolytic agents
have improved, large clinical trials have demon-
strated only modest improvements in 30-day mor-
tality [27]. The drugs currently approved for use in
the United States are outlined in Table 1. There
are several dosing regimens available for these
drugs, consisting of single- or double-bolus ad-
ministration or an accelerated front-loaded regi-
men consisting of a single bolus followed by
continuous intravenous administration over 60
to 90 minutes. Double-bolus alteplase (tPA) was
found to have an increased incidence of bleeding
complications, stroke, and death as compared
with front-loaded tPA [28]. Accelerated tPA has
an approximately 1% survival advantage over
streptokinase [23,29]. The benefit from tPA is
greatest in patients less than 75 years old and in
those who have anterior wall infarctions [23],
but a consistent benefit is seen in virtually all sub-
groups, including patients who are older, have
a nonanterior infarction or hypertension, or
have had a prior myocardial infarction (MI) or
coronary artery bypass grafting [29–31]. The ben-
efit of tPA over streptokinase is thought to be
caused by a significant increase in TIMI III (nor-
mal) blood flow in the infarct-related artery with
tPA administration [32]. Compared with tPA, re-
teplase (rPA) is less fibrin selective and has a lon-
ger half-life. Several studies suggest similar clinical
benefit with either tPA or rPA. The largest, trial,
Global Use of Strategies to Open Occluded Coro-
nary Arteries (GUSTO) III, compared 15,000 pa-
tients who had STEMI and found no significant
difference between accelerated alteplase and dou-
ble-bolus reteplase in 30-day and 1-year mortality
or stroke incidence [33]. As with tPA, rPA has
demonstrated a lower incidence of cardiogenic
shock compared with streptokinase [34]. Tenecte-
plase (TNK-tPA) is a genetically modified variant
 EARLY MANAGEMENT OF STEMI
Table 1
Characteristics of currently approved thrombolytics in the United States
Alteplase
Streptokinase (t-PA, activase)
Dosing 1.5 million 15 mg bolus,
30-60 min 0.75 mg/kg
(max 50 mg) Over
30 min, then 0.5 mg/kg
(max 35 mg) Over 1 hour
Systemic fibrinolytic Marked Mild
state (fibrinogen
depletion)
Antigenic Yes No
Bolus administration No No
Approximate cost 300 2200
(US dollars, 2001)
Intracranial hemorrhage (%) 0.5 0.8
of tPA which has a significantly longer half-life,
more fibrin specificity, and less inhibition by plas-
minogen activation inhibitor than standard tPA.
Single-bolus TNK-tPA is associated with less
bleeding than tPA and with similar 30-day mortal-
ity [35,36]. Although there is no proven mortality
benefit of TNK-tPA over tPA in patients present-
ing within 4 hours of symptom onset, these data
favor TNK-tPA in patients who present later
than 4 hours after the onset of symptoms [36]. Sig-
nificantly, although the rates of intracranial hem-
orrhage are similar for the two drugs, TNK-tPA
use is associated with fewer bleeding complica-
tions unrelated to intracranial hemorrhage and
less need for blood product transfusion [35]. It is
also easier and faster to use, because its longer
half-life permits single-bolus administration. Al-
though streptokinase is substantially less expen-
sive than any of the newer thrombolytic agents,
its potential antigenicity, short half-life, and lack
of fibrin specificity led to a reduction in its use
in the United States.
Indications and contraindications for
pharmacologic reperfusion
The ACC/AHA class I indications for use of
thrombolytic therapy in STEMI are patients who
had symptom onset less than 12 hours previously
and who have ST-segment elevation in at least
two contiguous leads or new/presumably new left
bundle branch block [3]. Thrombolytic therapy is
most effective when applied to younger (!75
years) patients who do not have heart failure or
39
Reteplase Tenecteplase
(r-PA, retevase) (TNK-tPA)
10 U bolus
2, Weight based: 30 mg
30 minutes apart (!60 kg), 35 mg
(60-69 kg), 40 mg
(70-79 kg), 45 mg
(80-89 kg), 50 mg
(R90 mg)
Moderate Minimal to mild
No No
Yes (double bolus) Yes
2200 2200
0.9 0.9
hemodynamic instability presenting early after
symptom onset and in whom the risk of major
bleeding is low [4]. Elderly patients (O75 years)
who do not have contraindications for thrombo-
lytic therapy also derive significant morbidity
and mortality benefits, although the absolute mor-
tality reduction is less than that for younger pa-
tients (Fig. 2) [37,38].
Indicators of successful reperfusion include
relief of symptoms within 60 to 90 minutes of
receiving thrombolytics or a ST-segment decrease
of at least 50% [3,4]. Echocardiographic contrast
perfusion is another marker of reperfusion. An-
giographic evaluation of epicardial antegrade
flow of the infarct related artery, as measured by
the TIMI flow grade classification, is an important
predictor of outcome. TIMI grade 0 refers to the
absence of any antegrade flow beyond a coronary
occlusion, whereas TIMI 1 flow is faint antegrade
coronary flow beyond the occlusion with incom-
plete filling of the distal coronary bed. TIMI 2
flow is delayed or sluggish antegrade flow with
complete filling of the distal territory, and TIMI
3 flow is defined as normal flow that fills the distal
coronary bed completely [39]. Only restoration of
normal epicardial flow (TIMI 3 flow) has been as-
sociated with improved left ventricular function
and survival; the outcome of patients who have
TIMI 2 flow is similar to those who have only
TIMI 0 or 1 flow [29,40,41]. Other angiographic
features of epicardial flow and myocardial perfu-
sion also correlate with outcome after thrombolytic
administration [42,43]. Patency of the infarct-re-
lated artery does not always result in cellular
40 SURA & KELEMEN
Fig. 2. Although there is increased risk, the probability of
death or intracranial hemorrhage is not increased in pa-
tients older than 75 years who receive thrombolytic ther-
apy. The figure shows the adjusted probability of death or
cerebral bleeding in relation to fibrinolytic therapy in pa-
tients who had STEMI and who were 75 years old or older
(dotted line) versus that among patients who had STEMI
who did not receive fibrinolysis (solid line). At 30 days and
1 year, the probability was 23% and 32% versus 26% and
36%, respectively. (Modified from Stenestrand U, Wallen-
tin L. Fibrinolytic therapy in patients 75 years and older
with ST-segment-elevation myocardial infarction:one-
year follow-up of a large prospective cohort. Arch Intern
Med 2003;163:968.)
reperfusion, and a current of injury may persist.
For this reason, early resolution of ST segment
elevation is a better representation of complete
reperfusion than angiographic criteria [17].
Several studies have demonstrated that the
mortality benefits of thrombolytics are gender
independent and are present regardless of blood
pressure (if less than 180 mm Hg systolic), heart
rate, presence of diabetes mellitus, or previous MI
[2]. The greatest absolute mortality benefit is seen
in patients who have anterior STEMI and signifi-
cant myocardium at risk; the benefit is less
in inferior STEMI, except in subgroups with asso-
ciated right ventricular infarction or anterior ST
segment depression [2].
Hemorrhage is the most severe risk of throm-
bolytic therapy, and most contraindications to
thrombolytic therapy are based on the risk of
bleeding after receiving thrombolytics. Absolute
contraindications include any prior intracranial
hemorrhage, recent face or head trauma (within
3 months), known central nervous system vascular
malformation or neoplasm, ischemic stroke
within 3 months (except acute ischemic stroke),
and suspicion of aortic dissection [3]. Clinical pre-
dictors of intracranial bleeding with use of
thrombolytics are age (O65 years), female gender,
systolic blood pressure higher than 180 mm Hg on
presentation, and low (!70 kg) body weight [17].
The risk of bleeding with these drugs is, in part,
related to relative fibrin specificity. The use of fi-
brinolytic drugs carries a category C (uncertain ef-
fect) rating in pregnancy, because there have been
no well-controlled studies. Streptokinase is anti-
genic, and there is a risk of allergic reactions [17].
Mechanical reperfusion
In the TIMI and GUSTO trials, attempts at
reperfusion using thrombolytics failed to restore
full patency of the infarct-related arteries in up to
54% of cases [44,45]. Limitations of thrombolytics
also included a risk of early recurrent ischemia of
10% to 15% and a 5% risk of reinfarction after
thrombolytic therapy. Mechanical reperfusion
with PCI has gradually replaced the used of lytics
in many centers. Primary PCI is defined as mechan-
ical reperfusion without prior thrombolytic therapy.
Rescue PCI is emergent PCI after failed throm-
bolysis. Immediate and delayed (1–7 days) PCI
refers to the timing of PCI after successful
thrombolysis. Facilitated PCI is a newer concept re-
ferring to emergent PCI immediately after reduced-
dose thrombolysis. Compared with thrombolysis,
PCI restores normal epicardial blood flow (TIMI
III) in 70% to 90% of cases [46]. Several studies
have demonstrated that normal flow in the in-
farct-related artery is associated with a mortality
of 3.7%, compared with 6.6% and 9.2% in patients
who have impaired flow and an occluded or nearly
occluded infarct-related artery [47,48]. PCI also al-
lows early identification of patients who may bene-
fit from surgical revascularization as well as the
application of other mechanical interventions,
such as the use of intra-arterial balloon counter-
pulsation therapy, thrombectomy, and distal pro-
tection devices, which may improve outcome.
Initial comparisons of PCI with thrombolytic
therapy were conducted in the pre–coronary stent
era and used percutaneous transluminal coronary
angioplasty (PTCA). In the Primary Angioplasty
in Myocardial Infarction trial of 395 patients who
had STEMI presenting within 12 hours of symp-
tom onset, PTCA therapy reduced the combined
occurrence of in-hospital death or nonfatal re-
infarction at 2-year follow-up, as compared with
tPA therapy [46]. These benefits persisted al-
though overall left ventricular systolic function
did not differ in the two groups. Subgroup analy-
sis revealed that location of the infarct affected

outcome, with the mortality benefit in the PTCA
group seen primarily in patients presenting with
anterior wall MI [49]. Patients in the non–anterior
wall MI group also derived benefit, however, hav-
ing a lower rate of recurrent ischemia (9.7% ver-
sus 27.8%) and fewer urgent catheterizations
and revascularization procedures [49].
A meta-analysis of 10 trials (Fig. 3) comparing
use of streptokinase, tPA, and accelerated tPA
with PTCA found that, at 30 days, primary angio-
plasty seemed to be superior to thrombolytic ther-
apy for treatment of patients who had acute MI
[50]. The rates of death or nonfatal reinfarction
were 7.2% for angioplasty and 11.9% for throm-
bolytic therapy (P ! .001). PTCA was also asso-
ciated with a significant reduction in total stroke
(0.7% versus 2.0%; P ¼ .007) and hemorrhagic
stroke (0.1% versus 1.1%; P ! .001) when
compared with the thrombolytics used in the
trials. A potential limitation of these data is the
limited use of newer, more effective thrombolytic
regimens, stents, and adjunctive therapies. These
issues were addressed in a subsequent meta-analy-
sis of 23 trials comparing thrombolytics with PCI,
8 of which involved primary stenting of the
infarct-related artery. PCI still reduced the com-
bined endpoint of death, nonfatal reinfarction,
recurrent ischemia, and stroke, and this benefit
was maintained at 18 months [51]. The advantages
of PCI in these meta-analyses should be viewed in
light of several limitations of the studies, such as
the relatively low numbers of patients and limited
use of cardiac biomarkers in the PCI arms.
There was initial reluctance to use intracoro-
nary stents in the setting of thrombus and an
active coagulopathic state, but subsequent studies
demonstrated improved outcomes of timely PCI
20
PTCA (n=1466)
Thrombolytic therapy (n=1443)
15
Frequency (%)
p=0.057
10
0
Death
Fig. 3. PTCA is associated with significantly fewer complications than thrombolysis for patients presenting with
STEMI. (From Keeley E, Boura J, Grines C. Primary angioplasty versus intravenous thrombolytic therapy for acute
myocardial infarction: a quantitative review of 23 randomised trials. Lancet 2003;361:17; with permission.)
EARLY MANAGEMENT OF STEMI 41
with stents over thrombolysis [52–54]. The Danish
Acute Myocardial Infarction-2(DANAMI-2) trial
randomly assigned 1572 patients to either front-
loaded tPA or primary PCI with stents and was
discontinued prematurely when PCI with stenting
significantly reduced the primary combined end-
point of stroke, reinfarction, and mortality at 30
days [54]. The majority of this benefit was derived
from a reduction in reinfarction rates. These re-
sults are independent of adjunctive glycoprotein
(Gp) IIb/IIIa use [53]. Increasingly, drug-eluting
stents are being used for primary PCI in STEMI,
having been shown to reduce target-vessel revas-
cularization at 300 days with no increase in stent
thrombosis [55]. Comparison studies of primary
stenting versus primary PTCA have shown that
stenting results in a higher rate of successful reper-
fusion and lower rates of acute vessel closure and
vessel restenosis [56–58]. These short-term bene-
fits, however, have not translated into long-term
reduction in recurrent MI or mortality benefit
with stenting. Reductions in composite endpoints
of death, stroke, reinfarction, and target-vessel
revascularization have all been driven primarily
by reductions in target-vessel revascularization
[59,60]. Although most trials comparing PCI and
thrombolytics have included patients eligible to
receive thrombolytic therapy, studies have also
shown benefits of PCI in thrombolytic-ineligible
patients [61].
Analysis of PCI trials reveals that certain
subgroups of STEMI patients derive particular
benefit from a strategy of primary PCI over
fibrinolysis. Several risk models (TIMI, GRACE,
Gruppo Italiano per lo Studio della Sopravvivenza
nell’Infarto-1 study [GISSI], prehospital risk in-
dex) identify a variety of individual patient
p<0.0001
p<0.0001 p=0.049 p=0.25
Non-fatal Total Haemorrhagic Death, non-fatal
myocardial stroke stroke reinfarction, or
infarction stroke
42 SURA & KELEMEN
features, such as age, hemodynamic status, pres-
ence of comorbid conditions, infarct location, and
time from onset of symptoms, to predict early
mortality in patients presenting with STEMI [62–
65]. These models allow rapid and early risk strat-
ification and maintain predictive accuracy regard-
less of the reperfusion strategy chosen. In general,
patients who have higher risk scores benefit from
PCI instead of thrombolysis. Elderly patients
(O75 years old) presenting with STEMI more often
have multivessel coronary disease and significantly
increased mortality compared with younger pa-
tients. Data from thrombolytic trials shows that
an age older than 65 years is an independent risk
factor for mortality and increased incidence of
stroke and intracranial hemorrhage with thrombo-
lytic therapy [66]. Observational studies and trials
comparing PCI and thrombolytics demonstrate
a significant mortality benefit in elderly patients
treated with PCI suggesting that primary PCI
should be the preferred choice for revascularization
in this patient group [37,67,68]. Another extremely
high-risk group of patients who have STEMI are
those presenting in cardiogenic shock, with mor-
tality greater than 80% if left untreated. Both the
GISSI-1 and the International Study Group trials
demonstrated extremely high mortality rates in pa-
tients who had cardiogenic shock treated with
thrombolytics; in some cases, results were no bet-
ter than with placebo [15,69]. Results of the Should
We Emergently Revascularize Occluded Coronar-
ies for Cardiogenic Shock (SHOCK) trial, in which
302 patients were randomly assigned to emergent
revascularization (64% PCI) or initial medical sta-
bilization showed no difference in mortality at 30
days but a significant mortality reduction at 6
months in the revascularization arm [70]. Subse-
quent follow-up has also shown than emergent re-
vascularization for patients who have cardiogenic
shock produces improved functional status over
time, with good quality of life and fewer symptoms
of heart failure as compared with medical ther-
apy. [71] The ACC/AHA has given a class I rec-
ommendation to a strategy of early reperfusion
with intra-arterial balloon counter-pulsation sup-
port for all patients younger than 75 years who
develop shock within 36 hours of MI and who
can be treated within 18 hours from the onset
of shock. Early revascularization for selected pa-
tients older than 75 years who have shock was
given a class IIA recommendation [3]. There are
also data to support use of PCI as the preferred
reperfusion strategy in patients presenting with
anterior MI or heart failure and in patients
who have had prior coronary artery bypass graft-
ing [70,72–74].
PCI involves risks and technical concerns that
must be considered in determining the optimal
reperfusion strategy for a patient. Mechanical
complications associated with PCI include vessel
dissection or rupture and are more frequent in
elderly patients. The potential adverse effects of
contrast medium must be considered. Resource
availability, time of day, and expertise of medical
personnel must also be taken into account and are
a much more prominent concern than with use of
thrombolytics. Data from the National Registry of
Myocardial Infarction (NRMI) demonstrated that
there are significantly longer door-to-balloon times
when STEMI patients present outside the hours of
8 AM and 6 PM [75]. The NRMI study also showed
no difference in outcome between thrombolytics
and PCI for patients who had STEMI in low-vol-
ume centers (%16 procedures/year), with lower
volume centers having higher mortality rates. In-
termediate-volume centers (17–48 procedures/
year) and high-volume centers (R49 procedures/
year) had lower mortality rates with reperfusion
by PCI [75]. A second NRMI analysis showed no
significant relationship between hospital volume
of thrombolytic interventions and mortality [76].
As with thrombolytic therapy, the time to
reperfusion is critical. There is a positive linear
relationship between duration of ischemic time
and poor outcome. There are discordant data
regarding time of symptom onset to PCI, how-
ever. Several studies claim no influence of time delay
when PCI is performed within 2 to 3 hours of
symptom onset [77,78]. These findings may result,
at least in part, from discrepancies between time
of symptom onset and presentation to hospital
and exclusion of patients who died before reach-
ing the hospital. Other studies, after adjusting
for baseline characteristics, show a significant cor-
relation between time from symptom onset to bal-
loon inflation and 1-year mortality [79–81]. Data
from a Dutch study showed a continuous relation-
ship between symptom duration and mortality,
with a relative risk of death of 1.08 for every
30-minute delay in reperfusion [80].
Adjunctive therapy in ST-segment elevation
myocardial infarction
Adjunctive medical therapy is a critical com-
ponent of therapy for STEMI and confers benefit
in addition to that gained by reperfusion therapies,
regardless of method of reperfusion (Table 2).
 EARLY MANAGEMENT OF STEMI 43

Ancillary therapy can be used to facilitate and en-
hance coronary reperfusion or to limit the conse-
quences of myocardial ischemia. Early, aggressive
use of antiplatelet therapies, such as aspirin, and
clopidogrel, confers significant additional mortal-
ity benefit when given as adjuncts to thrombolysis
or PCI [82].
Beta-blockers, angiotensin-converting enzyme
inhibitors in appropriately selected patients, and
3-hydroxy-3-methylglutaryl CoA reductase inhib-
itors (statins) have all been shown to reduce
the risk of cardiovascular events and mortality
in patients who have STEMI. Therapies such
as nitroglycerin and morphine have no mortality
benefit but may improve symptoms and re-
duce ischemic burden. Calcium-channel blockers
and prophylactic anti-arrhythmic drug therapy
(lidocaine) may increase mortality. The suggested
benefit of metabolic modulation at the myocyte
level with electrolytes, glucose, and insulin seen in
small, early trials has not been reproduced in
larger, randomized studies [83].
Thrombin inhibition enhances coronary throm-
bolysis and limits reocclusion. As a result, adjunc-
tive anticoagulation is administered to most
patients receiving thrombolytic therapy or PCI.
Although it is extensively used, the benefit of
heparin in the current era of antiplatelet therapy,
advanced thrombolytics, and PCI is controversial.
There are currently no data to support the routine
administration of adjunctive unfractionated heparin
in patients receiving earlier generation, non–fibrin-
specific thrombolytic agents, providing aspirin
is given [44,73]. Although streptokinase in

Table 2
Mechanisms and clinical effects of adjunctive therapies for patients who have ST elevation myocardial infarction
Agent Mechanism Clinical Effect
Adjunctive therapies for AMI
Aspirin Antiplatelet Improve survival
Decrease reinfarction, CVA
Thienopyridines Antiplatelet Recommended in aspirin-allergic patients
(clopidogrel, ticlopidine) Decrease death, MI, CVA in NSTE ACS
Glycoprotein IIb/IIIa inhibitors Antiplatelet Decrease MI, ischemic complications
following primary PCI
Decrease death or MI in high-risk NSTE ACS
Unfractionated heparin Antithrombin Decrease death and MI in prefibrinolytic era
Low molecular weight heparins Antithrombin Reduce cardiac events in NSTE ACS versus
unfractionated heparin
Direct thrombin inhibitors Antithrombin Recommended in heparin-induced
thrombocytopenia
Beta-brockers Decrease myocardial oxygen Improve survival
demand (YHR, YBP) Reduce infract size, ventricular arrhythmias,
recurrent ischemia
Adjunctive therapies for AMI
ACE Inhibitors Vasodilator (BP) Improve survival
Prevent LV remodeling Decrease heart failure. LV dysfunction
IV nitroglycerine Venous, arterial, coronary No effect on survival
vasodilator (YBP, Ypreload) Decrease recurrent ischemia
HMG CoA Reductase Inhibitors Lipid lowering Decrease future CV death and MI
Anti-inflammatory May decrease early ischemic events
Magnesium Myocardial protective Therapy for torsades depointes
Anti-arrhythmic May improve reperfusion outcomes
Calcium-channel blocker Decrease myocardial oxygen No survival benefit
demand (YHR, YBP) Possible use in beta-blocker–intolerant patients
without CHF or LV dysfunction
Warfarin Oral anticoagulant Reduced embolic risk with atrial fibrilliation.
LV thrombus or dysfunction
Glucose-insulin potassium Metabolic modulator May improve intracellular myocardial energy
infusion stores and outcome
Abbreviations: ACS, acute coronary syndrome; BP, blood pressure; CHF, congestive heart failure; CV, cardiovascu-
lar; CVA, cerebrovascular accident; HR, heart rate; IV, intravenous; LV, left ventricular; MI, myocardial infarction;
NSTE, non–ST-segment elevation; PCI, percutaneous coronary intervention.
44 SURA & KELEMEN
combination with enoxaparin has been shown to
have better ST-segment resolution and better an-
giographic patency at days 5 through 10, the possi-
ble benefit of enoxaparin on survival and long-term
mortality with first-generation agents remains un-
certain [84]. The 2004 ACC/AHA guidelines, how-
ever, did recommend unfractionated heparin
therapy followed by long-term oral anticoagulation
in patients who are at high risk for systemic throm-
boembolism (large or anterior MI, atrial fibrilla-
tion, previous embolus, or known left ventricular
thrombus) [3]. Unlike older-generation agents,
both unfractionated heparin and low molecular
weight heparin have shown additional benefit
when used with the newer agents tenecteplase, alte-
plase, and reteplase. Careful, monitored adminis-
tration of intravenous unfractionated heparin in
patients undergoing PCI has been shown to prevent
acute vessel closure caused by thrombosis. Studies
with Gp IIb/IIIa inhibitors as adjuvant therapy to
thrombolysis have shown increased rates of moder-
ate to severe bleeding without demonstrating sig-
nificant reductions in 30-day and 1-year mortality
rates, leading the Consensus Guideline Panel to
give combination therapy a class IIB recommen-
dation [3,85]. In contrast, Gp IIb/IIIa inhibitors
have demonstrated early and sustained mortality
benefit in patients undergoing PCI for reperfusion,
with increasing benefit in higher-risk patients. Al-
though thrombolytic therapy may be underused,
realization of the benefits of adjunctive therapies
has resulted in dramatic increases in their use over
time.
Time to balloon inflation and mortality in primary
percutaneous coronary intervention for
ST-segment elevation myocardial infarction
In many of the studies comparing PCI and
thrombolysis, PCI was performed within 3 hours
of symptom onset. Registry data confirm that
there is a significant difference in time from
presentation to PCI depending on the need for
transfer to a PCI-capable facility [19]. Several
studies confirm the benefits of PCI even when
there is some delay in therapy resulting from inter-
hospital transfer and preparation for PCI [86–88].
As the time delay for PCI increases, however, the
mortality benefit of PCI over thrombolytic ther-
apy decreases, and thrombolysis may have a greater
mortality benefit when there is a 60-minute or
longer delay in performing PCI. In this meta-
analysis, every 10-minute delay reduced the
absolute mortality benefit by approximately 1%,
and the benefit in combined endpoint of stroke,
reinfarction, and death was decreased by almost
2% [89].
The available data suggest that primary PCI is
preferable to thrombolysis if PCI is performed
within 90 minutes at the presenting institution by
experienced interventionalists in a cardiac cathe-
terization laboratory that performs more than 36
such procedures per year, even in community
hospitals without surgical back-up [90]. Primary
PCI is also preferable if transfer to a neighboring
institution can be accomplished within 30 to 60
minutes. Compared with thrombolysis, primary
PCI achieves a higher rate of TIMI 3 flow, does
not carry the risk of intracranial hemorrhage,
and is associated with improved outcomes. Pri-
mary PCI is not available at all institutions, however,
there are often delays in implementation, and lo-
cal expertise is an important determinant of out-
come. As a result, the ACC/AHA task force gave
a class I recommendation to the use of thrombolytic
therapy for any patient who does not have contra-
indications and who presents to a facility without
the capability for expert, prompt intervention
with primary PCI within 90 minutes of first medical
contact [3]. Thrombolytic therapy was also recom-
mended if the relative delay necessary to perform
primary PCI (the expected door-to-balloon time
minus the expected door-to-needle time) is greater
than 1 hour.
Combined reperfusion strategies
Although reperfusion with thrombolytics and
PCI has demonstrated significant mortality and
combined endpoint benefits in patients presenting
with STEMI, there are limited data regarding
combination therapy. The use of early, in some
cases prehospital, thrombolytic therapy followed
by early, planned PCI is known as facilitated PCI.
This approach must be differentiated from PCI
after thrombolytic failure, known as rescue PCI,
and from urgent PCI for threatened reocclusion
or hemodynamic instability. Initial data from the
1980s comparing PTCA immediately after admin-
istration of thrombolytics suggested no additional
benefit and increased morbidity, primarily in the
form of bleeding [91,92]. Advances in pharmaco-
logic regimens, in PCI, and in adjunctive thera-
pies, however, have led to re-examination of
facilitated PCI. The Plasminogen-activator An-
gioplasty Compatibility trial examined the benefit
 EARLY MANAGEMENT OF STEMI 45

Evaluate:
• Time since onset of symptoms
• Ml risk (patient and ECG)
• Risk of fibrinolysis
• Time to fibrinolysis or PCl

Reperfusion indicated

YES Contraindication to fibrinolysis? NO

Does the patient have

Perform PCI Killip class 3/4 or other
promptly high-risk AMI features?
if feasible

Is PCI reliably available

within 60 minutes of YES NO
time to fibrinolysis?

Is PCI reliably available

within 60 minutes of
time to fibrinolysis?
YES NO

YES NO
Transfer to PCI Give fibrinolysis and
centre and/or appropriate cardiopulmonary
perform PCI support and transfer to
tertiary cardiac centre
PCI or Give
fibrinolysis fibrinolysis

Fig. 4. A proposed model for the selection of reperfusion therapy for patients who have STEMI presenting within 12
hours of symptom onset. The figure shows the Canadian Cardiovascular Society Working Group algorithm for the se-
lection of patients for reperfusion after STEMI. The algorithm applies to patients presenting within 12 hours of symp-
tom onset. It assumes that the diagnosis of STEMI is not in doubt and indicates that, for most hospitals caring for
patients who have STEMI who are not at high risk, fibrinolysis is the preferred option. AMI, acute myocardial infarc-
tion; PCI, percutaneous cardiac intervention. (From Canadian Cardiovascular Society Working Group. Applying the
new STEMI guidelines: 1. Reperfusion in acute ST-segment elevation myocardial infarction. CMAJ 2004;171:10–41;
with permission.)

of alteplase immediately before PCI in 606 pa- Future directions

tients who had STEMI. This study suggested
Although improvements in the management of
that thrombolysis followed by immediate PCI
patients who have STEMI have led to a decline in
led to more frequent early recanalization of the
acute and long-term fatality rates, reperfusion and
infarct-related artery, preservation of left ventric-
ancillary therapies remain underused. Several
ular function, and no increase in adverse events
initiatives (NMRI, Get With the Guidelines) are
[93]. A synergistic approach offers the timeliness
designed to improve adherence to guidelines and
of thrombolytic therapy followed by early PCI
access to appropriate reperfusion therapies [22].
to consolidate the reperfusion process and prevent
To date, clinical advancement is judged on achiev-
reocclusion [94]. There are many issues to be re-
ing and maintaining epicardial artery patency.
solved, such as timing of therapy (early versus de-
New fibrinolytics and combination therapies will
layed PCI after thrombolysis, timing of hybrid
continue to evolve. There will be technological ad-
therapy with duration of symptoms), type of
vances and improvement in operator skills for
thrombolytic agent, need for adjuvant medical
PCI. There is increasing attention to improving
therapy, and appropriate patient selection. Sev-
outcome at the myocyte level, however, because
eral ongoing trials may provide insight into the
an open artery does not equate to adequate tissue
role of combination therapy [95].
46 SURA & KELEMEN
perfusion. Edema, inflammation, necrosis, micro-
emboli, and microvascular constriction contribute
to reperfusion injury and tissue death. Further ad-
vances for managing STEMI must focus on thera-
pies that mitigate these factors. Initial studies have
suggested benefit to anti-inflammatory and anti-
complement therapies and suggest an expanding
role for ancillary therapy [96]. The ultimate goal
for the management of STEMI remains un-
changed: to open occluded arteries quickly in
carefully screened patients and in a cost-effective
manner.
Current approach to initial treatment of
ST-segment elevation myocardial infarction
Fig. 4 shows one suggested approach for the
initial treatment of STEMI, taking into account
symptom duration, hemodynamic status, local ex-
pertise, and transport systems to help make deci-
sions regarding the choice of reperfusion therapy
for patients presenting within 12 hours of symp-
tom onset [4]. A directed history and physical ex-
amination, focusing on the time from onset of
symptoms to presentation and hemodynamic sta-
tus, should be completed quickly. If the history
correlates with a 12-lead ECG demonstrating
new or presumed new ST-segment elevation in
two or more contiguous leads, or left bundle
branch block the diagnosis of STEMI can be
made, and focus can be directed to the appropri-
ate mode of reperfusion. If there are no contrain-
dications for thrombolysis, it is a reasonable
option, especially if local PCI expertise is not on
site or if transfer to a PCI-capable facility will
add more than 60 minutes to the time to fibrinoly-
sis. For patients in whom thrombolytics are con-
traindicated or who have high-risk STEMI
features, such as cardiogenic shock, PCI is the
preferred method of revascularization. In the
worst-case scenario, a patient who has STEMI
and high-risk features, contraindications to fibri-
nolysis, and for whom PCI is not readily available
has an extremely high risk of dying. The approach
to such patients is not well delineated, but it may
be reasonable to try to obtain some level of reper-
fusion with thrombolytics and supportive meas-
ures before transfer to a tertiary level cardiac
center. Although both thrombolytic therapy and
PCI focus on epicardial artery patency, strategies
to improve myocardial-level perfusion must also
be employed. It is possible that near-maximal
benefit has been obtained by modification of
100
Mortality Reduction, %
80 D Shifts in Potential Outcomes
With Different Treatment Strategies
A to B No Benefit
60
C
A to C
B to C
Benefit
Benefit
40 D to B Harm
D to C Harm
20 B
Extent of A
Myocardial Salvage
0
0 4 8 12 16 20 24
Time From Symptom Onset to
Reperfusion Therapy, h
Critical Time-Dependent Period Time-Independent Period
Goal: Myocardial Salvage Goal: Open Infarct-Related Artery
Fig. 5. A hypothetical model demonstrating the rela-
tionship of symptom duration, myocardial salvage, and
mortality for patients presenting with STEMI. Mortality
reduction as a benefit of reperfusion therapy is greatest
in the first 2 to 3 hours after the onset of symptoms of
acute myocardial infarction, most likely a consequence
of myocardial salvage. The exact duration of this critical
early period may be modified by several factors, includ-
ing presence of functioning collateral coronary arteries,
ischemic preconditioning, myocardial oxygen demands,
and the duration of sustained ischemia. After this early
period, the magnitude of the mortality benefit is much
reduced, and as the mortality reduction curve flattens,
time-to-reperfusion therapy is less critical. If a treatment
strategy, such as facilitated PCI, is able to move patients
back up the curve, a benefit would be expected. The
magnitude of the benefit would depend on how far up
the curve the patient can be shifted. The benefits of a shift
from points A or B to point C would be substantial, but
the benefit of a shift from point A to pint B would be
small. A treatment strategy that delays treatment during
the critical early period, such as patient transfer for PCI,
would be harmful (shift from point D to point C or
point B). Between 6 and 12 hours after the onset of
symptoms, opening the infarct-related artery is the pri-
mary goal of reperfusion therapy, and primary PCI is
preferred over fibrinolytic therapy. The possible contri-
bution to mortality reduction of opening the infarct-re-
lated artery, independent of myocardial salvage, is not
shown. Shaded boxes indicate the critical time-depen-
dent period in which the goal is myocardial salvage;
open boxes indicate the time-independent period in
which the goal is to open the infarct-related artery.
(From Gersh B, Stone G, White H, et al. Pharmacolog-
ical facilitation of primary percutaneous coronary inter-
vention for acute myocardial infarction: is the slope of
the curve the shape of the future? JAMA 2005;298:980;
with permission.)
thrombolytic agents, dosing regimens, and ad-
junctive treatments. Although developments in
catheter, balloon, stent, and device design contin-
ue and make opening occluded arteries easier, cur-
rent technology allows opening the infarct-related
 EARLY MANAGEMENT OF STEMI
artery in more than 90% of cases, and the ability
to improve clinical endpoints meaningfully with
design advances seems limited.
Fig. 5 is a hypothetical construct of the rela-
tionship of duration of symptoms before reperfu-
sion, mortality reduction, and extent of
myocardial salvage [97]. It is a helpful way to
judge future changes in MI care. In the example
given, a reduction of time from 12 to 6 hours
(from point A to B) results in only minimal clini-
cal benefit. A high-cost solution, then, probably
would not be cost effective. On the other hand,
moving from point B to point C (or from C to
D) would probably be associated with a substan-
tial clinical benefit. Moving from point D to point
C (or from C to B) represents a hazardous delay.
In evaluating new therapies, whether in the pro-
cess of delivering care, the drugs used, or the devi-
ces used in the catheterization, the goal remains
rapid and complete reperfusion.
Significant mortality reductions can be realized
if citizens are taught to activate emergency med-
ical services immediately with the onset of chest
pain, thereby reducing the all-important time
from symptom onset to delivering therapy. Other
process improvements are also possible, such as
increased compliance with national guidelines and
continued encouragement to deliver therapy to
patients who qualify for it. As therapies become
increasingly complex, the development of special-
ized MI hospitals built on the model of trauma
centers may prove beneficial. Future treatment
strategies must also attempt to provide earlier and
sustained reperfusion at both the epicardial and
tissue level. This advance may involve continued
drug, device, and dosing developments as well as
novel therapies to mitigate reperfusion injury and
delay or limit myocyte necrosis. Ultimately, ear-
lier patient presentation to the health care system
followed by rapid, appropriately selected reperfu-
sion therapy will result in more lives saved.
References
[1] Reimer K, Lowe J, Rasmussen M, et al. The wave-
front phenomenon of ischemic cell death: 1. Myocar-
dial infarct size vs duration of coronary occlusion in
dogs. Circulation 1977;56:786–94.
[2] Fibrinolytic Therapy Trialists’ (FTT) Collaborative
Group. Indications for fibrinolytic therapy in sus-
pected acute myocardial infarction: collaborative
overview of early mortality and major morbidity
results from all randomised trials of more than
1000 patients. Lancet 1994;343:311–22.
47
[3] Antman E, Anbe D, Armstrong P, et al. ACC/AHA
guidelines for the management of patients with ST-
elevation myocardial infarctiondexecutive summary:
a report of the American College of Cardiology/
American Heart Association Task Force on Practice
Guidelines (Writing Committee to Revise the 1999
Guidelines for the Management of Patients With
Acute Myocardial Infarction). Circulation 2004;
110:588.
[4] Canadian Cardiovascular Society Working Group.
Applying the new STEMI guidelines: 1. Reperfusion
in acute ST-segment elevation myocardial infarc-
tion. CMAJ 2004;171:1039–41.
[5] Fletcher A, Alkjaersig N, Smyrniotis F, et al. Treat-
ment of patients suffering from early myocardial in-
farction with massive and prolonged streptokinase
therapy. Trans Assoc Am Physicians 1958;71:
287–96.
[6] Rentrop P, Blanke H, Karsch K, et al. Selective
intracoronary thrombolysis in acute myocardial in-
farction and unstable angina pectoris. Circulation
1981;63:307–17.
[7] Chazov E, Mateeva L, Mazaev A, et al. Intracoro-
nary administration of fibrinolysis in acute myocar-
dial infarction. Ter Arkh 1976;48:8–19.
[8] Collen D, Topol E, Tiefenbrunn A, et al. Coronary
thrombolysis with recombinant human tissue-type
plasminogen activator. A prospective, random-
ized, placebo-controlled trial. Circulation 1984;70:
1012–7.
[9] Schroder R, Biamino G, Leitner E-R, et al. Intrave-
nous short-term infusion of streptokinase in acute
myocardial infarction. Circulation 1983;67:536–48.
[10] Ganz W, Geft I, Shah P, et al. Intravenous strepto-
kinase in evolving acute myocardial infarction. Am
J Cardiol 1984;53:1209–16.
[11] Waters R, Mahaffey K, Granger C, et al. Current
perspectives on reperfusion therapy for acute ST-
segment elevation myocardial infarction: integrating
pharmacologic and mechanical reperfusion strate-
gies. Am Heart J 2003;146:958–68.
[12] Boersma E, Maas A, Deckers J, et al. Early throm-
bolytic treatment in acute myocardial infarction:
reappraisal of the golden hour. Lancet 1996;348:
771–5.
[13] Morrison L, Verbeek P, McDonald A, et al. Mortal-
ity and prehospital thrombolysis for acute myocar-
dial infarction: a meta-analysis. JAMA 2000;283:
2686–92.
[14] Baigent C, Collins R, Appleby P, et al. ISIS-2: 10
year survival among patients with suspected acute
myocardial infarction in randomised comparison
of intravenous streptokinase, oral aspirin, both, or
neither. The ISIS-2 (Second International Study of
Infarct Survival) Collaborative Group. BMJ 1998;
316:1337–43.
[15] Franzosi M, Santoro E, De Vita C, et al. Ten-year
follow-up of the first megatrial testing thrombolytic
therapy in patients with acute myocardial infarction:
48 SURA & KELEMEN
results of the Gruppo Italiano per lo Studio della
Sopravvivenza nell’Infarto-1 study. The GISSI
Investigators. Circulation 1998;98:2659–65.
[16] Furman M, Dauerman H, Goldberg R, Yarzebski J,
et al. Twenty-two year (1975 to 1997) trends in the
incidence, in-hospital and long-term case fatality
rates from initial Q-wave and non-Q-wave myocar-
dial infarction: a multi-hospital, community-wide
perspective. J Am Coll Cardiol 2001;37:1571–80.
[17] Khan I, Gowda R. Clinical perspectives and thera-
peutics of thrombolysis. Int J Cardiolol 2003;91:
115–27.
[18] Late Assessment of Thrombolytic Efficacy (LATE)
study with alteplase 6–24 hours after onset of acute
myocardial infarction. Lancet 1993;342:759–66.
[19] Rogers WJCJ, Lambrew CT, et al. Temporal
trends in the treatment of over 1.5 million patients
with myocardial infarction in the US from 1990
through 1999: the National Registry of Myocardial
Infarction 1, 2 and 3. J Am Coll Cardiol 2000;36:
2056–63.
[20] Fox KA. An international perspective on acute cor-
onary syndrome care: Insights from the global regis-
try of acute coronary events. Am Heart J 2004;
148(Suppl 5):S40–5.
[21] Fox K, Goodman S, Klein W, et al. Management of
acute coronary syndromes. Variations in practice
and outcome; findings from the Global Registry of
Acute Coronary Events (GRACE). Eur Heart J
2002;23:1177–89.
[22] Eagle K, Goodman S, Avezum A, et al. Practice var-
iation and missed opportunities for reperfusion in
ST–segment-elevation myocardial infarction: find-
ings from the Global Registry of Acute Coronary
Events (GRACE). Lancet 2002;359:373–7.
[23] The GUSTO investigators. An international ran-
domized trial comparing four thrombolytic strate-
gies for acute myocardial infarction. N Engl J Med
1993;329:673–82.
[24] Gore J, Granger C, Simoons M, et al. Stroke after
thrombolysis. Mortality and functional outcomes
in the GUSTO-I trial. Global Use of Strategies to
Open Occluded Coronary Arteries. Circulation
1995;92:2811–8.
[25] Berkowitz S, Granger C, Pieper K, et al. Incidence
and predictors of bleeding after contemporary
thrombolytic therapy for myocardial infarction.
The Global Utilization of Streptokinase and Tissue
Plasminogen activator for Occluded coronary
arteries (GUSTO) I Investigators. Circulation
1997;95:2508–16.
[26] Giugliano R. Braunwald E, for the TTS Group.
Selecting the best reperfusion strategy in ST-eleva-
tion myocardial infarction: it’s all a matter of time.
Circulation 2003;108:2828–30.
[27] Young JJ, Kereiakes DJ. Pharmacologic reperfusion
strategies for the treatment of ST-segment elevation
myocardial infarction. Rev Cardiovasc Med 2003;4:
216–27.
[28] The Continuous Infusion versus Double-Bolus Ad-
ministration of Alteplase (COBALT) Investigators.
A comparison of continuous infusion of alteplase
with double-bolus administration for acute myo-
cardial infarction. N Engl J Med 1997;337:
1124–30.
[29] Holmes DJ, Califf R, Topol E. Lessons we have
learned from the GUSTO trial. Global Utilization
of Streptokinase and Tissue Plasminogen Activator
for Occluded Arteries. J Am Coll Cardiol 1995;25:
10S–7S.
[30] Labinaz M, Sketch MJ, Ellis S, et al. Outcome of
acute ST-segment elevation myocardial infarction
in patients with prior coronary artery bypass surgery
receiving thrombolytic therapy. Am J Cardiol 2001;
141:469–77.
[31] Brieger D, Mak K, White H, et al. Benefit of early
sustained reperfusion in patients with prior myocar-
dial infarction (the GUSTO-I trial). Global Utiliza-
tion of Streptokinase and TPA for occluded arteries.
Am J Cardiol 1998;81:282–7.
[32] The GUSTO Angiographic Investigators. The
effects of tissue plasminogen activator, streptoki-
nase, or both on coronary-artery patency, ventricu-
lar function, and survival after acute myocardial
infarction. N Engl J Med 1993;329:1615–22.
[33] The Global Use of Strategies to Open Occluded
Coronary Arteries (GUSTO III) Investigators. A
comparison of reteplase with alteplase for acute
myocardial infarction. N Engl J Med 1997;337:
1118–23.
[34] International Joint Efficacy Comparison of Throm-
bolytics (INJECT). Randomised. double-blind com-
parison of reteplase double-bolus administration
with streptokinase in acute myocardial infarction:
trial to investigate equivalence. Lancet 1995;346:
329–36.
[35] Van de Werf F, Cannon C, Luyten A, et al.
Safety assessment of single-bolus administration
of TNK tissue-plasminogen activator in acute
myocardial infarction: the ASSENT-1 trial. The
ASSENT-1 Investigators. Am Heart J 1999;137:
786–91.
[36] Assessment of the Safety and Efficacy of a New
Thrombolytic Investigators. Single-bolus tenecte-
plase compared with front-loaded alteplase in acute
myocardial infarction: the ASSENT-2 double-blind
randomised trial. Lancet 1999;354:716–22.
[37] Thiemann D, Coresh J, Schulman S. Lack of benefit
for intravenous thrombolysis in patients with myo-
cardial infarction who are older than 75 years. Circu-
lation 2000;101:2239–46.
[38] White HD. Thrombolytic therapy in the elderly.
Lancet 2000;356:2028–30.
[39] Gibson C, Murphy S, Menown I, et al. Determi-
nants of coronary blood flow after thrombolytic
administration. TIMI Study Group. Thrombolysis
in Myocardial Infarction. J Am Coll Cardiol 1999;
34:1403–12.
 EARLY MANAGEMENT OF STEMI
[40] The GUSTO Investigators. An international ran-
domized trial comparing four thrombolytic strate-
gies for acute myocardial infarction. N Engl J Med
1993;329:673–82.
[41] Gibson C, Murphy S, Marble S, et al. Can we replace
the 90-minute thrombolysis in myocardial infarction
(TIMI) flow grades with those at 60 minutes as a pri-
mary end point in thrombolytic trials? Am J Cardiol
2001;87:450–3.
[42] French J, Ellis C, Webber B, et al. Abnormal coro-
nary flow in infarct arteries 1 year after myocardial
infarction is predicted at 4 weeks by corrected
Thrombolysis in Myocardial Infarction (TIMI)
frame count and stenosis severity. Am J Cardiol
1998;81:665–71.
[43] Gibson C, Cannon C, Murphy S, et al. Relationship
of TIMI myocardial perfusion grade to mortality af-
ter administration of thrombolytic drugs. Circula-
tion 2000;101:125–30.
[44] The GUSTO Investigators. The effects of tissue plas-
minogen activator, streptokinase, or both on coro-
nary-artery patency, ventricular function, and
survival after acute myocardial infarction. N Engl
J Med 1993;329:1615–22.
[45] The TIMI Study Group. The Thrombolysis in Myo-
cardial Infarction (TIMI) trial. N Engl J Med 1985;
312:932–6.
[46] Grines C, Browne K, Marco J. Primary Angioplasty
in Myocardial Infarction Study Group (PAMI). A
comparison of immediate angioplasty with throm-
bolytic therapy for acute myocardial infarction.
N Engl J Med 1993;329:673–9.
[47] Fath-Ordoubadi F, Huehns T, Al-Mohammad A,
et al. Significance of the thrombolysis in myocardial
infarction scoring system in assessing infarct-related
artery reperfusion and mortality rates after acute
myocardial infarction. Am Heart J 1997;134:62–8.
[48] Smith D. For and against: primary angioplasty
should be first line treatment for acute myocardial
infarction. BMJ 2004;328:1254–8.
[49] Stone G, Grines C, Browne K, et al. Influence of
acute myocardial infarction location on in-hospital
and late outcome after primary percutaneous trans-
luminal coronary angioplasty versus tissue plasmin-
ogen activator therapy. Am J Cardiol 1996;78:
19–25.
[50] Weaver W, Simes R, Betriu A, et al. Comparison
of primary coronary angioplasty and intravenous
thrombolytic therapy for acute myocardial infarc-
tion: a quantitative review. JAMA 1997;278:2093–8.
[51] Keeley E, Boura J, Grines C. Primary angioplasty
versus intravenous thrombolytic therapy for acute
myocardial infarction: a quantitative review of 23
randomised trials. Lancet 2003;361:13–20.
[52] Le May M, Labinaz M, Richard F, et al. Stenting
versus thrombolysis in acute myocardial infarction
trial (STAT). J Am Coll Cardiol 2001;37:985–91.
[53] Kastrati A, Mehilli J, Dirschinger J, et al. Myocar-
dial salvage after coronary stenting plus abciximab
49
versus fibrinolysis plus abciximab in patients with
acute myocardial infarction: a randomised trial.
Lancet 2002;359:920–5.
[54] Andersen H, Nielsen T, Rasmussen K, et al. A com-
parison of coronary angioplasty with fibrinolytic
therapy in acute myocardial infarction. N Engl J
Med 2003;349:733–42.
[55] Lemos P, Saia F, Hofma S, et al. Short- and long-
term clinical benefit of sirolimus-eluting stents com-
pared to conventional bare stents for patients with
acute myocardial infarction. J Am Coll Cardiol
2004;43:704–8.
[56] Saito S, Hosokawa F, Kim K, et al. Primary stent
implantation without Coumadin in acute myocar-
dial infarction. J Am Coll Cardiol 1996;28:74–81.
[57] Bauters C, Lablanche J, Van Belle E, et al. Effects of
coronary stenting on restenosis and occlusion after
angioplasty of the culprit vessel in patients with
recent myocardial infarction. Circulation 1997;96:
2854–8.
[58] Stone G, Brodie B, Griffin J, et al. Clinical and an-
giographic follow-up after primary stenting in acute
myocardial infarction: the Primary Angioplasty in
Myocardial Infarction (PAMI) stent pilot trial. Cir-
culation 1999;99:1548–54.
[59] Nordmann A, Hengstler P, Harr T, et al. Clinical
outcomes of primary stenting versus balloon angio-
plasty in patients with myocardial infarction: a
meta-analysis of randomized controlled trials. Am
J Med 2004;116:253–62.
[60] Zhu M, Feit A, Chadow H, Alam M, et al. Primary
stent implantation compared with primary balloon
angioplasty for acute myocardial infarction:
a meta-analysis of randomized clinical trials. Am J
Cardiol 2001;88:297–301.
[61] Grzybowski M, Clements E, Parsons L. Mortality
benefit of immediate revascularization of acute ST-
segment elevation myocardial infarction in patients
with contraindications to thrombolytic therapy:
a propensity analysis. JAMA 2003;290:1891–8.
[62] Morrow D, Antman E, Charlesworth A, et al. TIMI
risk score for ST-elevation myocardial infarction:
a convenient, bedside, clinical score for risk assess-
ment at presentation: an Intravenous nPA for Treat-
ment of Infarcting Myocardium Early II trial
substudy. Circulation 2000;102:2031–7.
[63] Morrow D, Antman E, Giugliano R, et al. A simple
risk index for rapid initial triage of patients with ST-
elevation myocardial infarction: an InTIME II sub-
study. Lancet 2001;358:1571–5.
[64] Morrow DA, Antman EM, Parsons L, et al. Appli-
cation of the TIMI risk score for ST-elevation MI
in the National Registry of Myocardial Infarction
3. JAMA 2001;286:1356–9.
[65] Marchioli R, Avanzini F, Barzi F, et al. Assessment
of absolute risk of death after myocardial infarction
by use of multiple-risk-factor assessment equations:
GISSI-Prevenzione mortality risk chart. Eur Heart J
2001;22:2085–103.
50 SURA & KELEMEN
[66] Stone G, Grines C, Browne K. Predictors of in-hos-
pital and 6-month outcome after acute myocardial
infarction in the reperfusion era: the Primary Angio-
plasty in Myocardial Infarction (PAMI) trial. J Am
Coll Cardiol 1995;25:370–7.
[67] O’Neill W, Menko J, Gibbons R. Lessons from the
pooled outcome of the PAMI, ZWOLLE and
Mayo Clinic randomized trials of primary angio-
plasty versus thrombolytic therapy of acute myocar-
dial infarction. J Invest Cardiol 1998;10:4A–10A.
[68] Mehta R, Sadiq I, Goldberg R, et al. Effectiveness of
primary percutaneous coronary intervention com-
pared with that of thrombolytic therapy in elderly
patients with acute myocardial infarction. Am Heart
J 2004;147:253–9.
[69] Bates E, Topol E. Limitations of thrombolytic ther-
apy for acute myocardial infarction complicated by
congestive heart failure and cardiogenic shock. J
Am Coll Cardiol 1991;18:1077–84.
[70] Hochman J, Sleeper L, Webb J, et al. Early revascu-
larization in acute myocardial infarction compli-
cated by cardiogenic shock. SHOCK Investigators.
Should We Emergently Revascularize Occluded
Coronaries for Cardiogenic Shock. N Engl J Med
1999;341:625–34.
[71] Sleeper L, Ramanathan K, Lejemtel T, et al. Func-
tional status and quality of life after emergency re-
vascularization for cardiogenic shock complicating
acute myocardial infarction. SHOCK Investigators.
J Am Coll Cardiol 2005;46:266–73.
[72] Garcia E, Elizaga J, Perez-Castellano N, et al. Pri-
mary angioplasty versus systemic thrombolysis in
anterior myocardial infarction. J Am Coll Cardiol
1999;33:605–11.
[73] O’Keefe J, Bailey W, Rutherford B. Primary angio-
plasty for acute myocardial infarction in 1000
consecutive patients: results in an unselected
population and high-risk subgroups. Am J Cardiol
1993;72:107G–15G.
[74] DeGeare V, Dangas G, Stone G, et al. Interven-
tional procedures in acute myocardial infarction.
Am Heart J 2001;141:15–24.
[75] Magid D, Calonge B, Rumsfeld J, et al. Relation be-
tween hospital primary angioplasty volume and
mortality for patients with acute MI treated with pri-
mary angioplasty versus thrombolytic therapy.
JAMA 2000;284:3131–8.
[76] Canto J, Every N, Magid D, et al. The volume of
primary angioplasty procedures and survival after
acute myocardial infarction. National Registry of
Myocardial Infarction 2 Investigators. N Engl J
Med 2000;342:1573–80.
[77] Cannon C, Gibson C, Lambrew C, et al. Relation-
ship of symptom-onset-to-balloon time and door-
to-balloon time with mortality in patients undergo-
ing angioplasty for acute myocardial infarction.
JAMA 2000;283:2941–7.
[78] Brodie B, Stuckey T, Wall T, et al. Importance of
time to reperfusion for 30 day and late survival
and recovery of left ventricular function after pri-
mary angioplasty for acute myocardial infarction.
J Am Coll Cardiol 1998;32:1312–9.
[79] De Luca G, Suryapranata H, Zijlstra F, et al. Symp-
tom-onset-to-balloon time and mortality in patients
with acute myocardial infarction treated by primary
angioplasty. J Am Coll Cardiol 2003;42:991–7.
[80] De Luca G, Suryapranata H, Ottervanger JP, et al.
Time delay to treatment and mortality in primary
angioplasty for acute myocardial infarction: every
minute of delay counts. Circulation 2004;109:
1223–5.
[81] De Luca G, van’t Hof AW, de Boer MJ, et al.
Time-to-treatment significantly affects the extent
of ST-segment resolution and myocardial blush
in patients with acute myocardial infarction trea-
ted by primary angioplasty. Eur Heart J 2004;25:
1009–13.
[82] Sabatine M, Cannon C, Gibson C, et al. Addition of
clopidogrel to aspirin and fibrinolytic therapy for
myocardial infarction with ST-segment elevation.
N Engl J Med 2005;352:1179–89.
[83] Mehta S, Yusuf S, Diaz R, et al. Effect of glucose-
insulin-potassium infusion on mortality in patients
with acute ST-segment elevation myocardial infarc-
tion: the CREATE-ECLA randomized controlled
trial. JAMA 2005;293:437–46.
[84] Simoons M, Krzeminska-Pakula M, Alonso A, et al.
Improved reperfusion and clinical outcome with
enoxaparin as an adjunct to streptokinase thrombol-
ysis in acute myocardial infarction. The AMI-SK
study. Eur Heart J 2002;23:1282.
[85] Topol E. GUSTO V Investigators. Reperfusion
therapy for acute myocardial infarction with fibrino-
lytic therapy or combination reduced fibrinolytic
therapy and platelet glycoprotein IIb/IIIa inhibition:
the GUSTO V randomised trial. Lancet 2001;357:
1905–14.
[86] Dalby M, Bouzamondo A, Lechat P, et al. Transfer
for primary angioplasty versus immediate thrombol-
ysis in acute myocardial infarction: a meta-analysis.
Circulation 2003;108:1809–14.
[87] Widimsky P, Budesinsky T, Vorac D, et al. Long dis-
tance transport for primary angioplasty vs immedi-
ate thrombolysis in acute myocardial infarction.
Final results of the randomized national multicentre
trial–PRAGUE-2. Eur Heart J 2003;24:94–104.
[88] Grines C, Westerhausen DJ, Grines L, et al. A ran-
domized trial of transfer for primary angioplasty
versus on-site thrombolysis in patients with high-
risk myocardial infarction: the Air Primary Angio-
plasty in Myocardial Infarction study. J Am Coll
Cardiol 2002;39:1713–9.
[89] Nallamothu B, Bates E. Percutaneous coronary in-
tervention versus fibrinolytic therapy in acute myo-
cardial infarction: is timing (almost) everything?
Am J Cardiol 2003;92:824–6.
[90] Aversano T, Aversano L, Passamani E, et al.
Thrombolytic therapy vs primary percutaneous
 EARLY MANAGEMENT OF STEMI
coronary intervention for myocardial infarction in
patients presenting to hospitals without on-site car-
diac surgery: a randomized controlled trial. JAMA
2002;287:1943–51.
[91] TIMI Study Group. Comparison of invasive and
conservative strategies after treatment with intrave-
nous tissue plasminogen activator in acute myocar-
dial infarction: results of the thrombolysis in
myocardial infarction (TIMI) phase II trial. N
Engl J Med 1989;320:618–27.
[92] The TIMI Research Group. Immediate vs
delayed catheterization and angioplasty following
thrombolytic therapy for acute myocardial in-
farction. TIMI II A results. JAMA 1988;260:
2849–58.
[93] Ross A, Coyne KS, Reiner JS, et al. A randomized
trial comparing primary angioplasty with a strategy
of short-acting thrombolysis and immediate planned
rescue angioplasty in acute myocardial infarction:
the PACT trial. PACT investigators. Plasminogen-
activator Angioplasty Compatibility Trial. J Am
Coll Cardiol 1999;34:1954–62.
51
[94] Antman E, Van de Werf F. Pharmacoinvasive ther-
apy: the future of treatment for ST-elevation myo-
cardial infarction. Circulation 2004;109:2480–6.
[95] Ellis SGAP, Betriu A, Brodie B, et al. Facilitated In-
tervention with Enhanced Reperfusion Speed to
Stop Events Investigators. Facilitated percutaneous
coronary intervention versus primary percutaneous
coronary intervention: design and rationale of the
Facilitated Intervention with Enhanced Reperfusion
Speed to Stop Events (FINESSE) trial. Am Heart J
2004;147:1–7.
[96] Mahaffey K, Granger C, Nicolau J, et al. Effect of
pexelizumab, an anti-C5 complement antibody, as
adjunctive therapy to fibrinolysis in acute myocar-
dial infarction: the Complement Inhibition in Myo-
cardial Infarction Treated with Thrombolytics
(COMPLY) trial. Circulation 2003;108:1176–83.
[97] Gersh B, Stone G, White H, et al. Pharmacological
facilitation of primary percutaneous coronary inter-
vention for acute myocardial infarction: is the slope
of the curve the shape of the future? JAMA 2005;
293:979–86.
 Cardiol Clin 24 (2006) 53–65

Emergency Cardiac Imaging: State of the Art

Dick Kuo, MDa,b,*, Vasken Dilsizian, MDa,b, Rajnish Prasad, MDa,b,
Charles S. White, MDa,b
a
University of Maryland School of Medicine, 655 West Baltimore Street, Baltimore, MD 21201, USA
b
University of Maryland Medical System, 22 South Greene Street, Baltimore, MD 21201, USA

One of the most difficult challenges for emer-
gency physicians is to determine whether chest
pain is cardiac related and if the patient is at
increased risk for a cardiac event (eg, nonfatal
myocardial infarction or death). A certain group
of high-risk patients can be identified readily
based on history, ECG changes, and cardiac
enzyme elevations. Based on recommendations
by the American College of Cardiology (ACC)
and American Heart Association (AHA), these
patients usually undergo urgent coronary angiog-
raphy [1]. A larger group of patients presenting
with a less urgent clinical scenario and varying de-
grees of pretest likelihood for coronary artery dis-
ease require additional testing to evaluate their
cardiac risk. Several noninvasive imaging modali-
ties are helpful in this group of patients. Nuclear
stress perfusion testing and stress echocardiogra-
phy are useful in risk stratifying these patients,
and new-generation CT scanners and MRI may
soon develop their own roles.
Cardiac angiography remains the reference
standard for imaging of the coronary vessels and
provides an avenue for intervention, but cardiac
catherization is an invasive procedure. An esti-
mated 1.46 million cardiac catherizations were
performed in 2002, although only 657,000 percu-
taneous transluminal coronary angioplasty pro-
cedures were performed [2]. Because most
cardiac catherizations are diagnostic, there has
been a long search for a noninvasive technique
to diagnose coronary artery disease and visualize
* Corresponding author. 110 South Paca Street Sixth
Floor, Suite 200, Baltimore, MD 21201.
E-mail address: dkuo@umaryland.edu (D. Kuo).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.001
the coronary vessels, but until recently few techni-
ques have been satisfactory.
Imaging of the heart in the emergency de-
partment begins with the plain chest radiograph.
Although the plain chest radiograph states little
about the coronary vessels, it provides important
background information. Other options now
available to visualize the coronary arteries with-
out cardiac catherization include electron beam
CT (EBCT), multidetector or multislice CT
(MDCT) with CT angiography (CTA), and car-
diac MR (CMR) with angiography. Provocative
and nuclear testing can also provide much useful
information in the evaluation of the patient who
has suspected angina.
Plain chest radiograph
The plain chest radiograph has served for many
years as a first-line imaging technique in the
assessment of the cardiac patient in the emergency
room. The chest radiograph is quite sensitive for
diagnosis of some noncardiac causes of chest pain
including pneumonia, pneumothorax, and rib
fractures.
Direct evidence of myocardial ischemia is often
absent on chest radiographs, but indirect evidence
may be present in the form of atherosclerotic
calcification of vessels. This calcification is usually
most evident in the aorta but is more specific when
found in the coronary arteries. The sensitivity for
detection of coronary artery calcification on
radiography is less than 50%. The usual location
of visible coronary artery calcification is in the
coronary triangle in the mid-upper part of the left
heart corresponding to the proximal portions of
the left coronary arteries [3]. Calcification may
cardiology.theclinics.com
54
also be present on the lateral radiograph arising
from the aortic root (Fig. 1). Data from a fluoro-
scopic study suggest that radiographically evident
coronary calcification is associated with a higher
likelihood of significant coronary artery stenosis
[4].
A plain chest radiograph may also be useful for
assessment of complications of early episodes of
myocardial ischemia. An enlarged cardiac silhou-
ette may be evidence of a previous myocardial
event. Calcification along the left heart border is
often an indication of a prior myocardial infarction
(Fig. 2) [5]. The presumed mechanism is impaired
wall motion followed by local thrombus forma-
tion, which may ultimately calcify. A focal bulge
of the left heart border may represent a postinfarct
myocardial aneurysm or pseudoaneurysm [6].
Other indirect signs of ischemic chest pain may
be identified, including congestive heart failure.
There is some correlation between the radiographic
findings and the severity of congestive heart
failure. In mild heart failure, cephalization may
be present consisting of reduced flow to lower lobe
vessels and diversion of flow to upper lobes.
Cephalization requires a gravitational gradient
and therefore is difficult to recognize on a supine
or semierect radiograph. More severe heart failure
is associated with interstitial and alveolar (air-
space) pulmonary edema, respectively [7]. In a re-
cent study of patients presenting to the
emergency department with acute dyspnea, a plain
chest radiograph showing enlarged heart size iden-
tified patients for whom a final diagnosis of heart
failure was confirmed by two cardiologists with
a sensitivity of 88% and a specificity of 72% [8].
Fig. 1. Coronary calcification. (A) Lateral chest radiograph shows coronary artery calcification overlying the anterior
cardiac silhouette (arrow). (B) Nonenhanced CT scan shows that the anterior calcification corresponds to the right
coronary artery (arrow). The left circumflex artery is also calcified (arrowhead).
KUO et al
The plain chest radiograph is unlikely to be
replaced because it provides a large amount of
useful information, and cardiac patients will al-
most uniformly require a chest radiograph to
exclude other potential diagnoses associated with
cardiac symptoms.
Electron beam CT
Current EBCT scanners deploy temporal reso-
lutions of 50 to 100 milliseconds and ECG gating
to evaluate the cardiac anatomy. For comparison,
conventional angiography has a temporal resolu-
tion of less than 10 milliseconds. Multiple studies
have evaluated the ability of EBCT to evaluate
coronary stenosis and provide prognostic infor-
mation for patients who have coronary calcifica-
tions. CT scanning, in particular EBCT, has been
used to risk stratify patients who have suspected
coronary artery disease by demonstrating coro-
nary calcium. The presence of coronary calcium
indicates coronary artery disease and is conven-
tionally measured with the method described by
Agatston and colleagues [9], in which the extent
and density of coronary calcification are used to
derive a global score. Coronary calcium is strongly
associated with coronary artery disease and has
been shown to have an odds ratio of 13.7 for any
coronary artery disease and an odds ratio of 10.3
for obstructive coronary artery disease [10,11].
This association must be balanced with the results
of a recent meta-analysis that found only a mo-
derately increased risk for cardiac events (un-
stable angina, myocardial infarction, need for
 EMERGENCY CARDIAC IMAGING
Fig. 2. Myocardial calcification. (A) Posteroanterior radiograph shows curvilinear calcification along the left heart bor-
der (arrow) in a patient with an implantable defibrillator. (B) Nonenhanced CT shows curvilinear myocardial calcifica-
tion in the left ventricular apex (arrow). Defibrillator wires are also identified (arrowhead).
revascularization, cardiac death) associated with
coronary calcifications in asymptomatic popula-
tions [12].
Several studies have also sought to evaluate
coronary stenosis greater than 50% with EBCT.
Reddy and colleagues [13] found an overall sensi-
tivity of 88% and specificity of 79% in 23 patients
although coronary artery calcifications resulted in
decreased specificity. Budoff and colleagues [14]
studied 52 patients and reported an overall sensi-
tivity of 78% and a specificity of 91%. In this study
11% of the cardiac segments were noninterpret-
able, usually because of motion. The authors also
noted difficulty in viewing the right coronary and
circumflex arteries. Schmermund and colleagues
[15] also reported increased false negatives second-
ary to segmental calcification with a sensitivity of
82% and a specificity of 88% in 28 patients. An-
other study by Achenbach and colleagues [16] dis-
covered a sensitivity of 92% and a specificity of
94% in 125 patients although there 25% of seg-
ments were noninterpretable in this study. These
studies illustrate the capabilities of EBCT but
may not be applicable to patients in an emergency
department where risk stratification and outcomes
are more important measures of the success of
a particular test, and where patients cannot be
excluded secondary to their ‘‘noninterpretable
segments.’’
A few studies have assessed the use of EBCT in
the emergency department setting to evaluate
patients who have angina-like chest pain [17–19].
Table 1 summarizes studies in which patients pre-
senting to the emergency department with chest
pain were evaluated by new techniques. Using
the presence of coronary calcium as a marker,
55
these studies demonstrate a sensitivity ranging
from 88% to 100% for coronary stenosis or cardi-
ac events and negative predictive values of 97% to
100%. In addition, a high negative predictive val-
ue was found. In the study by Laudon and
colleagues [17], no patient presenting with chest
pain who had a negative EBCT had a cardiac
event in the 4 months after presentation to the
emergency department [17]. In the study by
McLaughlin and colleagues [19], the 1-month car-
diac event rate was 2%, compared with 8% in pa-
tients who had coronary artery calcium (CAC)
scores greater than one. More recently Georgiou
and colleagues [18] found a strong association be-
tween the age- and gender-adjusted CAC score
and a subsequent cardiac event in a cohort of
192 patients who had undergone EBCT during
the course of their emergency department evalua-
tion for chest pain. Overall, the 1-year annualized
rate for cardiac events was 0.6% for patients who
had a CAC score of zero compared with a cardiac
event rate of 13.9% in patients who had CAC
scores greater than 400 (average follow-up after
presentation in the emergency department was
50 G 10 months with a range of 1–84 months).
The results of this study are somewhat difficult
to generalize, because follow-up time was not
standard; instead, results were statistically annual-
ized. This study, however, seems to confirm that
a CAC score of zero has a high negative predictive
value for cardiac events.
Other strategies for managing patients who
have chest pain include EBCT replacement of
stress testing, the combination of EBCT and stress
testing [20], or EBCT scanning in patients who
have indeterminate stress results [21].
56 KUO et al

Table 1
Studies evaluating patients presenting to the emergency department patients with chest pain
FU period
Study n Sensitivity Specificity PPV NPV Standards and outcome Event Rate
EBCT
Lauden [17] 1999 100 100 63 30 100 þ test, sten O 40 4 mo CE 0% in neg EBCT
vs CAC O 0
Georgiou [18] 2001 192 97 55 48 97 CAC O 0 vs CE 1 mo–84 mo 1 yr annualized
(mean 50 mo)
CE
93 59 55 95 CAC O 4 vs CE 0.6% CAC ¼ 0
13.9%
CAC O 400
McLaughlin [19] 1999 134 88* 37* 8* 98% CAC O 1 1 mo CE 2% versus 8%
MDCT
White [31] 2005 69 83 96 83 96 CTA O 50% not studied
stenosis versus
final diagnosis
CMR
Kwong [36] 2003 161 84 85 ns ns ACS, NSTEMI, not studied
UA 70%sten or
true þ stress test
Takahashi [37] 2004 18 78 89 78 89 AMI and not studied
ultrasound
Abbreviations: ACS, acute coronary syndrome; AMI, acute myocardial infarction; CAC, coronary artery calcifica-
tion score; CE, cardiac events; CMR, cardiac magnetic resonance imaging; EBCT, electron beam CT; FU, follow up;
MDCT, multidetector CT; NPV, negative predictive value; NSTEMI, nonST elevation myocardial infarction; PPV,
positive predictive value; UA, unstable angina.
* Value not provided in text but calculated using standard 2
2 table from which NPV was calculated.

The consensus statement on CAC and EBCT
from the ACC/AHA can be summarized briefly.
Negative tests occur in most patients who have
angiographically normal coronaries and may be
consistent with a low risk of cardiovascular risk in
the next 2 to 5 years. A positive EBCT confirms
the presence of atherosclerotic plaque and is best
correlated with total amount of plaque burden.
The greater the amount of calcium, the greater the
likelihood of occlusive disease, and a high calcium
score may be consistent with moderate to high
cardiovascular risk in the next 2 to 5 years [22].
Multidetector CT
The latest generation of MDCT scanners fea-
tures ECG gating, submillimeter spatial resolution,
and relatively good temporal resolution that per-
mits increasingly accurate assessment of coronary
artery anatomy. Currently, scanners are available
with 64 detectors, a spatial resolution of 0.5 to 0.6
mm, and temporal resolution of 50 to 100 milli-
seconds. CT scanners are increasingly being placed
in the emergency suite, alleviating concerns about
monitoring some patients who have chest pain.
The improved technical parameters of MDCT
allow determination of the extent of coronary
calcification and acquisition of acceptable coro-
nary CTA, ventricular function, and, perhaps
myocardial perfusion.
The scanning protocol may be optimized to
assess the heart alone or be acquired as a compro-
mise between a coronary and lung CT protocol to
assess for pulmonary emboli and aortic dissection
also. Generally, 10 evenly spaced phases through-
out the cardiac cycle are obtained. This approach
permits selection of the phase with the least amount
of coronary artery motion, typically in early or late
diastole. Reconstructions centered along the curv-
ing centerline of the individual coronary arteries
are produced and evaluated for hard and soft
plaque and critical stenoses. Software exists for
a quantitative assessment of the extent of the
stenosis. Left ventricular ejection fraction can
also be derived from determination of end-systolic
and end-diastolic volumes. This postscan process-
ing currently is labor-intensive, and much effort is
being directed to streamlining this analysis.
 EMERGENCY CARDIAC IMAGING
MDCT technique requires intravenous con-
trast material and is usually done with an auto-
matic triggering mechanism that times the contrast
bolus so that opacification is optimized. Patients
who have renal insufficiency or significant contrast
allergies are thus not eligible. Beta-blockade,
typically with metoprolol, can be used in patients
who have a heart rate greater than 70 beats per
minute. This strategy has been shown to improve
image quality because of image degradation at
higher rates [23].
As with EBCT, there have been multiple
investigations with multidetector CTA to evaluate
stenosis. A study by Nieman and colleagues [24]
reported a sensitivity of 95% and a specificity of
86% in 59 patients and also reported higher accu-
racy for left main and left anterior descending ar-
teries than for circumflex and right coronary
arteries. This finding is thought to result from
the increased motion of those vessels. Ropers
and colleagues [25] conducted a study with 77 pa-
tients; with 12% excluded segments, sensitivity
was 92% and specificity was 93%.
In a more recent study evaluating coronary
arteries greater than 2 mm in patients undergoing
elective evaluation for chest pain, coronary CTA
has shown considerable potential, with sensitivity
and specificity of 83% and 97%, respectively [26].
Thirty-seven percent of patients in this study had
CAC scores of 400 or higher; if these patients were
excluded, CTA sensitivity increased to 89% and
specificity increased to 98%.
A PUB MED search (search terms: chest pain,
multislice or multidetector, and emergency) found
no studies using multislice CT for CTA of the
coronary anatomy to evaluate patients presenting
to the emergency department with chest pain,
although there have been a couple of case reports
showing acute myocardial infarction in patients
presenting with chest pain [27,28]. Also, a couple
of studies have used MDCT in acute coronary syn-
drome (ACS). Although it may be assumed that
these patients presented to the emergency depart-
ment initially, this information is not confirmed
in the text of the articles. One study evaluated ejec-
tion fraction and stenosis with a 4-slice scanner
[29], and another study used a predictive model
to determine if a use of a 16-slice MDCT could de-
crease the number of diagnostic cardiac angiogra-
phies [30].
Currently, the appropriate use and timing of
the MDCT in the emergency setting is unclear,
and the authors have devised a protocol to test
one scenario [31]. The authors have proposed that
57
the examination be obtained in patients who have
chest pain and an intermediate probability of an-
gina, as initially assessed by the emergency physi-
cian by examination and ECG. Patients who
have a high probability would be taken for emer-
gent coronary angiography; those with low proba-
bility are unlikely to benefit from MDCT. In the
authors’ protocol, patients were brought to the
emergency suite scanner between 30 minutes and
1 hour after initial assessment. The 16-slice CT
scan using a dual heart–lung protocol was intended
to provide a comprehensive evaluation of both cor-
onary and noncoronary causes of chest pain.
In this study, 69 patients met all criteria for
enrollment, 45 (65%) of whom otherwise would
not have undergone CT scanning [29]. Fifty-two
patients (75%) had no significant CT findings
and a final diagnosis of clinically insignificant
chest pain. Thirteen patients (18%) had significant
CT findings concordant with the final diagnosis
(10 cardiac, 3 noncardiac). Fig. 3 shows an exam-
ple of a curved planar reconstruction used to visu-
alize the coronary vessels. CT failed to suggest
a diagnosis in two patients (3%), both of whom
proved to have clinically significant coronary ar-
tery stenoses. In two patients (3%), CT overdiag-
nosed a coronary stenosis. Sensitivity and
specificity for the establishment of a cardiac cause
of chest pain were 83% and 96%, respectively.
Overall sensitivity and specificity for all cardiac
and noncardiac causes of chest pain were 87%
and 96%, respectively. The cardiac assessment
was done several hours or more after acquisition
of the CT scan because of software limitations.
The study suggests that MDCT is logistically
Fig. 3. CT scanning in the emergency room. A curved
planar reconstructed image of a diagonal branch shows
calcification and narrowing (arrow).
58 KUO
feasible and may prove useful if hardware and
software improvements continue. Current tech-
nology involves the use of 40- and 64-slice scan-
ners; studies evaluating their use in patients
presenting to the emergency department with
chest pain have yet to be published.
Cardiac MR and angiography
MRI is effective in evaluating myocardial
ischemia and thus has potential applications in
the emergency room setting. After intravenous
infusion of gadolinium chelate, myocardial perfu-
sion can be assessed with rapid temporal imaging.
Wall-motion abnormalities can be delineated with
bright blood cine imaging. Delayed images after
gadolinium enhancement are valuable to depict
myocardial viability. Delayed hyper-enhancement
10 to 20 minutes after injection is a strong in-
dicator of myocardial infarction (Fig. 4) [32].
In one of the earlier studies evaluating CMR
for detection of coronary stenoses, Regenfus and
colleagues [33] reported sensitivity of 94.4% and
specificity of 57.1% on a patient basis with 50 pa-
tients in the study. Only 76.6% of segments could
be evaluated, and the left circumflex could be eval-
uated in only 50% of cases. In a larger study of 109
patients, coronary magnetic resonance angiogra-
phy was performed before elective radiographic
coronary angiography, and the results of the two
diagnostic procedures were compared. Six hun-
dred thirty-six of 759 proximal and middle seg-
ments were interpretable on magnetic resonance
angiography (84%). In these segments, the sensi-
tivity, specificity, and accuracy for patients who
had disease of the left main coronary artery or
three-vessel disease were 100%, 85%, and 87%, re-
spectively. The negative predictive values for any
coronary artery disease and for left main artery
or three-vessel disease were 81% and 100%, re-
spectively [34]. In a recent brief report by van
Geuns and colleagues [35], CMR was found to
have only 46% sensitivity but 90% specificity for
stenosis that was greater than 50% in a small study
of 27 patients.
Again, fewer studies have examined the use-
fulness in emergency department patients. Kwong
and colleagues [36] assessed the use of CMR in
a prospective study of 161 patients who presented
to the emergency room with suspected ACS. In-
clusion criteria were an episode of chest pain last-
ing more than 30 minutes and an abnormal
but nondiagnostic ECG. Resting CMR was
et al
Fig. 4. CMR of myocardial infarction. Short-axis image
from CMR viability study shows an area of hyper en-
hancement in the anterior wall, indicating myocardial
infarction (arrow).
performed within 12 hours of presentation. The
image protocol consisted of perfusion, wall mo-
tion, and viability sequences. CMR demonstrated
a sensitivity of 84% and specificity of 85%, re-
spectively, for ACS.
Another small study of 18 patients in 2004 by
Takahashi and colleagues [37] also evaluated
CMR in patients who had ACS as defined by
acute myocardial infarction and ultrasound. Be-
cause the majority of patients in this study actual-
ly were classified as having acute myocardial
infarction, this study probably more accurately
appraises the utility of CMR in acute myocardial
infarction and than in true ACS.
CMR has the advantage of good spatial and
excellent temporal resolution, and Gadolinium
contrast agent is widely available. Nevertheless,
CMR is limited by the need for specialized, often
expensive equipment that may not be located near
the emergency room. Other issues that potentially
make CMR unfeasible for many emergency de-
partment patients are patient claustrophobia and
the need to monitor an acutely ill patient appro-
priately in the bore of the MR imager.
Other techniques
Although there are other new techniques avail-
able to evaluate patients who have chest pain that
better establish plaque composition, these new
techniques are invasive and investigational and
 EMERGENCY CARDIAC IMAGING 59

have little if any application in the emergency
department or in the initial evaluation of the
patient who has potential cardiac disease. These
techniques include intravascular ultrasound, opti-
cal coherence tomography, thermography, and
angioscopy and are beyond the scope of this
discussion.
Stress echocardiography
Stress echocardiography is readily available, is
relatively low in cost, and can assess cardiac
anatomy and function during stress. It also has
the advantage of providing incremental informa-
tion of value by evaluating baseline ventricular
function, valvular function, aortic root morphol-
ogy, and pericardial anatomy. Such information
can provide further insight into the possible causes
of the chest pain. Regional wall-motion abnor-
malities are early signs of myocardial ischemia and
provide an indirect evaluation of abnormal myo-
cardial perfusion [1] and coronary blood flow.
Wal-motion abnormalities at rest identify patients
who have had ischemic injury. The number of ab-
normal wall-motion segments is quantified by the
wall-motion score index (Fig. 5). The higher the
wall-motion score index, the greater the number
of abnormal segments and, thus, the higher risk
for the patient [38].
The decision to perform either an exercise or
pharmacologic stress echocardiogram depends on
the functional status of the patient. Ideally, an
exercise stress test should be performed because it
provides valuable physiologic information includ-
ing functional capacity. A normal exercise stress
echocardiogram confers an excellent prognosis.
The overall cardiac event rate (cardiac death,
nonfatal myocardial infarction) ranges from
0.9% to 1.1% per year. An abnormal study
increases the risk of a cardiac event by three to
four times [39–42].
Other factors may affect prognosis as well.
Exercise stress–induced wall-motion abnormali-
ties in the left anterior descending distribution
predict a fivefold higher cardiac event rate at
5 years than wall-motion abnormalities in other
regions [43]. Even with a normal stress echocar-
diogram, patients who have diabetes mellitus
have significantly higher cardiac event rates (6%
per year) than nondiabetics (2.7% per year)
[44], and hypertensive patients who have a normal

Fig. 5. Exercise echocardiogram recorded in a patient with a disease of the right coronary artery. The two left panels
were recorded at rest and the two right panels immediately after treadmill exercise; the top panels show diastole, and
the bottom panels show systole. In each panel the arrows note the location of the inferior wall endocardium at end-
diastole. At rest there is appropriate thickening and inward motion of the inferior wall that can be seen to move inward
through the body of the arrows. Immediately after exercise the proximal inferior wall (lower two arrows) becomes frankly
dyskinetic, and the mid and diastole portion of the inferior wall is akinetic. There is no incursion of the endocardium into
the previously placed arrows. (From Braunwald E. Heart disease: a textbook of cardiovascular medicine. 6th edition.
Philadelphia: Elsevier Inc.; 2001. p. 214; with permission.)
60 KUO
dobutamine stress echocardiogram have an over-
all higher cardiac event rate (1.8% per year) than
the general population (approximately 1% per
year) but a significantly lower rate than those
who have an abnormal study (3.8% per year)
[45].
When the patient is unable to exercise, phar-
macologic testing provides a valuable alternative
and provides similar prognostic information. The
most common drug used for a pharmacologic
stress test is dobutamine. The infusion begins at 5
mg/kg/min and is increased by 5- to 10-mg/kg/min
increments until target heart rate is achieved. Fre-
quently, atropine may be required if the dobut-
amine infusion does not achieve the target heart
rate. Continuous ECG monitoring is performed
throughout the stress test and the recovery period.
Echocardiogram images are obtained during the
pre-infusion period (resting state), at low-dose
stress, at peak stress (when target heart rate is
achieved), and then in recovery.
Several studies have found death and nonfatal
myocardial infarction rates of approximately
1.1% per year for a normal test [46,47]. The death
and nonfatal myocardial infarction rates for ab-
normal studies are about 7% per year [46–51].
In studies evaluating exercise and dobutamine
stress echocardiography, a normal study trans-
lates into a low cardiac event rate (0.8%–0.9%
per year) [52,53]. Abnormal studies could be fur-
ther stratified into intermediate (3.1% per year
cardiac event rate) and high (5.2% per year car-
diac event rate) risk groups based on the wall-
motion score index [52].
Because regional wall-motion abnormalities
usually precede the onset of definitive echocardio-
graphic signs of ischemia, echocardiographic de-
tection of regional left ventricular dysfunction has
been assessed as a tool to improve the diagnosis of
acute cardiac ischemia in the emergency room.
Among patients undergoing echocardiographic
study during active chest pain in the emergency
room, Peels and colleagues [54] found echocardio-
graphy to be highly sensitive for the detection of
myocardial infarction and acute ischemia (92%
and 88%, respectively). The specificity of this ap-
proach was limited, at 53% for infarction and
78% for ischemia. In the absence of ongoing
symptoms, the sensitivity of echocardiography
was limited [55]. Echocardiographic analysis in
these studies was limited to patients exhibiting
normal conduction systems and no prior myocar-
dial infarction, because both conduction distur-
bances and prior areas of infarction can cause
et al
regional wall-motion abnormalities in the absence
of acute ischemia.
Sabia and colleagues [56] examined the value of
regional wall-motion abnormality for the diagno-
sis of acute myocardial infarction in the emergency
room. The sensitivity for echocardiographically
detected regional wall-motion abnormalities to
identify acute ischemic heart disease presenting
as myocardial infarction was 93%. The specificity,
however, was modest (57%). These investigators
estimated that the use of echocardiography in the
emergency room could result in a 32% reduction
in hospital admissions, but this estimate was not
demonstrated in a prospective manner. In a small
subset of patients ultimately diagnosed as having
non-Q wave infarction, echocardiography failed
to demonstrate regional wall-motion abnormality.
False-negative findings by echocardiography have
also been observed by other investigators [57,58].
Recently, Kontos and colleagues [59] demon-
strated a high negative predictive value of normal
echocardiographic studies in patients who had
chest pain, which correlated with a benign prog-
nosis at 10 months. In a subsequent study [60],
these authors also compared myocardial perfusion
imaging with single-photon emission CT (SPECT)
and echocardiography in 185 patients who pre-
sented to the emergency room with chest pain
and who were considered to have low to moderate
risk of coronary ischemia based on history and
echocardiography. In 90% of the patients, acute
rest sestamibi perfusion and echocardiographic
studies were performed within 1 hour of each
other. The two techniques had similar sensitivities
and specificities for the detection of acute myocar-
dial infarction or acute myocardial ischemia. Fur-
ther confirmatory studies are needed to determine
the impact of symptom resolution on this compar-
ison, because the earlier studies of echocardiogra-
phy demonstrated that optimal sensitivity is
dependent on the presence of symptoms during
the emergency room evaluation [54–58]. The
most recent studies evaluating the role of dobut-
amine tele-echocardiography [61] and contrast
echocardiography [62] for patients presenting to
the emergency room with chest pain have also re-
ported favorable results. These techniques or tech-
nologies are not in widespread clinical use,
however.
These studies suggest that regional wall-mo-
tion assessment by echocardiography to determine
early signs of ischemia in patients presenting to
the emergency room with chest pain is feasible.
For optimal sensitivity this approach requires
 EMERGENCY CARDIAC IMAGING
ongoing symptoms during the study. The sub-
optimal specificity suggests this technique has
limited use in decreasing the number of false-
positive admissions for patients presenting to the
emergency department with chest pain. Moreover,
no study has evaluated the actual impact of the
use of echocardiography on triage from the
emergency room.
Stress echocardiography has been shown to be
comparable with nuclear stress perfusion scanning
for detecting coronary disease and for predicting
short- and long-term cardiac events [63–65]. Exer-
cise stress testing and pharmacologic stress testing
seem to provide comparable short- and long-term
prognostic information. Dobutamine stress echo-
cardiography is equal to dipyridamole Technetium
sestamibi scanning in sensitivity and has greater
specificity for detecting single-vessel and multives-
sel disease [64]. An abnormal exercise stress echo-
cardiography or Thallium perfusion study
predicted a 4.1-fold and 4.9-fold increase, respec-
tively, in the risk of all cardiac events over an al-
most 4-year follow-up [65]. For even a longer
follow-up period (mean, 7.3 years), a normal do-
butamine stress echocardiogram predicted a car-
diac event rate of 3.6% per year, whereas
a normal dobutamine stress Technetium sestamibi
scan predicted a cardiac event rate of 2.8% per
year. Abnormal stress echocardiograms predicted
a cardiac event rate of 6.5%, whereas abnormal
Technetium sestamibi scans predicted a rate of
6.9% per year [63]. Therefore, exercise stress echo-
cardiography and pharmacologic stress echocardi-
ography provide similar detection and prognostic
information when compared with nuclear stress
studies.
Application of single-photon emission CT
myocardial perfusion imaging in the emergency
department
In patients who present in the emergency
departments with chest pain and are suspected
to be experiencing ACS, radionuclide myocardial
perfusion imaging techniques can provide both
diagnostic and prognostic information. Evidence
from controlled, randomized trials suggests that
incorporating SPECT myocardial perfusion im-
aging in emergency department patients who have
suspected ACS but no definitive ECG changes can
improve triage decisions. The ACC/AHA/Amer-
ican Society of Nuclear Cardiology Radionuclide
Imaging Guidelines classify myocardial perfusion
imaging in this setting as a class I, level A
61
indication [66] for patients in whom the diagnosis
is uncertain.
Among ACS patients who present with ST
segment elevation myocardial infarction or non-
ST segment elevation myocardial infarction/un-
stable angina, the typical role for imaging in the
stabilized patient is to provide risk-stratification
information to drive a management strategy
aimed at improving natural history. Thus, the
role of myocardial perfusion SPECT early during
ST segment elevation myocardial infarction or
non-ST segment elevation myocardial infarction/
unstable angina is to identify the location and
extent of myocardial injury. After therapeutic
intervention a follow-up study is compared with
the earlier study to identify the extent of myocar-
dial salvage and final infarct size (Fig. 6).
The role of myocardial perfusion imaging in
patients presenting with chest pain and
nondiagnostic ECG changes
In patients presenting with chest pain and non-
diagnostic ECG changes, myocardial perfusion
SPECT data have been shown to have an in-
cremental risk stratification value over clinical
data for predicting unfavorable cardiac events
[67]. The injection of Technetium-99m–based per-
fusion tracer in a patient during chest pain and im-
aging 45 to 60 minutes later allows the assessment
of myocardial blood flow at the time of injection.
In all observational studies, the negative predictive
value for ruling out myocardial infarction has
equaled or exceeded 99% in this setting. This find-
ing suggests that a normal myocardial perfusion
study in this setting portends very small risk of
myocardial infarction or ischemic event [68]. In
contrast, patients exhibiting abnormal regional
perfusion defect have a higher risk of cardiac events
during the index hospitalization as well as during
follow-up. One study by Kontos and colleagues
[69] found the sensitivity of SPECT sestamibi per-
formed in the emergency department to be 92%
for detecting acute myocardial infarction, whereas
initial troponin I values drawn at the same time had
a sensitivity of only 39%. The maximum troponin I
over the first 24 hours had sensitivity similar to rest
sestamibi imaging, but at a distinctly later time
point. Thus, acute myocardial perfusion imaging
has the potential to identify ACS earlier than bio-
markers, thereby providing assistance in patient
triage decisions (admit or discharge) in the emer-
gency department.
62 KUO et al

Fig. 6. Inferior myocardial perfusion defect in a patient with chest pain but no ischemic ECG abnormalities. (A) Vertical
long-axis resting SPECT myocardial perfusion images of a 67-year-old-man who presented to the emergency room with
chest pain and no ischemic ECG changes. His troponin T was negative and troponin I was less than 0.1. He was injected
with Technetium-99m sestamibi at rest in the emergency room and underwent SPECT imaging soon thereafter. The im-
ages show severely reduced inferior perfusion defect (arrows), which in the setting of ongoing symptoms was suggestive
of acute coronary syndrome. (B) Cardiac catheterization showed totally occluded right coronary artery and graft. (C) A
prior myocardial perfusion SPECT study performed showed normal perfusion in all myocardial regions, including the
inferior region (arrows).

Although these observational studies empha-
size the importance of myocardial perfusion
imaging for ruling out ACS, in none of those
studies were the imaging data allowed to affect
patient triage decisions in the emergency depart-
ment. In a prospective study by Stowers and
colleagues [70], 46 patients who had ongoing chest
pain and a nondiagnostic ECG were randomly as-
signed to an image-guided strategy (in which
patient management was based on the SPECT re-
sults) or a conventional strategy (in which imaging
results were kept blinded, and patient manage-
ment was independent of the SPECT data). The
results showed that an image-guided strategy in-
curred approximately 50% lower costs and re-
sulted in shorter lengths of hospital stay. In
a larger prospective trial (the ERASE Chest
Pain Trial) [71], 2475 patients who had symptoms
suggestive of ACS and a normal or nondiagnostic
ECG were randomly assigned to a usual emergen-
cy department evaluation strategy or a strategy in-
cluding acute rest SPECT myocardial perfusion
information. The results showed that the imaging
data were among the most powerful factors asso-
ciated with the appropriate decision to discharge
the patient from the emergency department. For
patients ultimately determined not to have ACS
as the presenting syndrome, SPECT myocardial
imaging was associated with a 32% reduction in
the odds of being admitted unnecessarily to the
hospital for treatment or observation [71]. On
30-day follow-up of all patients, there were no dif-
ferences in outcomes between the usual emergency
department evaluation strategy and SPECT
image-guidance. These findings suggest that the
incorporation of SPECT perfusion imaging into
the emergency department triage decision-making
process reduces unnecessary hospital admissions
without inappropriately reducing admission for
patients who have ACS. In the future, metabolic
imaging with a fatty acid tracer called methyl-
[123I]-iodophenyl-pentadecanoic acid (BMIPP)
may extend the time window for identifying myo-
cardial ischemia in the emergency room up to 30
hours after the cessation of chest pain [72].
The chest pain center protocol: stress myocardial
perfusion single-proton emission CT
Another strategy that has been proposed for
patients who have suspected ACS but a nondiag-
nostic ECG is serial evaluation of cardiac specific
enzymes over 6 to 24 hours, followed by stress
testing if the enzymes are negative. Among
patients who are considered clinically to be at
very low risk, however, stress myocardial perfu-
sion SPECT study can be performed rather early.
SPECT myocardial perfusion imaging in this
setting can potentially allow earlier patient triage
decisions than serial enzyme evaluation. The
current data suggest that if stress myocardial
perfusion studies are normal, the risk of ACS or
unfavorable cardiac events is low, and therefore
early discharge from the emergency department
may be considered. On the other hand, if the stress
 EMERGENCY CARDIAC IMAGING
imaging results are abnormal (ischemia or in-
farction), rapid admission and entry into an
appropriate evidence-based treatment pathway
for ACS are in order.
Summary
Multiple strategies and testing modalities are
available to evaluate patients presenting to the
emergency department with cardiac complaints.
Many provide anatomic and prognostic informa-
tion about coronary stenosis and long-term out-
comes. Although nuclear and stress echo imaging
have the ability to predict outcomes in patients in
the emergency department population, the newer
modalities of cardiac imaging (EBCT, MDCT,
and CMR) continue to show promising results
and may soon be incorporated into emergency
department chest pain centers. Protocols can be
developed within an institution to meet the needs
of the patient population while minimizing risk
and improving outcomes for all patients.
References
[1] Braunwald E, Antman EM, Beasley JW, et al. ACC/
AHA 2002 guideline update for the management of
patients with unstable angina and non-ST-segment
elevation myocardial infarctiondsummary article:
a report of the American College of Cardiology/
American Heart Association task force on practice
guidelines (Committee on the Management of
Patients With Unstable Angina). J Am Coll Cardiol
2002;40(7):1366–74.
[2] American Heart Association. 2005 heart disease and
stroke statisticsd2005 update. Dallas (TX): Ameri-
can Heart Association; 2005.
[3] Souza AS, Bream PR, Elliott LP. Chest film detec-
tion of coronary artery calcification. The value of
the CAC triangle. Radiology 1978;129:7–10.
[4] Margolis JR, Chen JT, Kong Y, et al. The diagnostic
and prognostic significance of coronary artery calci-
fication. A report of 800 cases. Radiology 1980;137:
609–16.
[5] MacGregor JH, Chen JT, Chiles C, et al. The radio-
graphic distinction between pericardial and myocar-
dial calcifications. AJR Am J Roentgenol 1987;148:
675–7.
[6] Higgins CB, Lipton MJ, Johnson AD, et al. False
aneurysms of the left ventricle. Identification of dis-
tinctive clinical, radiographic, and angiographic fea-
tures. Radiology 1978;127:21–7.
[7] Higgins CB, Lipton MJ. Radiography of acute myo-
cardial infarction. Radiol Clin North Am 1980;18:
359–68.
63
[8] Dao Q, Krishnaswamy P, Kazanegra R, et al. Utility
of B-type natriuretic peptide in the diagnosis of con-
gestive heart failure in an urgent-care setting. J Am
Coll Cardiol 2001;37(2):379–85.
[9] Agatston AS, Janowitz WR, Hildner FJ, et al. Quan-
tification of coronary artery calcium using ultrafast
computed tomography. J Am Coll Cardiol 1990;
15:827–32.
[10] Rich S, McLaughlin VV. Detection of subclinical
cardiovascular disease: the emerging role of electron
beam computed tomography. Prev Med 2002;34(1):
1–10.
[11] Rumberger JA, Sheedy PF II, Breen JF, et al. Elec-
tron beam computed tomography and coronary ar-
tery disease: scanning for coronary artery
calcification. Mayo Clin Proc 1996;71(4):369–77.
[12] O’Malley PG, Taylor AJ, Jackson JL, et al. Prognos-
tic value of coronary electron-beam computed to-
mography for coronary heart disease events in
asymptomatic populations. Am J Cardiol 2000;
85(8):945–8.
[13] Reddy GP, Chernoff DM, Adams JR, et al. Coronary
artery stenoses: assessment with contrast-enhanced
electron-beam CT and axial reconstructions. Radiol-
ogy 1998;208(1):167–72.
[14] Budoff MJ, Oudiz RJ, Zalace CP, et al. Intravenous
three-dimensional coronary angiography using con-
trast enhanced electron beam computed tomogra-
phy. Am J Cardiol 1999;83(6):840–5.
[15] Schmermund A, Rensing BJ, Sheedy PF, et al. Intrave-
nous electron-beam computed tomographic coronary
angiography for segmental analysis of coronary artery
stenoses. J Am Coll Cardiol 1998;31(7):1547–54.
[16] Achenbach S, Moshage W, Ropers D, et al. Value of
electron-beam computed tomography for the nonin-
vasive detection of high-grade coronary-artery ste-
noses and occlusions. N Engl J Med 1998;339(27):
1964–71.
[17] Lauden DA, Vukov LF, Breen JF, et al. Use of elec-
tron-beam computed tomography in the evaluation
of chest pain patients in the emergency department.
Ann Emerg Med 1999;33(1):15–21.
[18] Georgiou D, Budoff MJ, Kaufer E, et al. Screening
patients with chest pain in the emergency depart-
ment using electron beam tomography: a follow-up
study. J Am Coll Cardiol 2001;38:105–10.
[19] McLaughlin VV, Balogh T, Rich S. Utility of elec-
tron beam computed tomography to stratify patients
presenting to the emergency room with chest pain.
Am J Cardiol 1999;84:327–8.
[20] Shavelle DM, Budoff MJ, LaMont DH, et al. Exer-
cise testing and electron beam computed tomogra-
phy in the evaluation of coronary artery disease.
J Am Coll Cardiol 2000;36(1):32–8.
[21] Schmermund A, Baumgart D, Sack S, et al. Assess-
ment of coronary calcification by electron-beam
computed tomography in symptomatic patients
with normal, abnormal or equivocal exercise stress
test. Eur Heart J 2000;21(20):1674–82.
64 KUO
[22] O’Rourke RA, Brundage BH, Froelicher VF, et al.
American College of Cardiology/American Heart
Association Expert Consensus Document on elec-
tron-beam computed tomography for the diagnosis
and prognosis of coronary artery disease. J Am
Coll Cardiol 2000;36(1):326–40.
[23] Hoffmann MH, Shi H, Manzke R, et al. Noninvasive
coronary angiography with 16-detector row CT:
effect of heart rate. Radiology 2005;234(1):86–97.
[24] Nieman K, Cademartiri F, Lemos PA, et al. Reliable
noninvasive coronary angiography with fast submil-
limeter multislice spiral computed tomography.
Circulation 2002;106(16):2051–4.
[25] Ropers D, Baum U, Pohle K, et al. Detection of cor-
onary artery stenoses with thin-slice multi-detector
row spiral computed tomography and multiplanar
reconstruction. Circulation 2003;107(5):664–6.
[26] Morgan-Hughes GJ, Roobottom CA, Owens PE,
et al. Highly accurate coronary angiography with
submillimetre, 16 slice computed tomography.
Heart 2005;91:308–13.
[27] Silberman S, Dambrin G, Ghostine S, et al. [Diagno-
sis of acute myocardial infarction using multislice
computed tomography in emergency room.] Arch
Mal Coeur Vaiss 2004;97(4):366–9 [in French].
[28] Hastreiter D, Lewis D, Dubinsky TJ. Acute myocar-
dial infarction demonstrated by multidetector CT
scanning. Emerg Radiol 2004;11(2):104–6.
[29] Dirksen MS, Jukema JW, Bax JJ, et al. Cardiac
multidetector-row computed tomography in pa-
tients with unstable angina. Am J Cardiol 2005;
95(4):457–61.
[30] Dorgelo J, Willems TP, Geluk CA, et al. Multide-
tector computed tomography-guided treatment
strategy in patients with non-ST elevation acute cor-
onary syndromes: a pilot study. Eur Radiol 2005;
15(4):708–13.
[31] White CS, Kuo D, Keleman M, et al. Chest pain
evaluation in the emergency room: can multi-slice
CT provide a comprehensive evaluation? AJR Am
J Roentgenol, 2005;185(2):533–40.
[32] Wagner A, Mahrholdt H, Sechtem U, et al. MR im-
aging of myocardial perfusion and viability. Magn
Reson Imaging Clin N Am 2003;11(1):49–66.
[33] Regenfus M, Ropers D, Achenbach S, et al. Nonin-
vasive detection of coronary artery stenosis using
contrast-enhanced three-dimensional breath-hold
magnetic resonance coronary angiography. J Am
Coll Cardiol 2000;36(1):44–50.
[34] Kim WY, Danias PG, Stuber M, et al. Coronary
magnetic resonance angiography for the detection
of coronary stenoses. N Engl J Med 2001;345(26):
1863–9.
[35] van Geuns RJ, Oudkerk M, Rensing BJ, et al. Com-
parison of coronary imaging between magnetic
resonance imaging and electron beam computed
tomography. Am J Cardiol 2002;90(1):58–63.
[36] Kwong RY, Schussheim AE, Rekhraj S, et al.
Detecting acute coronary syndrome in the
et al
emergency department with cardiac magnetic reso-
nance imaging. Circulation 2003;107:531–7.
[37] Takahashi N, Inoue T, Oka T, et al. Diagnostic use
of T2-weighted inversion-recovery magnetic reso-
nance imaging in acute coronary syndromes com-
pared with 99mTc-Pyrophosphate, 123I-BMIPP
and 201TlCl single photon emission computed to-
mography. Circ J 2004;68(11):1023–9.
[38] Gottdiener JS. Overview of stress echocardiography:
uses, advantages, and limitations. Prog Cardiovasc
Dis 2001;43(4):315–34.
[39] McCully RB, Roger VL, Mahoney DW, et al. Out-
come after abnormal exercise echocardiography for
patients with good exercise capacity: prognostic
importance of the extent and severity of exercise-
related left ventricular dysfunction. J Am Coll
Cardiol 2002;39(8):1345–52.
[40] Elhendy A, Mahoney DW, Burger KN, et al. Prog-
nostic value of exercise echocardiography in patients
with classic angina pectoris. Am J Cardiol 2004;
94(5):559–63.
[41] Arruda-Olson AM, Juracan EM, Mahoney DW,
et al. Prognostic value of exercise echocardiography
in 5,798 patients: is there a gender difference? J Am
Coll Cardiol 2002;39(4):625–31.
[42] Marwick TH, Case C, Short L, et al. Prediction of
mortality in patients without angina: use of an exer-
cise score and exercise echocardiography. Eur Heart
J 2003;24(13):1223–30.
[43] Elhendy A, Mahoney DW, Khandheria BK, et al.
Prognostic significance of the location of wall mo-
tion abnormalities during exercise echocardiogra-
phy. J Am Coll Cardiol 2002;40(9):1623–9.
[44] Kamalesh M, Matorin R, Sawada S. Prognostic
value of a negative stress echocardiographic study
in diabetic patients. Am Heart J 2002;143(1):163–8.
[45] Sozzi FB, Elhendy A, Rizzello V, et al. Prognostic
value of dobutamine stress echocardiography in
patients with systemic hypertension and known or
suspected coronary artery disease. Am J Cardiol
2004;94(6):733–9.
[46] Chuah SC, Pellikka PA, Roger VL, et al. Role of
dobutamine stress echocardiography in predicting
outcome in 860 patients with known or suspected
coronary artery disease. Circulation 1998;97(15):
1474–80.
[47] Amici E, Cortigiani L, Coletta C, et al. Usefulness of
pharmacologic stress echocardiography for the long-
term prognostic assessment of patients with typical
versus atypical chest pain. Am J Cardiol 2003;
91(4):440–2.
[48] Steinberg EH, Madmon L, Patel CP, et al. Long-
term prognostic significance of dobutamine echocar-
diography in patients with suspected coronary artery
disease: results of a 5-year follow-up study. J Am
Coll Cardiol 1997;29(5):969–73.
[49] Marwick TH, Case C, Sawada S, et al. Prediction
of mortality using dobutamine echocardiography.
J Am Coll Cardiol 2001;37(3):754–60.
 EMERGENCY CARDIAC IMAGING
[50] Bholasingh R, Cornel JH, Kamp O, et al. Prognostic
value of predischarge dobutamine stress echocardi-
ography in chest pain patients with a negative cardiac
troponin T. J Am Coll Cardiol 2003;41(4):596–602.
[51] Biagini E, Elhendy A, Bax JJ, et al. Seven-year fol-
low-up after dobutamine stress echocardiography:
impact of gender on prognosis. J Am Coll Cardiol
2005;45(1):93–7.
[52] Yao SS, Qureshi E, Sherrid MV, et al. Practical
applications in stress echocardiography: risk stratifi-
cation and prognosis in patients with known or sus-
pected ischemic heart disease. J Am Coll Cardiol
2003;42(6):1084–90.
[53] Chung G, Krishnamani R, Senior R. Prognostic
value of normal stress echocardiogram in patients
with suspected coronary artery disease–a British
general hospital experience. Int J Cardiol 2004;
94(2–3):181–6.
[54] Peels CH, Visser CA, Kupper AJF, et al. Usefulness
of two-dimensional echocardiography for immedi-
ate detection of myocardial ischemia in the emer-
gency room. Am J Cardiol 1990;65:687–91.
[55] Sasaki H, Charuzi Y, Beeder C, et al. Utility of echo-
cardiography for the early assessment of patients
with non-diagnostic chest pain. Am Heart J 1986;
112:494–7.
[56] Sabia P, Afrookteh A, Touchstone D, et al. Value of
regional wall motion abnormality in the emergency
room diagnosis of acute myocardial infarction.
Circulation 1991;84:I-85–92.
[57] Villanueva FS, Sabia PJ, Afrookteh A, et al. Value
and limitations of current methods of evaluating
patients presenting to the emergency room with car-
diac-related symptoms for determining long-term
prognosis. Am J Cardiol 1992;69:746–50.
[58] Loh IK, Charuzi Y, Beeder C, et al. Early diagnosis
of non-transmural myocardial infarction by two-di-
mensional echocardiography. Am Heart J 1982;104:
963–8.
[59] Kontos MC, Arrowood JA, Paulsen WHJ, et al.
Early echocardiography can predict cardiac events
in emergency department patients with chest pain.
Ann Emerg Med 1998;31:550–7.
[60] Kontos MC, Arrowood JA, Jesse RL, et al. Com-
parison between 2-dimensional echocardiography
and myocardial perfusion imaging in the emergency
department in patients with possible myocardial is-
chemia. Am Heart J 1998;136:724–33.
[61] Trippi JA, Lee KS, Kopp G, et al. Dobutamine
stress tele-echocardiography for evaluation of emer-
gency department patients with chest pain. J Am
Coll Cardiol 1997;30:627–32.
[62] Kaul S, Senior R, Firschke C, et al. Incremental
value of cardiac imaging in patients presenting to
the emergency department with chest pain and with-
out ST-segment elevation: a multicenter study. Am
Heart J 2004;148:129–36.
65
[63] Schinkel AF, Bax JJ, Elhendy A, et al. Long-term
prognostic value of dobutamine stress echocardiog-
raphy compared with myocardial perfusion scan-
ning in patients unable to perform exercise tests.
Am J Med 2004;117(1):1–9.
[64] Smart SC, Bhatia A, Hellman R, et al. Dobutamine-
atropine stress echocardiography and dipyridamole
sestamibi scintigraphy for the detection of coronary
artery disease: limitations and concordance. J Am
Coll Cardiol 2000;36(4):1265–73.
[65] Olmos LI, Dakik H, Gordon R, et al. Long-term
prognostic value of exercise echocardiography
compared with exercise 201Tl, ECG, and clinical
variables in patients evaluated for coronary artery
disease. Circulation 1998;98(24):2679–86.
[66] Klocke FJ, Baird MG, Lorell BH, et al. ACC/
AHA/ASNC guidelines for the clinical use of
cardiac radionuclide imaging-executive summary:
a report of the American College of Cardiology/
American Heart Association Task Force on Prac-
tice Guidelines (ACC/AHH/ASNC Committee to
revise the 1995 Guidelines for the clinical use of
cardiac radionuclide imaging). Circulation;108(11):
1404–18. 2003.
[67] Heller GV, Stowers SA, Hendel RC, et al. Clinical
value of acute rest technetium-99m tetrofosmin to-
mographic myocardial perfusion imaging in patients
with acute chest pain and non diagnostic electrocar-
diogram. J Am Coll Cardiol 1998;31:1011.
[68] Wackers FJ, Brown KA, Heller GV, et al. American
Society of Nuclear Cardiology position statement
on radionuclide imaging in patients with suspected
acute ischemic syndrome in the emergency depart-
ment or chest pain center. J Nucl Cardiol 2002;9:
246.
[69] Kontos MC, Jesse RL, Anderson P, et al. Compar-
ison of myocardial perfusion imaging and cardiac
troponin I in patients admitted to the emergency
department with chest pain. Circulation 1999;99:
2073.
[70] Stowers SA, Eisenstein EL, Wackers FJ, et al. An
economic analysis of an aggressive diagnostic stra-
tegy with single photon emission computed tomo-
graphy myocardial perfusion imaging and early
exercise stress testing in emergency department
patients who present with chest pain but nondiag-
nostic electrocardiograms. Ann Emerg Med 2000;
35:17.
[71] Udelson JE, Beshansky JR, Ballin DS, et al. Myocar-
dial perfusion imaging for evaluation and triage of
patients with suspected acute cardiac ischemia: a ran-
domized controlled trial. JAMA 2002;288:2693.
[72] Dilsizian V, Bateman TM, Bergmann SR, et al.
Metabolic imaging with methyl-[123I]-iodophenyl-
pentadecanoic acid (BMIPP) identifies ischemic
memory following demand ischemia. Circulation
2005;112(14):2169–74.
 Cardiol Clin 24 (2006) 67–78
Cardiac Risk Assessment: Matching Intensity
of Therapy to Risk
Mark R. Vesely, MD, Mark D. Kelemen, MD, MSc, FACC*
Division of Cardiology, University of Maryland School of Medicine, 22 South Greene Street,
Baltimore, MD 21202, USA
The concept of cardiac risk assessment is appli-
cable to many settings, from primary prevention
and preoperative evaluation to post–myocardial
infarction (MI) management. In the context of
emergency cardiac care, cardiac risk assessment
is typically focused on the evaluation of acute
coronary syndrome (ACS): ST-segment elevation
MI (STEMI), non–ST-segment elevation MI
(NSTEMI), and unstable angina (UA). The likeli-
hood of cardiac death and significant morbidity
guides management decisions, including the need
for further diagnostic evaluation, specific treat-
ment, and the degree of monitoring for complica-
tions. Risk scores of varying complexity are
available to guide early management. This article
briefly describes the clinical spectrum of NSTE
ACS, which encompasses NSTEMI and UA, and
focuses on the use of scoring systems in tandem
with clinical guidelines to determine the intensity
and timing of early therapy. Discussion then
focuses on the process of developing these tools
and potential future developments in risk
assessment.
The rationale of a risk stratification system
The goal of risk assessment is to predict the
likelihood of occurrence of a clinically significant
outcome, given a complex initial presentation. Risk
models can be used in stratification within clinical
trials, in quality of care evaluation based on
expected outcomes, and, as described here, in
medical decision making [1]. Once a risk model is
* Corresponding author.
E-mail address: mkelemen@medicine.umaryland.edu
(M.D. Kelemen).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.010
developed and validated, the frequency of clinical
outcomes (beneficial and adverse) following a spe-
cific therapy can be determined within the risk lev-
els. The high risk of morbidity and mortality from
ACS must be balanced with the degree of benefit
and risk of adverse events inherent in the various
therapeutic options available for ACS. For
instance, antiplatelet therapy with a platelet
glycoprotein (Gp) IIb/IIIa receptor antagonist
improves outcome in many patients who have
ACS but is associated with increased rates of
thrombocytopenia and bleeding. A risk stratifica-
tion system (RSS), through a standardized assess-
ment scheme, allows a concise and simplified
method for characterizing this risk. Once derived
and validated, risk models can be tested prospec-
tively to assess their predictive capacity. If effective,
they can help the clinician make the risk–benefit
calculation. Optimal risk stratification and delivery
of care remains a moving target, however. The un-
derstanding of the pathophysiology of ACS and
available treatment options constantly evolve. As
additional risk factors and therapies arise, an
RSS can become outdated because these new fea-
tures are not included in the model. Therefore,
the optimal use of an RSS depends on an un-
derstanding of the strengths and weaknesses of
the RSS, including the setting in which it was
developed.
Mortality from acute ischemic heart disease has
fallen considerably during the past 2 decades
because of advances in prevention, diagnosis, and
treatment. Despite well-established consensus
guidelines, however, many patients do not receive
recommended treatments or are given medications
that may actually increase risk. Observational
studies from the early 1990s (Thrombolysis In
cardiology.theclinics.com
68 VESELY & KELEMEN
Myocardial Infarction [TIMI] 3 registry) [2] to
2002 (Can Rapid Risk Stratification of Unstable
Angina Patients Suppress Adverse Outcomes
with Early Implementation of the ACC/AHA
Guidelines [CRUSADE]) [3] confirm the increased
use of aspirin (from 82% to 90%) and beta-block-
ers (from 45% to 76%) but demonstrate low rates
of use of heparin (53%) and, recently, Gp IIb/IIIa
inhibitors (31%).
RSSs may improve compliance with the na-
tional guidelines, however, they have the potential
for being very complicated. The measurement of
the biomarkers including troponin, C-reactive
peptide, and B-type natriuretic peptide (BNP),
each give independent and additive prognostic
information [4]. Table 1 provides a more complete
list of variables that have been associated with in-
creased mortality in NSTE ACS. A scoring system
that incorporates all of them would be time-con-
suming; and time-to-treatment itself is an impor-
tant risk factor. While some variables can be
determined quickly, others require improvements
in the speed of laboratory testing or in computer
processing. Furthermore, certain predictors are
strong and consistently associated with risk,
such as the peak level of cardiac troponin I (or
T) [2]. The peak level is not known for several
days in NSTE ACS, however, so therapy cannot
be guided in the first few hours of presentation.
The consensus guidelines for NSTE ACS in-
troduce the concept of matching intensity of
therapy to risk identified during a hospitalization,
but no randomized controlled trials have focused
on a strategy based entirely on risk scoring.
According to the class I recommendations from
the guidelines, high-risk and low-risk patients have
clearly identified pathways with different treatment
strategies. Patients at intermediate risk typically
30% to 40% of all patients, probably account for
much of the variability in treatment prescription.
Interaction of consensus guidelines and quality
improvement studies
An individual’s risk during an episode of ACS
is determined by factors unique to that patient
and factors related to the pathophysiology of
ACS. Patient factors include age, known prior
ischemic heart disease, and evidence of heart
failure on presentation. Angina pectoris can be
classified as unstable in three scenarios: rest
angina (angina occurring at rest and continuing
for longer than 20 minutes), new-onset angina
(new onset with slight limitation of ordinary
activity), and increasing angina (previously pres-
ent but now more frequent, longer in duration,
and with decreased exertional threshold) [14]. The
Canadian Cardiovascular Society [15] classifica-
tion system differentiates symptoms into four
grades based on exertional tolerance, but descrip-
tive classification of angina pectoris based on
severity of symptoms has been found to be
unreliable in providing prognostic information
[15,16].
There are five mechanisms underlying myocar-
dial ischemia in NSTE ACS, resulting in dispro-
portionate supply and demand of myocardial
oxygen [17]. Decreased myocardial perfusion and
oxygen delivery are common to four mechanisms;
nonocclusive thrombus and microembolization
are the most common. Dynamic obstruction
from epicardial vasospasm, progressive severe
narrowing without spasm, and localized inflam-
mation also decrease oxygen delivery. Secondary
UA is caused by either decreased supply (anemia
and hypoxemia) or increased demand (fever and
hypotension), typically in the setting of underlying
CAD. These mechanisms are nonexclusive and do
not offer reliable prognostic information within
the initial evaluation period.
When facing an ACS presentation, the clini-
cian’s first critical decision point is excluding
STEMI. An ECG should be obtained within the
first 10 minutes of arrival (or sooner if possible).
The first set of biomarkers should be available
within 60 minutes [18]. This inherent delay results
in reduced efficiency in identifying and treating
high-risk patients; therefore risk assessment tools
were developed. Aspirin, heparin (unfractionated
or low molecular weight) and beta-blockers are
given to all patients who do not have contraindi-
cations, and Gp IIb/IIIa inhibitors are used in
high-risk patients for whom early cardiac catheter-
ization is also recommended. Although patients
who have ACS are, as a group, at increased risk
of death and nonfatal ischemic complications,
the spectrum of disease severity and subsequent
outcomes remains wide. Initial evaluation should
focus on immediate management of hemodynamic
instability and identifying patients at significant
risk for death, recurrent MI, stroke, heart failure,
and recurrent ischemic symptoms. Because the in-
tensity of therapy typically depends on this prog-
nostic information, risk assessment facilitates
further decisions on the type of antithrombin
and antiplatelet therapy, the indications for and
timing of invasive angiographic evaluation, and
the appropriate level of further monitoring
Table 1
Variables associated with increased mortality in acute coronary syndrome
TIMI [5] PURSUIT [6] GRACE [7] Other
Demographics
Age R65 y þ . . .
Latin American race . þ . .
Female gender . . þ .
Medical history
R3 CAD risk factors* þ . . .
Diabetes mellitus * þ þ .
Hypertension * . þ .
Hyperlipidemia * þ þ .
Current smoking * þ - .
Family history of CAD * þ . .
Congestive Heart Failure . þ þ .
Stroke . þ þ .
Peripheral vascular disease . þ þ .
Renal dysfunction . . þ þ [8]
Angina . . - .
Atrial fibrillation . . þ .
Prior (þ) exercise stress test . . þ .
Bleeding . . þ .
CAD history
Coronary stenosis R50% þ . . .
Prior angioplasty þ þ - .
Coronary artery bypass . þ þ .
Myocardial infarction . þ . .
Presentation
Cardiac arrest . . þ .
Severe angina þ þ . .
Enrolled as MI, not UA . þ . .
Rales (R1/3 of lung fields) . þ . .
ECG changes
ST deviation þ . þ .
ST depression . þ þ .
Anterior ST depression . . þ .
Inferior ST depression . . þ .
T-wave inversion . þ - .
Significant Q-wave . . þ .
Left bundle branch block . . þ .
Right bundle branch block . . þ .
Prior medication use
Aspirin þ þ - .
Statin . . - .
b-blocker . þ . .
Calcium channel blocker . þ . .
Nitrates . þ . .
ACE inhibitor . þ . .
Biomarkers
Troponin I (or T) þ . þ þ [9]
CK-MB þ . þ .
Myoglobin . . . þ [10]
NT-proBNP . . . þ [9]
C-reactive protein . . . þ [11]
Hemoglobin . . . þ [12]
White blood cell count . . . þ [13]
Abbreviations: ACE, angiotensin-converting enzyme; CAD, coronary artery disease; CK-MD, MB isoenzyme of cre-
atine kinase; GRACE, Global Registry of Acute Coronary Events; NT-proBNP, N-terminal probrain natriuretic pep-
tide; PURSUIT, Platelet Glycoprotein IIb/IIIa in Unstable Angina: Receptor Suppression Using Integrilin Therapy
Trail.
70 VESELY & KELEMEN
(ranging from evaluation in the chest pain unit
and outpatient follow-up to critical care
admission).
What makes a risk stratification system useful?
As a clinical tool, the usefulness of an RSS
depends on its accuracy (ie, predictive capacity)
and applicability (ie, ease of use). The number and
type of variables used in the system, the endpoints
for which the system is meant to predict, and the
populations in which a system is tested and to
which a system is applied are important for the
success of an RSS. Ideally, an RSS can be used at
the bedside to provide highly accurate and rele-
vant prognostic information within a short time
frame. The clinical variables should be informa-
tive and obtained easily and quickly. Intuitively,
the more variables there are within a risk strati-
fication model, the more informative it is likely to
be. As knowledge of the pathophysiology of ACS
continues to expand, so do the number and
complexity of the potential variables available
for use within an RSS. As additional factors are
used to provide more predictive capacity, each
additional variable can add potential time and
difficulty for use, making the system more cum-
bersome and thus less applicable.
A typical RSS design is based on a point
system, in which points are assigned for the
presence (or absence) of specific clinical factors.
The sum of these points correlates to a specific
risk level. An RSS with absolute discriminatory
capacity would account for all of the variables
influencing a particular outcome. If such a system
were composed of five independent risk factors
(A, B, C, D, and E) of equal impact to the overall
risk, each factor would account for 20% of the
overall risk burden. If, however, factor A portrays
twice the level of risk as the others A would
represent 33.3% of the overall risk burden, and
each of the remaining four factors would repre-
sent 16.6% of the overall risk. This variability in
attributable risk must be accounted for in risk
stratification to maximize accuracy. In point-
system risk models, accuracy is maximized by
weighing each factor by the percentage of its
contribution to total risk of the outcome (eg,
factor A, 2 points; factors B, C, D, and E, 1 point
each) [1]. As additional, appropriately weighted
factors are added to a system for improved accu-
racy, the system’s complexity expands and can po-
tentially become a hindrance to its quick and
appropriate use.
ACS has many potential adverse outcomes,
including need for urgent revascularization, re-
current MI, and death. To provide clinically
useful prognostic information, an RSS may be
developed with these relevant outcomes combined
into a composite endpoint, with the perspective
that the risk of any one of the outcomes is
pertinent to optimizing management. A composite
endpoint such as a major adverse cardiac event
(MACE), is easier to use clinically, in that
a therapeutic option can be provided if the risk
of MACE (any one of the many outcomes in-
cluded in MACE) is not too high. If a health care
provider is attempting to avoid any and all of
these outcomes, combining them in one RSS is
simpler than having an RSS for each of the
individual outcomes. Specific factors within an
RSS, however, may be predictive only of certain
events within a composite endpoint (ie, urgent
revascularization but not death). If the practitioner
is primarily concerned about mortality, a scale
with an accurate predictor of risk of death should
be chosen.
Finally, the practitioner using the RSS should
be aware of the patient population in which the
scale was derived. An RSS is best developed in
a diverse patient population to allow applicability
to a similarly wide population. If an RSS is
developed in a selective population, significant
factors may exist within that population that are
not accounted for but that uniformly add risk to
the general population. If those factors do not
exist in a subsequent patient, the predicted risk
may be less accurate. For example, if the deriva-
tion cohort of an RSS for ACS outcomes contains
a high prevalence of smoking, then smoking may
not add predictive capacity to the RSS and may
not be included in the final model. If that RSS is
then applied to a nonsmoking patient, the RSS
may suggest higher risk than the patient really
has. The application of an RSS to a patient not
similar to the derivation cohort or subsequently
verified population may thus provide inaccurate
information. The responsibility for the appropri-
ate use of an RSS thus falls on the user, who must
maintain this perspective when relying on an RSS
to assist in medical decision making.
Anatomy of a risk stratification system: the
Thrombolysis In Myocardial Infarction risk score
for unstable angina/non–ST-segment elevation
myocardial infarction
Among the types of RSS available for clinical
use, models developed through multivariable
 INTENSITY OF THERAPY
regression are frequently employed [19]. Multiple
clinical characteristics are first identified as as-
sociated with the likelihood for the event. These
characteristics are then compiled into a com-
posite score that functions as a mathematical rep-
resentation of the probability of the clinical event’s
occurrence [1].
As an example of an RSS, the development of
the TIMI risk score (TIMI RSS) for UA/NSTEMI
is discussed. The purpose of the TIMI RSS is to
assist in the identification of the patients at highest
risk for further adverse events such as death and
recurrent ischemia or infarction. Given the vari-
ability of potential adverse outcomes, the primary
endpoint (within follow up for 14 days) was chosen
as a composite of all-cause mortality, new or
recurrent MI, or recurrent ischemia leading to
urgent revascularization. The creation of the TIMI
RSS included two stages of production, the initial
development of the system using one cohort of
patients and a subsequent validation of the system
with a different set of patients. The initial patient
population used for the development of the
system, defined as the ‘‘derivation cohort,’’ con-
sisted of the 1957 patients in the TIMI 11B trial
who were randomly assigned to receive standard
treatment including aspirin and unfractionated
heparin (UFH) [20]. Study subjects were enrolled
between August 1996 and March 1998, with an ad-
justment in inclusion criteria after the first 10
months. Initial parameters for enrollment included
the presence of ischemic discomfort at rest for at
least 5 minutes and additional evidence of ischemic
heart disease including history of CAD, ST-seg-
ment deviation, or elevated serum cardiac markers.
After 10 months of enrollment, the criteria were
adjusted to focus on higher-risk patients who
have evidence of ischemic heart disease and
was limited to only those with ST-segment devi-
ation or elevated serum cardiac markers. The
study subjects had a mean age of 66 years,
and 64% were male. CAD risk factors within
the patient population included family history
(34%), hypertension (50%), hypercholesterol-
emia (32%), diabetes mellitus (20%), and cur-
rent smoking (27%). Prior cardiac history was
also present in many patients, such as previous
MI (32%), history of coronary artery bypass graft
surgery (CABG) (13%), and history of percutane-
ous transluminal coronary angioplasty (12%).
ECG abnormality, including ST-segment devia-
tions or T-wave inversion, was present in 83%
of the patients, and elevated serum cardiac
markers were present at enrollment in 40% of
71
the patients. The majority of patients were ulti-
mately diagnosed as having UA (58%), but
34.5% of the patients had a non–Q-wave MI,
and 3.8% had a Q-wave MI. Through the full
14-day follow-up period, 16.7% of the patients
experienced the composite endpoint.
The choice of potential predictor variables was
completed with the goal of producing an easily
applicable stratification system with useful pre-
dictive capacity. The number of predictors for the
model was thus critical, because too few variables
might limit predictive capacity, but too many
variables would make the system cumbersome
and hamper clinical utility. Thus the potential
predictor variables selected were those that were
easily identifiable as a baseline characteristic and
previously demonstrated to be predictive of out-
come. Twelve variables were chosen as potential
contributors to the scoring system:
1. Age greater than 65 years
2. Presence of more than three CAD risk fac-
tors (including family history of CAD, hy-
pertension, hypercholesterolemia, diabetes,
current smoking)
3. Prior coronary stenosis greater than 50%
4. Prior MI
5. Prior CABG
6. Prior percutaneous transluminal coronary
angioplasty
7. ST-segment deviation greater than 0.5 mV
8. Severe anginal symptoms (two anginal
events in the prior 24 hours)
9. Use of aspirin in last 7 days,
10. Use of intravenous UFH within 24 hours of
enrollment
11. Elevated serum cardiac markers (creatine
kinase myocardial band or troponin)
12. Prior history of congestive heart failure
These 12 variables were then analyzed by
univariate logistic regression, and the variables
with a significant correlation (P ! .20) were sub-
jected to multivariate logistic regression analysis.
The seven variables found to correlate signifi-
cantly with the endpoint were chosen subsequently
for the risk stratification model (Fig. 1) [21].
Two additional statistical tests, the C-statistic
and the Hosmer-Lemeshow statistic, were applied
to the final variable set to assess the model’s
performance in terms of discrimination and cali-
bration. The C-statistic measures a stratification
model’s predictive capacity by assessing discrim-
inative ability to classify an individual to the
72 VESELY & KELEMEN

Inclusion Criteria Outcomes:**Rate (%) at 14 Days

5 min of ischemic discomfort at rest & Point D or
D or
history of CAD or Total D MI UR nMI
nMI or UR
ST deviation or
elevated CK-MB or troponin 0&1 1 2 3 1 5
2 1 2 3 6 8
Exclusion Criteria 3 2 4 5 10 13
planned revascularization within 24 hrs 4 3 5 7 12 20
or correctable cause of angina 5 6 9 12 14 26
or anticoagulation contraindication 6&7 7 16 19 21 41

CLINICAL FEATURES (1 point assigned for each)

Historical At Presentation
age 65 years severe angina (twice within 24 hrs)
aspirin use within 7 days elevated CK-MB or troponin
coronary artery stenosis 50% ST deviation 0.05 mV
3 CAD risk factors*

Fig. 1. The TIMI risk score for UA/NSTEMI was derived from the TIMI 11B patient cohort. Multivariate logistic re-
gression identified seven clinical features, which provide equally weighted prognostic information. One point is assigned
for each present feature, and points are summed to derive the patient’s TIMI risk score. The risk scores of 0 and 1 and
the risk scores of 6 and 7 are combined to derive six TIMI risk levels. Rates of individual events and composite outcomes
were determined from the entire TIMI 11B patient population, including both the unfractionated heparin and enoxaparin
arms. Likelihood of death, MI, or urgent revascularization increases with each risk level (P ! .001). Risk factors for
coronary artery disease are a family history of coronary artery disease, hypertension, hypercholesterolemia, diabetes mel-
litus, active smoking. D, all-cause mortality; MI, myocardial infarction; nMI, nonfatal myocardial infarction; UR, ur-
gent revitalization. (From Antman EM, Cohen M, Bernink PJ, et al. The TIMI risk score for unstable angina/non-ST
elevation MI: a method for prognostication and therapeutic decision making. JAMA 2000;284:835.)

appropriate risk level. The Hosmer-Lemeshow
statistic assesses a model’s ‘‘goodness of fit’’ or
calibration by comparing rates of actual events
with predicted events within each risk group.
The C-statistic and Hosmer-Lemeshow statistic
showed the TIMI RSS to be well balanced for
discriminatory capacity and calibration (Hosmer-
Lemeshow statistic: 3.56, 8 degrees of freedom,
P ¼ 0.89; C-statistic: 0.65 with 0.83 the best bal-
ance of discrimination and calibration) [21].
Although a multivariate logistic regression
model could provide a weighted score for each
variable to be used in predicting a risk score, it
would require complex calculations probably re-
quiring computer assistance. Given the desire to
produce an easily applicable bedside tool, each of
the seven significant variables was weighted equally
because the seven variables had similar prognostic
significance (odds ratio). Thus the TIMI RSS was
produced as a summed score of one point for each
of the seven variables, making eight possible scores,
ranging from 0 to 7. The summed scores of 0 and 1
were combined, as were those of 6 and 7, to make
a final total of six levels of risk (Fig. 1).
When examining the clinical applicability of
the TIMI RSS, several favorable characteristics
were identified, each adding to the overall utility
of the model. The six risk levels were found to
have a normal distribution within the derivation
cohort. In addition, each subsequent risk level
correlated with significantly increased rates of
clinical endpoints, producing a wide range of
risk across the six levels. The risk of experiencing
death (from any cause), new or recurrent MI, or
urgent revascularization secondary to recurrent
ischemia within 14 days of presentation ranged
from 4.7% (at the 0–1 risk level) to 40.9% (at the
6–7 risk level.) Finally, the trend of 14-day risk for
each of the individual components of the com-
posite endpoint increased in parallel with the
composite endpoint.
Once set in its final form, the risk stratification
model was retrospectively applied to three
additional cohorts for validation, including the
patients receiving enoxaparin in TIMI 11B
(n ¼ 1953) [11] and both the UFH (n ¼ 1564)
and enoxaparin (n ¼ 1564) arms of the Efficacy
and Safety of Subcutaneous Enoxaparin in Non-
Q-wave Coronary Events (ESSENCE) trial [22].
Both discriminatory capacity and calibration
were maintained within these additional patient
cohorts.
 INTENSITY OF THERAPY 73

Application of the Thrombolysis In Myocardial within each TRS, enoxaparin was shown to provide
Infarction risk score for unstable angina/ greater benefit than UFH to patients who had
non–ST-segment elevation myocardial infarction a TRS of 4 or higher. A superior outcome in the
high-risk groups with enoxaparin is conceptually
The purpose of identifying high-risk patients is
plausible, because bolus subcutaneous dosing
to alter management to mitigate risk. The TIMI
may reach therapeutic levels faster than the slowly
RSS was assessed for the ability to predict which
titrated UFH continuous intravenous infusion.
patients gained benefit from the choice of enox-
Unlike UFH, however, the degree of antico-
aparin over UFH. Patients within the TIMI 11B
agulation with enoxaparin is not monitored. If the
and ESSENCE trials were stratified by their TIMI
presence of low molecular weight heparin delays
RSS to test retrospectively for improved outcome
a needed cardiac catheterization and revasculari-
with enoxaparin (Fig. 2). When assessing patients
zation, the benefit for high-risk patients may not

(0-3) (4-7)
PRISM-PLUS
RR 1.3 RR 0.75
p = 0.2 p = 0.01

(0-2) (3-4) (5-7) TIMI 11B

RR 0.9 RR 0.8 RR 0.49 ESSENCE
p=0.5 p=0.016 P<0.0001
95% CI: 0.50 – 1.84 95% CI: 0.50 – 0.76 95% CI: 0.35 – 0.68
45

40

35 LMWH UFH (b)

14 Day Event Rate (%)

30 UFH (a) IIB/IIIA

25 *
20 NS

15

10

0
0-2 3 4 5 6-7
TIMI RISK SCORE for UA/NSTEMI

Fig. 2. Clinical outcomes as a function of TIMI risk scores. Efficacy of strategies for antithrombotic therapy in patients
with NSTE ACS following stratification with the TIMI risk score (TRS) for UA/NSTEMI. All strategies included as-
pirin and determined composite endpoint event rates at 14 days. LMWH, patients receiving enoxaparin in the TIMI 11B
and ESSENCE trials; UFH (a), patients receiving unfractionated heparin in the TIMI 11B and ESSENCE trials; UFH
(b), patients receiving unfractionated heparin in the PRISM-PLUS trial; IIb/IIIa, patients receiving unfractionated
heparin and the glycoprotein IIb/IIIa receptor antagonist tirofiban in the PRISM-PLUS trial. Compared with UFH
alone, patients receive greater benefit with UFH and tirofiban if their TRS is 4 or higher (TRS 0–3: relative risk [RR],
1.3; P ¼ .2; TRS 4–7: RR, 0.75; P ¼ .01). When risk scores are further grouped into low-risk (TRS 0–2), intermediate-
risk (TRS 3–4), and high-risk (TRS 5–7) groups, patients with intermediate or high risk have greater benefit with enoxaparin
(TRS 0–2: RR, 0.91; 95% CI, 0.67–1.22; P ¼ .5; TRS 3–4: RR, 0.8; 95% CI 0.67–0.96; P ¼ .016; TRS 5–7: RR, 0.49; 95%
CI 0.35–0.68; P ! 0.0001). In analysis of clinical benefit within individual TIMI risk scores, however, there is greater
benefit with enoxaparin than with UFH if the TRS is 4 or higher (TRS 3: RR, 0.91; 95% CI, 0.71–1.16; P ¼ .43;
TRS 4: RR, 0.71; 95% CI, 0.54–0.93; P ¼ .009). (Data from Antman EM, Cohen M, Bernink PJ, et al. The TIMI
risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making.
JAMA 2000;284:835; Morrow DA, Antman EM, Snapinn SM, et al. An integrated clinical approach to predicting
the benefit of tirofiban in non-ST elevation acute coronary syndromes. Application of the TIMI Risk Score for UA/
NSTEMI in PRISM-PLUS. Eur Heart J 2002;23:223.)
74 VESELY & KELEMEN
be fully realized. Since the publication of the con-
sensus guidelines, no randomized clinical trial has
prospectively assessed the relative benefit of enox-
aparin over UFH in high-risk patients. On the
other end of the risk spectrum, therapeutic antico-
agulation with heparin (UFH or enoxaparin) can
increase the risk of hemorrhage, stroke, and
thrombocytopenia [22,23]. No randomized clini-
cal trial has examined the benefit of heparin
(UFH or enoxaparin) anticoagulation in low-
risk NSTE ACS.
In addition to medical management of NSTE
ACS with antianginal and antithrombotic ther-
apy, the treating clinician must decide whether
invasive cardiac catheterization is indicated. Two
general strategies for approaching cardiac cathe-
terization are generally used: an invasive ap-
proach in which patients routinely undergo
catheterization within 48 hours of admission, or
a conservative approach in which catheterization
is completed if recurrent ischemia occurs sponta-
neously or is provoked through noninvasive
testing. Between December 1997 and December
1999, 2220 patients were enrolled into the Treat
Angina with Aggrastat and Determine Cost of
Therapy with an Invasive or Conservative Strat-
egy (TACTICS)-TIMI 18 trial and were randomly
assigned to one of these two strategies [24]. Med-
ical antithrombotic therapy consisted of aspirin,
UFH, and tirofiban. Over a 6-month follow-up
period, patients who underwent routine early
catheterization experienced significantly less
recurrent ischemia but suffered higher rates of
protocol-defined bleeding (5.5% versus 3.3%;
P ! .01). Stratification with the TIMI RSS
demonstrated patients who have intermediate
risk (3–4) or high risk (5–7) benefited from the
routine early catheterization. Patients at low risk
(0–2) did not obtain greater benefit from one
strategy over the other (Fig. 3). The benefit of rou-
tine early catheterization in the intermediate-risk
group of patients was barely significant (P ¼
.048; upper bound of 95% CI, .999). Whether
analysis of patient groups within individual scores
(3 versus 4; Fig. 2) or redefining the bounds of the
intermediate risk group would enhance predictive
capacity remains unclear. Current consensus
guideline class I recommendations call for an early
invasive strategy if any of the following high-risk
indicators are present:
1. Recurrent angina/ischemia with rest or low-
level exertion in setting of intense anti-ische-
mic therapy
45 UFH, IIb/IIIa + CON UFH, IIb/IIIa + INV
6 Month Event Rate (%)
40
*
35
30
*
25 NS
20
15
10
5
0
0-2 3-4 5-7
TIMI RISK SCORE for UA/NSTEMI
Fig. 3. Assessment of cardiac catheterization strategies
in addition to aspirin, unfractionated heparin (UFH),
and tirofiban (IIb/IIIa) in the TACTICS-TIMI 18 trial.
Rates of a composite endpoint (death, nonfatal MI, or
rehospitalization for ACS) at 6 months were compared
between conservative (CON: selective catheterization)
and invasive (INV: routine catheterization) strategies.
Risk levels were stratified as low (TIMI risk score 0–
2), intermediate (TIMI risk score 3–4) and high (TIMI
risk score 5–7). Although there was no difference be-
tween strategies in the low-risk group, routine catheter-
ization was beneficial in the intermediate-risk (P ¼
.048) and high-risk (P ¼ .018) groups. (From Cannon
CP, Weintraub WS, Demopoulos LA, et al. Comparison
of early invasive and conservative strategies in patients
with unstable coronary syndromes treated with the gly-
coprotein IIb/IIIa inhibitor tirofiban. N Engl J Med
2001;344:1879.)
2. Elevated troponin (I or T)
3. New ST-segment depression
4. Recurrent angina/ischemia with signs or
symptoms of congestive heart failure
5. High-risk results from noninvasive stress
testing
6. Left ventricular ejection fraction less than
0.40
7. Hemodynamic instability
8. Sustained ventricular tachycardia
9. PCI within 6 months
10. Prior CABG
Either the early invasive strategy or early
conservative strategy is recommended if none of
these findings are present. Risk stratification by
multivariate RSS has not yet been addressed in
consensus guidelines. Likewise, the best combina-
tion of medical and invasive management for
NSTE ACS is likely to remain undetermined. As
new therapies and equipment continue to arise,
the time required to complete a clinical trial to
assess a specific combination of therapies will
persist as a barrier to this goal. Nonetheless,
stratified trials to assess benefit from routine early
 INTENSITY OF THERAPY
catheterization with drug-eluting stents, a different
antithrombotic regimen (enoxaparin in place of
UFH), or additional therapeutic agents such
clopidogrel [25] have not been completed.
Other risk stratification systems
Because risk stratification is a multivariable
issue with continuously evolving risk factors and
management options, current consensus guide-
lines describe specific clinical features indepen-
dently associated with various risk levels rather
than endorsing one single RSS. Multiple tables
are used to illustrate how specific factors derived
from the Duke Cardiovascular Databank are
associated with risk of clinical events such as
true ACS from CAD, death, and nonfatal MI [26].
With this system, patients are assessed as being at
high, intermediate, or low risk with the presence
of any one factor in that risk level grouping [27].
In efforts to maximize the predictive capacity
in risk stratification during the initial time period
of an ACS presentation, the Platelet Glycoprotein
IIb/IIIa in Unstable Angina: Receptor Suppres-
sion Using Integrilin Therapy Trial (PURSUIT)
investigators examined both dichotomous and
continuous baseline characteristics for prognostic
utility [6]. Outcomes (30-day rates of mortality
and nonfatal MI) were assessed in 9461 patients
randomly assigned to receive the Gp IIb/IIIa in-
hibitor eptifibatide or placebo upon presentation
with NSTE ACS. After univariate and subsequent
multivariate logistic regression analysis, an equa-
tion was derived to calculate the probability of
30-day mortality, including 23 weighted clinical
variables. Because information on all 23 variables
and use of a computer would be required to com-
plete this calculation, this model is not easily ap-
plicable at the bedside. A simplified model was
then developed by removal of the 16 least infor-
mative features and subsequent placement of the
remaining seven variables into a weighted point
system. To estimate risk of 30-day mortality or
nonfatal MI, or mortality alone, the sum of points
is then applied against a curve. Although the dis-
criminatory capacity of the mortality model is ex-
cellent (C-statistic 0.814), the complexity of the
model reduces its applicability at the bedside.
To assess the risk of death across the entire ACS
spectrum (UA, NSTEMI, and STEMI) and to
allow application to a more generalized patient
population, the Global Registry of Acute Coro-
nary Events (GRACE) RSS was more recently
developed from patients who had ACS sampled
75
from 94 hospitals located in 14 different countries
[7]. A nomogram was developed to facilitate risk as-
sessment by summing eight weighted variables, in-
cluding Killip class, systolic blood pressure, heart
rate, age, creatinine level, presence of cardiac arrest
at admission, ST-segment deviation, and elevated
cardiac enzyme levels. Discriminatory ability was
again demonstrated to be excellent (C-statistic
0.83). A comparison of the GRACE RSS and the
PURSUIT RSS was recently completed with appli-
cation to the Canadian ACS Registry, comprising
4627 patients who had ACS enrolled between Sep-
tember 1999 and June 2001 [28]. Again, both mod-
els were found to have excellent predictive capacity
(C statistic: 0.84 for PURSUIT, 0.83 for GRACE).
Although the GRACE RSS was found to be
well calibrated (Hosmer Lemeshow P ¼ .40), the
PURSUIT RSS calibration was suboptimal
(Hosmer Lemeshow P ! .001). This observation
demonstrates the potential for differences between
clinical trial populations and real-world patients
and reinforces the appropriateness of validation
among diverse patient populations before wide-
spread use of an RSS. In an effort to maximize pre-
dictive capacity, large numbers of variables with
associated weighting for attributable risk are em-
ployed in the PURSUIT and GRACE systems.
These systems, however, are too complex to be
used easily straight from memory, and their com-
plexity detracts from their clinical applicability. Al-
ternatively, the TIMI RSS employs seven equally
weighted variables to allow improved applicability
but at the cost of predictive capacity (C statistic
0.65). With greater ease of use, the TIMI RSS has
been more routinely tested for ability to assist in
medical decision making, but as hand-held com-
puters become more available to clinicians, the
more complicated systems are likely to be more
useful.
Future developments in risk stratification
The goal of developing optimal management
for NSTE ACS will remain a moving target.
Improvements in risk stratification tools and their
appropriate use are needed to guide clinicians
through the complicated options for management.
As such, methods and tools for risk stratification
will continue to evolve. Since the development of
the RSS discussed here, additional variables have
been recognized as prognostically important and
subsequently have been shown to provide addi-
tional stratification potential. These new risk
factors are likely to be spliced into existing
76 VESELY & KELEMEN
systems or used in the creation of new scoring
systems. As one example of an attempt to improve
stratification beyond that by existing systems,
Bazzino and colleagues [9] assessed numerous
molecular markers in conjunction with clinical
features for predictive capacity, including
N-terminal probrain natriuretic peptide (NT-
proBNP), high sensitivity C-reactive protein,
troponin T, and myoglobin. Predictive capacity
for 6-month mortality improved with additional
stratification with NT-proBNP level, beyond
that provided by the TIMI risk score for UA/
NSTEMI (Fig. 4) and the American College of
Cardiology/American Heart Association classifi-
cation system. Other clinical variables are fre-
quently examined for independent prognostic
utility and hence qualify for future RSS. Features
associated with the pathophysiologic processes in-
volved in ACS, including markers of inflamma-
tion (white blood cell count) [13] and myocardial
injury (myoglobin) [10], as well as comorbidities
such as anemia (hemoglobin) [12] and renal insuf-
ficiency (creatinine clearance) [29], have been
shown to predict adverse clinical outcomes inde-
pendently. How to best use this additional infor-
mation within the perspective of clinical decision
making and future risk models remains to be
determined.
Although molecular markers that are directly
reflective of cardiomyocyte death, such as tropo-
nin I (and T), provide prognostic information, the
delay in achieving elevated levels significant
enough for detection can be prevent their use as
a variable for early risk stratification. Biomarkers
of underlying processes leading to cellular death
have potential for providing early prognostic
information, because they will potentially reach
clinically significant elevated levels earlier within
the course of an ACS presentation. Some poten-
tial upstream biomarkers with early promise
include myeloperoxidase, metalloproteinase-9,
soluble CD-40 ligand, pregnancy-associated
plasma protein A, choline, ischemia-modified un-
bound free fatty acids, placental growth factor,
and glycogen phosphorylase isoenzyme BB [30].
Although the many clinical and biomarker factors
currently used in stratification systems account for
the majority of variance in outcomes, molecular
markers have potential to differentiate risk more
quickly and accurately. As this field continues to
develop with newly recognized biomarkers and
organization of biomarker panels, more precise
stratification within the intermediate-risk group
may allow improved delivery of optimized care.
NT-proBNP (<586) NT-proBNP (>586)
*
20
6 Month Death Rate (%)
*
15
10 *
5
0
0-2 3-4 5-7
TIMI Risk Score for UA/NSTEMI
Fig. 4. Risk stratification for 6-month mortality is en-
hanced when N-terminal probrain natriuretic peptide
(NT-proBNP levels) are checked within 7 hours of ad-
mission for NSTE ACS. Within each TIMI risk level,
likelihood of the composite death or nonfatal MI, or
of death alone, was higher with the NT-proBNP level
greater than 586 pg/mL (P ! .001 for interaction test).
(From Bazzino O, Fuselli JJ, Botto F, et al. Relative value
of N-terminal probrain natriuretic peptide, TIMI risk
score, ACC/AHA prognostic classification and other
risk markers in patients with non-ST-elevation acute
coronary syndromes. Eur Heart J 2004;25:859.)
Summary
Simple RSS allow for rapid decision making in
the emergency department. The data presented in
this article suggest that for patients at the highest
risk and the lowest risk for complications of NSTE
ACS, the scoring systems work well and allow
effective triage and treatment. For patients at
intermediate risk (30%–40% of all patients who
have ACS), however, it is not clear whether early
aggressive treatment with cardiac catheterization
or routine conservative management should be the
standard of care. The consensus guidelines are
vague, and the scoring systems discriminate less
well for these patients. The authors think that
patients at intermediate risk are best served by
initial screening with an RSS like the TRS (with
risk scores of 3–4), followed by a multimarker
strategy to define risk better. They also think that
the next step is to design clinical trials to test
strategies of care defined prospectively by risk. This
step would, in the authors’ opinion, begin the next
round of the cycle of clinical therapeutics [31]. The
treatment of patients who have NSTE ACS has
been characterized in the past 2 decades by care
based on evidence from many excellent clinical tri-
als. The consensus panels have convened and guide
patient management. Quality-improvement initia-
tives such as CRUSADE and GRACE give
 INTENSITY OF THERAPY
feedback to improve compliance with guidelines.
The understanding of risk is developing with the
help of these scoring systems. Discovery is ongoing.
The next decade of acute cardiac care will focus on
early identification of patients at high risk and on
matching the most intensive treatments to the pa-
tients most in need. Excessive testing and care pro-
motes cost inefficiency and, perhaps, increased
hazard for some patients. New trials are needed
to move these new hypotheses back into practice.
References
[1] Mourouga PGC, Rowan KM. Does it fit? Is it good?
Assessment of scoring systems. Curr Opin Crit Care
2000;6:176.
[2] Stone PH, Thompson B, Anderson HV, et al. Influ-
ence of race, sex, and age on management of unsta-
ble angina and non-Q-wave myocardial infarction:
The TIMI III registry. JAMA 1996;275:1104.
[3] Bhatt DL, Roe MT, Peterson ED, et al. Utilization
of early invasive management strategies for high-
risk patients with non-ST-segment elevation acute
coronary syndromes: results from the CRUSADE
Quality Improvement Initiative. JAMA 2004;292:
2096.
[4] Sabatine MS, Morrow DA, de Lemos JA, et al. Mul-
timarker approach to risk stratification in non-ST el-
evation acute coronary syndromes: simultaneous
assessment of troponin I, C-reactive protein, and
B-type natriuretic peptide. Circulation 2002;105:
1760.
[5] Antman EM, Cohen M, Bernink PJ, et al. The TIMI
risk score for unstable angina/non-st elevation MI:
A method for prognostication and therapeutic deci-
sion making. JAMA 2000;284:835–42.
[6] Boersma E, Pieper KS, Steyerberg EW, et al. Predic-
tors of outcome in patients with acute coronary syn-
dromes without persistent ST-segment elevation.
Results from an international trial of 9461 patients.
The PURSUIT Investigators. Circulation 2000;101:
2557.
[7] Granger CB, Goldberg RJ, Dabbous O, et al. Pre-
dictors of hospital mortality in the global registry
of acute coronary events. Arch Intern Med 2003;
163:2345.
[8] Gibson CM, Pinto DS, Murphy SA, et al. Associa-
tion of creatinine and creatinine clearance on pre-
sentation in acute myocardial infarction with
subsequent mortality. J Am Coll Cardiol 2003;42:
1535–43.
[9] Bazzino O, Fuselli JJ, Botto F, et al. Relative value
of N-terminal probrain natriuretic peptide, TIMI
risk score, ACC/AHA prognostic classification
and other risk markers in patients with non-ST-
elevation acute coronary syndromes. Eur Heart J
2004;25:859.
77
[10] de Lemos JA, Morrow DA, Gibson CM, et al.
The prognostic value of serum myoglobin in
patients with non-ST-segment elevation acute cor-
onary syndromes. Results from the TIMI 11B and
TACTICS-TIMI 18 studies. J Am Coll Cardiol
2002;40:238.
[11] Morrow DA, Rifai N, Antman EM, et al. C-reactive
protein is a potent predictor of mortality indepen-
dently of and in combination with troponin T in
acute coronary syndromes: A TIMI 11a substudy.
Thrombolysis in myocardial infarction. J Am Coll
Cardiol 1998;31:1460–5.
[12] Sabatine MS, Morrow DA, Giugliano RP, et al. As-
sociation of hemoglobin levels with clinical out-
comes in acute coronary syndromes. Circulation
2005;111:2042.
[13] Bhatt DL, Chew DP, Lincoff AM, et al. Effect of
revascularization on mortality associated with an
elevated white blood cell count in acute coronary
syndromes. Am J Cardiol 2003;92:136.
[14] Braunwald E. Unstable angina. A classification. Cir-
culation 1989;80:410.
[15] Campeau L. Grading of angina pectoris [letter]. Cir-
culation 1976;54:522.
[16] Campeau L. The Canadian Cardiovascular Society
grading of angina pectoris revisited 30 years later.
Can J Cardiol 2002;18:371.
[17] Braunwald E. Unstable angina: an etiologic ap-
proach to management. Circulation 1998;90:2219.
[18] Wu AH, Apple FS, Gibler WB, et al. National Acad-
emy of Clinical Biochemistry Standards of Labora-
tory Practice: recommendations for the use of
cardiac markers in coronary artery diseases. Clin
Chem 1999;45:1104.
[19] Harrell FE Jr, Lee KL, Mark DB. Multivariable
prognostic models: issues in developing models,
evaluating assumptions and adequacy, and measur-
ing and reducing errors. Stat Med 1996;15:361.
[20] Antman EM, McCabe CH, Gurfinkel EP, et al.
Enoxaparin prevents death and cardiac ischemic
events in unstable angina/non-Q-wave myocardial
infarction. Results of the thrombolysis in myocar-
dial infarcion (TIMI) 11B trial. Circulation 1999;
100:1593.
[21] Sabatine MS, Antman EM. The thrombolysis in
myocardial infarction risk score in unstable angina/
non-ST-segment elevation myocardial infarction.
J Am Coll Cardiol 2003;41:89S.
[22] Cohen M, Demers C, Gurfinkel EP, et al. A compar-
ison of low-molecular-weight heparin with unfrac-
tionated heparin for unstable coronary artery
disease. Efficacy and Safety of Subcutaneous Enox-
aparin in Non-Q-Wave Coronary Events Study
Group. N Engl J Med 1997;337:447.
[23] Theroux P, Ouimet H, McCans J, et al. Aspirin, hep-
arin, or both to treat acute unstable angina. N Engl J
Med 1988;319:1105.
[24] Cannon CP, Weintraub WS, Demopoulos LA, et al.
Comparison of early invasive and conservative
78 VESELY & KELEMEN
strategies in patients with unstable coronary syn-
dromes treated with the glycoprotein IIb/IIIa inhib-
itor tirofiban. N Engl J Med 2001;344:1879.
[25] Mitka M. Results of CURE trial for acute coronary
syndrome. JAMA 2001;285:1828.
[26] Antiplatelet Trialists’ Collaboration. Collabora-
tive overview of randomised trials of antiplatelet
therapy. I: prevention of death, myocardial in-
farction, and stroke by prolonged antiplatelet
therapy in various categories of patients. BMJ
1994;308:81.
[27] Braunwald E, Antman EM, Beasley JW, et al. ACC/
AHA 2002 guideline update for the management of
patients with unstable angina and non-ST-segment
elevation myocardial infarction–summary article:
a report of the American College of Cardiology/
American Heart Association Task Force on Practice
Guidelines (Committee on the Management of
Patients With Unstable Angina). J Am Coll Cardiol
2002;40:1366.
[28] Yan AT, Jong P, Yan RT, et al. Clinical trial–de-
rived risk model may not generalize to real-world
patients with acute coronary syndrome. Am Heart
J 2004;148:1020.
[29] Gibson CM, Dumaine RL, Gelfand EV, et al. Asso-
ciation of glomerular filtration rate on presentation
with subsequent mortality in non-ST-segment eleva-
tion acute coronary syndrome; observations in
13,307 patients in five TIMI trials. Eur Heart J
2004;25:1998.
[30] Apple FS, Wu AH, Mair J, et al. Future biomarkers
for detection of ischemia and risk stratification in
acute coronary syndrome. Clin Chem 2005;51:810.
[31] Califf RM, Peterson ED, Gibbons RJ, et al. Inte-
grating quality into the cycle of therapeutic develop-
ment. J Am Coll Cardiol 2002;40:1895.
 Cardiol Clin 24 (2006) 79–85

Improving Systems of Care in Primary Angioplasty

Stanley Watkins, MD, MHSa, Lynnet Tirabassi, RN, BSNb,
Thomas Aversano, MDb,*
a
Division of Cardiology, 600 North Wolfe Street, Johns Hopkins Hospital, Baltimore, MD 21287, USA
b
Division of Cardiology, Johns Hopkins School of Medicine, 5501 Hopkins Bayview Circle,
Baltimore, MD 21224, USA

The authors have had the honor and unique
opportunity to develop primary angioplasty pro-
grams at more than 40 hospitals in the United
States as part of the Cardiovascular Patient
Outcomes Research Team (C-PORT) projects.
C-PORT projects have included a completed ran-
domized trial and an on-going primary angio-
plasty registry.
The idea that simply providing appropriate
equipment and expert physicians to use it results
in a good or even adequate primary angioplasty
program, although commonly held, is deeply
naive. Expert interventionalists and state-of-the-
art equipment are necessary but far from sufficient
conditions for the appropriate, prompt, safe, and
effective application of this therapy for acute
myocardial infarction (AMI).
Primary percutaneous coronary intervention
(PPCI) is not simply a procedure that occurs in
the cardiac catheterization laboratory. Rather, it
is a strategy of care that at a minimum involves
the emergency room, coronary care unit (CCU),
and the step-down unit as well as the catheteriza-
tion laboratory. In certain circumstances, PPCI
may importantly involve prehospital care, as well.
Failure to recognize PPCI as a strategy of care,
not simply a procedure, is the most common and
most dangerous misconception about this form of
therapy.
What follows is not a manual, guideline, or
prescription for development of a PPCI program.
* Corresponding author. Johns Hopkins Medical In-
stitutions JHAAC 1B.40, 5501 Hopkins Bayview Circle,
Baltimore, MD 21224.
E-mail address: taversan@jhmi.edu (T. Aversano).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.009
Program development is an immensely detailed
undertaking whose description is far beyond the
scope of this report. Furthermore, because each
institution is unique in its resources, requirements,
and culture, development procedures and specifics
are peculiar to each institution. There is no
cookbook for PPCI development. Nevertheless,
there are elements of program development and
issues that must be addressed that are common
to all or many institutions. This report provides
a general description of the process of PPCI
program development.
Program elements
The original C-PORT PPCI development pro-
gram is divided arbitrarily into four main compo-
nents: (1) the setting of standards, (2) training of
staff, (3) development of logistics, and (4) imple-
mentation of a quality and error management
program. This program has been incorporated
into the American College of Cardiology (ACC)/
American Heart Association guidelines for both
AMI and PCI [1,2].
Standards
‘‘Standards’’ refer to those for physicians,
facilities, equipment, treatment guidelines and
the nurses and technicians caring for the PPCI
patient. Interventional physician and treatment
guidelines (eg, ACC guidelines) are provided by
national organizations; facilities standards are
typically set by state regulation. Balloons, stents,
and rheolytic or suction thrombectomy devices
are the main devices used. Various forms of
atherectomy or ablating devices are specifically
cardiology.theclinics.com
80 WATKINS
excluded; otherwise, equipment standards are set
by the physicians performing intervention. Com-
petency standards for nurses and technicians
caring for the PPCI patient in the catheterization
laboratory and postprocedure units do not exist.
The training of these staff members is an impor-
tant part of program development.
Training
No amount of training can take the place of
actual experience. Nevertheless, training in the
elements of care specifically related to PPCI can
be useful for catheterization laboratory nurses
and technicians who are already expert at caring
for patients undergoing diagnostic catheterization
and for CCU nurses already expert at caring for
patients who have AMI.
Training can be didactic, observational, or
hands-on, depending on need and expectation.
Didactic training takes the form of lectures given
by experts in a particular area (eg, postprocedure
care discussed by an experienced postprocedure
nurse, drugs used in PPCI discusses by practi-
tioners or pharmacists) and in-services given by
vendors of devices and drugs used in PPCI. These
programs are extremely useful but are limited by
their nature as lectures. Supplementing didactic
teaching with observational training in the staff
member’s area of care is extremely helpful. Thus
the catheterization laboratory nurse and techni-
cian staff can spend anywhere from 1 day to a
week or more observing elective PCI procedures
at an affiliated tertiary hospital, and the CCU and
step-down unit nursing staff may spend a similar
amount of time in the postprocedure care area of
the same tertiary hospital. Having this experience
in an affiliated tertiary hospital (one in which the
interventionalists practice) has the added benefit
of allowing community hospital staff to become
acquainted with the policies, procedures, and
methods of care to which their interventionalists
are accustomed and also allows personal contact
with highly experienced individuals (other nurses
and other technicians) who may be resources in
the future.
If an individual in the catheterization labora-
tory is to be a second operator, experience in the
elective setting is important before being involved
in a procedure during an AMI. This hands-on
training is best accomplished in the affiliated
tertiary hospital.
Although this program development is neces-
sary, it is not sufficient for development of a PPCI
et al
program, because it fails to provide a mechanism
for sustaining program excellence.
Logistics
‘‘Logistics’’ refers to the policies and proce-
dures that allow delivery of excellent PPCI treat-
ment. Literally scores of issues must be addressed
in development of logistics within a community
hospital, and they tend to be peculiar to that
hospital. Again, there is no cookbook. Despite the
unique nature of specific logistical elements, there
are nevertheless logistical goals common to all
institutions. These are (1) immediate aspirin ad-
ministration, (2) door-to-balloon time of 90 min-
utes or less, and (3) door-to-needle (or transfer)
time of 30 minutes or less. The first two elements
need no explanation. The goal of thrombolytic
administration within 30 minutes of arrival at the
emergency department recognizes that, no matter
how well a system for performing PPCI is de-
veloped, it will fail. A formal plan for the fallback
therapy (whether thrombolytics or transfer) needs
to be in place. The 30-minute timeframe is justified
because that is the national standard for door-to-
needle time. When transfer for PCI elsewhere is the
fallback plan, to this 30 minutes must be added the
time to recall an ambulance for interhospital
transport, pick up of the patient, delivery to the
transfer hospital, removal to the cardiac catheter-
ization laboratory, diagnostic catheterization, and
then balloon inflation if appropriate. This process
will probably take at least 1 hour and much longer
in most instances.
It is impossible to detail logistics here. But
attention to detail is critical. The authors can
highlight several important steps in the acute care
of the patient who has AMI.
Emergency medical services
Prehospital care
Multiple therapies have been proven to limit
infarct size and improve mortality associated with
myocardial infarction (MI). All of these therapies
are time dependant, having significantly more
efficacy when applied earlier than later. In most
cases of acute myocardial infraction, a large
portion of the delay arises from delay in the
patient’s activating emergency medical services
(EMS). This delay from the time from onset of
signs and symptoms of AMI to seeking medical
assistance is termed ‘‘patient delay.’’ ‘‘Transfer
delay’’ describes the amount of time it takes for
a patient to arrive at a care center once EMS has
been activated. ‘‘Prehospital delay’’ refers to the
 IMPROVING CARE IN PRIMARY ANGIOPLASTY
total of patient delay plus transfer delay. The
average prehospital delay in the United States is
several hours, and the majority of this delay
comes from the patient’s delaying seeking medical
attention [3].
Several studies have attempted to improve the
prehospital delay using community-based inter-
ventions such as direct mail, radio, newspaper,
and television advertising. This advertising is an
attempt to educate the populace about the signs
and symptoms of AMI and about the need for
rapid diagnosis and treatment. In summary, the
data regarding the effect of these community-
based interventions are inconsistent. Earlier non-
randomized trials had suggested a significant
decline in patient delay following an education
campaign [4–6]. A large prospective, randomized
controlled trial was undertaken to address this is-
sue in the late 1990s. The Rapid Early Action for
Coronary Treatment trial randomly assigned 20
communities in the United States to an aggressive,
multifaceted education campaign versus no inter-
vention for an 18-month period. The intervention
communities did decrease their prehospital delay
time by 4.7% per year, but this change was signif-
icantly different from that of the control commu-
nities, which decreased 6.7% per year. EMS use
increased by 20% in the intervention communities
compared with the reference communities, but
this increased use did not translate into an in-
crease in emergency department visits or hospital
discharges for acute cardiac care [7]. Thus it seems
that aggressive community education does not
directly affect prehospital delay and that new
strategies are needed to improve patient delay sig-
nificantly. The authors’ group is conducting re-
search to identify a group of high-risk patients
who take and transmit a 12-lead ECG at the ini-
tial onset of symptoms.
Emergency medical services
Once the patient activates EMS, the blame for
delays in care falls on the efficiency of the health
care system. The delay from EMS activation to
EMS arrival on the scene is largely a function of
EMS availability and is beyond the scope of this
discussion. Researchers have examined multiple
ways to improve the transfer delay, which results
from post-EMS arrival delays.
It is standard of care to do a brief assessment
and administer chewed aspirin immediately upon
arrival at the scene of a chest pain victim.
Currently, the standard of care is to apply
81
a three-lead rhythm monitor upon initial assess-
ment of the patient and to have immediate access
to external defibrillation. Although adequate to
identify most rhythm disturbances, the rhythm
strip does not give reliable information regarding
acute ischemia. Furthermore, most of these de-
vices are unable to transfer data electronically for
physician interpretation. In experienced hands,
performing a prehospital 12-lead ECG adds only
2 to 3 minutes to the initial assessment and can
significantly affect downstream management. Less
than 25% of patients who have chest pain trans-
ported by EMS have a final diagnosis of MI, and
the non-ECG clinical predictors (diaphoresis, heart
block) are insensitive [8,9]. Some have suggested
that prehospital ECGs may result in further trans-
port delay. A study examining on-scene times for
paramedics using 12-lead ECGs compared with
standard three-lead monitors found an on-scene
time of 21.9 minutes for the ECG group and 22
minutes for the 3-lead group [10].
Issues have also been raised regarding the
ability of EMS personnel to interpret a 12-lead
ECG. EMS staff have been shown to diagnose
94% to 97% of physician-confirmed AMI cor-
rectly with prehospital ECGs after only 4 to 8
hours of training [11,12]. The legal issues remain
an obstacle, in that nonphysician interpretation
of the ECG is an off-label use of this device.
Most EMS systems that routinely employ preho-
spital ECG testing transmit the ECG to an emer-
gency department–based physician who confirms
the interpretation and aids in guiding therapy.
The ECG data can be transmitted by cellular tele-
phone (in transit) or landline (on site). The former
is less time-consuming but more prone to failed
transmission.
Several studies have examined the downstream
impact of the prehospital ECG. The Cincinnati
Heart Project documented a reduction in preho-
spital treatment delay from 50 minutes in patients
who were transported by EMS but did not receive
a prehospital ECG to 30 minutes in patients who
were transported by EMS but did receive a pre-
hospital ECG [13]. The Myocardial Infarction
Triage and Intervention Trial demonstrated a re-
duction in prehospital delay from 102 minutes to
46 minutes after the institution of system-wide
prehospital ECGs [14].
Following initial assessment, EMS personnel
routinely administer aspirin, nitroglycerin, and
beta-blockers. These preliminary therapies can
result in immediate improvement of ischemia
and the ECG changes associated with ischemia.
82 WATKINS
If the first 12-lead ECG is performed upon arrival
to the emergency department, the initial ischemic
changes may no longer be present, which can lead
to underdiagnosis of an ACS. In several European
countries the prehospital ECG is routinely used to
guide the administration of fibrinolytic therapy in
the ambulance. The largest of the trials to examine
the treatment effect of prehospital fibrinolysis
was the European Myocardial Infarction Project
trial in Europe. Patients who had ST-segment
elevation myocardial infarction (STEMI) (N ¼ 4767)
were randomly assigned to prehospital versus
in-hospital fibrinolysis. The patients randomly
assigned to prehospital treatment received a fibrino-
lytic medication a median of 55 minutes earlier
than those who received in-hospital medication.
There was a trend toward a reduction in mortality,
with a 30-day mortality rate of 9.7% in the
prehospital group compared with 11.1% in the
in-hospital group (P ¼ .08) [15]. The Grampian Re-
gion Early Anistreplase Trial was a randomized,
double-blind parallel-group clinical trial that en-
rolled 311 patients suspected of having AMI.
Patients were randomly assigned to receive intrave-
nous anistreplase (30 U) either at home or later, af-
ter arrival in the hospital. The median time saved
by prehospital thrombolysis was 130 minutes. By
the end of 1 year after trial entry, 17 (10.4%) of
163 patients given anistreplase at home died, com-
pared with 32 (21.6%) of 148 patients allotted anis-
treplase in the hospital (P ¼ .007) [16,17]. A recent
meta-analysis of all prehospital fibrinolysis trials
found an overall significant 20% reduction in hos-
pital mortality favoring prehospital therapy [18].
This treatment effect is significantly greater in rural
communities where transport times are prolonged
and time from symptom onset to treatment can
be significantly decreased by giving medication
on route.
Beyond giving EMS personnel the ability to
diagnose and treat AMI earlier, the prehospital
ECG allows efficient triage of patients who have
AMI to centers where primary angioplasty is
offered. Primary angioplasty has become the pre-
ferred method of acute reperfusion therapy for
many reasons. For the average patient who has
STEMI, primary angioplasty confers a survival
advantage over thrombolysis. A recent meta-
analysis documented a 30% reduction in short-
term death and nearly a 50% reduction in nonfatal
reinfarction [19]. The main barrier to the universal
implementation of PPCI is the availability of this
technology at local hospitals. The Danish Acute
Myocardial Infarction trial supported the safety
et al
of transporting patients who had AMI to facilities
with angioplasty capabilities compared with on-
site thrombolysis [20]. The clear superiority of pri-
mary angioplasty has sparked a debate over
prehospital triage of AMI patients. How far is
too far when a patient requires reperfusion and
a thrombolytic-only hospital is closer than the
nearest primary angioplasty center? The issue is
further clouded by political and economic factors
that have complicated the formation of an orga-
nized triage system for AMI similar to the one
that exists for trauma in the United States. Cur-
rently, triage of patients who have AMI by EMS
personnel should be guided by the goal of prompt
reperfusion, and formal protocols should exist for
consideration of hospital transfer.
Emergency department
Development of logistics in the emergency
department is conditioned and constrained by
the goal of achieving a reperfusion time of 90
minutes. The best mode of reperfusion is de-
pendant on the availability of primary angioplasty
at that institution. Because, on average, fibrino-
lysis takes longer to achieve reperfusion than PCI,
a door-to-needle time of only 30 minutes is the
goal for fibrinolysis. Primary angioplasty is more
effective than fibrinolysis provided the door-
to-balloon time is less than 90 minutes. During
off-hours when the angioplasty team is not on site,
activating the team can take a significant amount
of time. Therefore rapid upstream identification
of reperfusion-eligible patients is imperative. This
identification requires well-considered formal pro-
tocols to ensure that no patients who have AMI
are burdened with unnecessary delay while being
assessed in the emergency department. The for-
mation of such protocols improves outcomes and
is cost effective [21].
The emergency department chest pain protocol
must be straightforward, robust, and user-friendly.
Furthermore, with emergency department visits
increasing every year, the burden logistics places
on the emergency department staff must be mini-
mized. The logistical goals in the emergency de-
partment are that every patient who has STEMI
receive aspirin immediately unless contraindicated,
the door-to-needle time be 30 minutes or less, and
the door-to-balloon time be 90 minutes or less. The
third goal translates to a door-to-catheterization
laboratory time of approximately 60 minutes.
Although these goals are common to every PPCI
program, the way those goals are achieveddhow
 IMPROVING CARE IN PRIMARY ANGIOPLASTY
the many details of local logistics are addresseddis
unique to each institution.
It is important that an agreed-upon set of rules
defining eligibility criteria for PPCI be developed
and easily accessed for review for every potential
candidate. Chest pain and ST-segment elevation
(or new or presumed new left bundle branch
block) are important inclusion criteria, but other
issues also need to be addressed. Is the patient in
cardiogenic shock going to be included or always
transferred to a tertiary center? Do severe allergies
to aspirin or contrast exclude patients? A carefully
thought-out and universally agreed-upon set of
criteria for considering patients for PPCI is
mandatory. It is important that physician prefer-
ences not be criteria for inclusion or exclusion:
having two standards of care, one for some
physicians’ patients and one for other physicians’
patients, is a bad idea from every point of view.
Emergency department physicians cannot be put
in the position of having to recommend two
different care standards for the same disease
depending on whose patient they are seeing.
When the system required for performing
PPCI fails, an agreed-upon response must be in
place. For thrombolytic-ineligible patients, this
response usually means transfer to a tertiary
institution for PCI. For thrombolytic-eligible
patients, this response may mean either adminis-
tration of thrombolytic therapy or transfer to
a tertiary facility. Combinations of these ap-
proaches may also be considered (so-called ‘‘drip
and ship’’ treatment). Whatever the fallback
therapy may be for these patients, it is important
that they are specified, clear, reviewable by the
emergency department physician as he or she is
caring for the patient, and agreed upon by all
concerned.
Delay in the application of PPCI is most often
caused by failure to identify the patient as having
a possible AMI. This failure of identification may
occur at the point of triage and typically involves
unusual presentations (eg, no chest pain but only
fatigue or other nonspecific complaint) and a busy
emergency department. These patients may be
sent to the waiting room while more acutely ill
patients are cared for, with acute, ST-segment
elevation discovered only later on the first ECG.
The patient also may not be rapidly identified
because of an ECG that is unclear or equivocal.
The mechanism to determine the availability of
catheterization laboratory itself and its physicians
and staff is a critical piece of emergency depart-
ment logistics. In the more than 40 institutions in
83
which C-PORT has developed PPCI programs,
this system almost always has been considered to
be in place already. It almost never is.
It is impossible to specify how team recruitment
should work in all hospitalsdlike so many logis-
tical elements, its specifics depend upon the in-
stitution itself, its culture, and its resources. The
goals of the system, however, are uniform. The
emergency department cannot be in charge of such
a system. Rather, the emergency department must
be able to activate the team recruitment system
with a single outgoing call to an activator. The
activator then sends a call out to the team
(physicians, nurses, technicians) who must then
respond back to the activator. The activator logs
responses and, when the team is complete, informs
the emergency department that team is assembled.
If the team cannot be assembled, the activator
informs the emergency department that team re-
cruitment has failed so that fallback therapy can be
instituted.
Because fallback therapy is often thrombolytic
therapy, there is goal of a total elapse time of 30
minutes from arrival at the emergency department
to application of thrombolytic therapy. If it takes
15 minutes to identify the patient and 5 minutes to
apply thrombolytic therapy after the decision is
made to do so, 10 minutes are available for team
recruitment to succeed or fail. Reducing the time
from arrival to patient identification to 10 minutes
means 15 minutes are available. This is a very
short time. Changes in the on-call schedule, pager
failure, and numerous other common problems
can influence the success of this system. It is
critical that the system be developed and imple-
mented in a way that identifies all potential failure
points and minimizes their influence.
It is important that the emergency department
and interventional physicians agree on initial care.
Routine AMI laboratory tests should be obtained,
but also a tube of blood should be sent to the blood
bank for potential crossmatching in case of signif-
icant bleeding. All patients should receive aspirin
unless otherwise contraindicated. Similarly, beta-
blockers, nitrates, oxygen, and analgesics are given
almost uniformly. At this time, there is a wide
variety of further initial therapies that may or may
not include clopidogrel, a glycoprotein IIb/IIIa
receptor antagonist, unfractionated or low molec-
ular weight heparin, and occasionally thrombolytic
therapy (eg, with facilitated PCI). Within limits,
agreement among interventionalists and emer-
gency department physicians about what is done
is more important than exactly what is done.
84 WATKINS
The ability to conference the interventionalist
with the emergency department physician is
extremely helpful in defining patient-specific
treatment.
Catheter laboratory
Almost uniformly, the major issue involving
the catheterization laboratory concerns staffing.
Community hospitals without on-site cardiac
surgery usually do not have sufficient catheteriza-
tion laboratory staff to cover PPCI 24 hours
a day, 7 days a week. Ideally, there are four
individuals in each on-call position. Thus, if the
call team is one interventionalist, two nurses, and
a technician, there should be four intervention-
alists, eight nurses, and four technicians covering
call. This staffing leads to a livable and therefore
sustainable on-call schedule of approximately 1
weekday a week and 1 weekend per month (with
a weekend being Friday, Saturday, and Sunday).
This level of staffing can be difficult to secure, and
often PPCI programs begin with fewer individuals
covering call.
Coronary care unit
The CCU is the postprocedure care area for
the patient who has undergone PPCI. Most
hospitals without on-site cardiac surgery care for
a large number of patients who have AMI, so
typically the capability of the CCU staff to
monitor and manage complications of AMI is
already expert. There are, however, a number of
issues specific to patients undergoing PPCI that
need to be addressed. Chief among these is
management of the procedure site, typically the
femoral area. In most units, the nurse is the first
line in both the detection and management of
sheath complications. These individuals must,
therefore, have training in identifying the signs
and symptoms of postprocedure groin complica-
tions as well as training in initial management.
Management of bleeding complications should
include actual hands-on training of a representa-
tive group of CCU nurses who may be called on
in each shift to compress a bleeding site. This
training may be accomplished by compressing
patients after diagnostic catheterization patients.
Observational training at a high-volume tertiary
center is useful for all CCU nurses who care for
the patient after PCI. The details of how training
is done, what kind of training is done, how many
nurses are trained, and for how long is site
et al
specific. In addition to developing competency,
maintenance of competency must also be defined.
Predischarge education is a critical aspect of
PPCI. In addition to the usual predischarge
education that includes life style, drugs, activity,
rehabilitation, and so forth, discussions specific to
PPCI must be added. A stent card and a closure
device card should be given to the patient along
with education about both devices. Particularly
important is making sure that clopidogrel is
available to the patient for the duration required.
This requirement may mean involving social
services and pharmacy early in the patient’s
course so that it can be confirmed that the patient
can afford the drug. If not, programs by the
manufacturer can be used to provide this critical
medication. If the patient leaves the hospital and
fails to take this drug because the patient did not
recognize its importance, because of ineffective
education, or because the patient cannot afford it,
the entire PPCI strategy of care can be much less
effective and more dangerous than thrombolytic
therapy.
Summary
AMI is a life-threatening condition. Poor per
formance on the part of caregivers can result in
the death of a patient. It is critical that a PPCI
capability be developed in such a way that error is
minimized. It is not enough that the system works
well or very well. Aviation is often used as the
example that medical systems should emulate. In
developing the many interrelated systems required
to function properly to ensure safe, effective,
prompt, and appropriate application of PPCI,
an aviation parallel should be kept in mind. If you
were walking on the jetway toward a plane and
were greeted by the pilot who said to you, ‘‘You
know, I can land this thing 99% of the time,’’ you
would never get on that plane. It is important to
develop a PPCI system that is absolutely never the
cause of harm to any patient. Doing so requires
exquisite attention to detail, algorithms of care
when possible, redundancy, and clear orders for
all drugs and procedures.
References
[1] Antman EM, Anbe DT, Armstrong PW, et al. ACC/
AHA guidelines for the management of patients with
ST-elevation myocardial infarction; a report of the
American College of Cardiology/American Heart
Association Task Force on Practice Guidelines
 IMPROVING CARE IN PRIMARY ANGIOPLASTY
(Committee to Revise the 1999 Guidelines for the
Management of Patients with Acute Myocardial
Infarction). J Am Coll Cardiol 2004;44:E1–211.
[2] Smith SC Jr, Dove JT, Jacobs AK, et al. ACC/AHA
guidelines for percutaneous coronary intervention
(revision of the 1993 PTCA guidelines)dexecutive
summary: a report of the American College of Car-
diology/American Heart Association Task Force on
Practice Guidelines (Committee to Revise the 1993
Guidelines for Percutaneous Transluminal Coro-
nary Angioplasty) endorsed by the Society for Car-
diac Angiography and Interventions. Circulation
2001;103:3019–41.
[3] Kainth A, Hewitt A, Sowden A, et al. Systematic re-
view of interventions to reduce delay in patients with
suspected heart attack. Emerg Med J 2004;21:506–8.
[4] Blohm M, Hartford M, Karlson BW, et al. A media
campaign aiming at reducing delay times and in-
creasing the use of ambulance in AMI. Am J Emerg
Med 1994;12:315–8.
[5] Gaspoz JM, Unger PF, Urban P, et al. Impact of
a public campaign on pre-hospital delay in patients
reporting chest pain. Heart 1996;76:150–5.
[6] Herlitz J, Blohm M, Hartford M, et al. Follow-up of
a 1-year media campaign on delay times and ambu-
lance use in suspected acute myocardial infarction.
Eur Heart J 1992;13:171–7.
[7] Luepker RV, Raczynski JM, Osganian S, et al. Ef-
fect of a community intervention on patient delay
and emergency medical service use in acute coronary
heart disease: The Rapid Early Action for Coronary
Treatment (REACT) Trial. JAMA 2000;284:60–7.
[8] Tresch DD, Brady WJ, Aufderheide TP, et al. Com-
parison of elderly and younger patients with out-of-
hospital chest pain. Clinical characteristics, acute
myocardial infarction, therapy, and outcomes.
Arch Intern Med 1996;156:1089–93.
[9] Hargarten KM, Aprahamian C, Stueven H, et al.
Limitations of prehospital predictors of acute myo-
cardial infarction and unstable angina. Ann Emerg
Med 1987;16:1325–9.
[10] Brown JL Jr. An eight-month evaluation of preho-
spital 12-lead electrocardiogram monitoring in Bal-
timore County. Md Med J 1997;(Suppl):64–6.
[11] Foster DB, Dufendach JH, Barkdoll CM, et al.
Prehospital recognition of AMI using independent
nurse/paramedic 12-lead ECG evaluation: impact
85
on in-hospital times to thrombolysis in a rural
community hospital. Am J Emerg Med 1994;12:
25–31.
[12] Ferguson JD, Brady WJ, Perron AD, et al. The pre-
hospital 12-lead electrocardiogram: impact on
management of the out-of-hospital acute coronary
syndrome patient. Am J Emerg Med 2003;21:
136–42.
[13] Kereiakes DJ, Gibler WB, Martin LH, et al. Relative
importance of emergency medical system transport
and the prehospital electrocardiogram on reducing
hospital time delay to therapy for acute myocardial
infarction: a preliminary report from the Cincinnati
Heart Project. Am Heart J 1992;123:835–40.
[14] Weaver WD, Eisenberg MS, Martin JS, et al. Myo-
cardial Infarction Triage and Intervention Projectd
phase I: patient characteristics and feasibility of
prehospital initiation of thrombolytic therapy.
J Am Coll Cardiol 1990;15:925–31.
[15] Boissel JP. The European Myocardial Infarction
Project: an assessment of pre-hospital thrombolysis.
Int J Cardiol 1995;49(Suppl):S29–37.
[16] GREAT Group. Feasibility, safety, and efficacy of
domiciliary thrombolysis by general practitioners:
Grampian region early anistreplase trial. BMJ 1992;
305:548–53.
[17] Rawles J. Halving of mortality at 1 year by domicil-
iary thrombolysis in the Grampian Region Early
Anistreplase Trial (GREAT). J Am Coll Cardiol
1994;23:1–5.
[18] Morrison LJ, Verbeek PR, McDonald AC, et al.
Mortality and prehospital thrombolysis for acute
myocardial infarction: a meta-analysis. JAMA
2000;283:2686–92.
[19] Keeley EC, Boura JA, Grines CL. Primary angio-
plasty versus intravenous thrombolytic therapy for
acute myocardial infarction: a quantitative review
of 23 randomised trials. Lancet 2003;361:13–20.
[20] Andersen HR, Nielsen TT, Rasmussen K, et al. A
comparison of coronary angioplasty with fibrinolytic
therapy in acute myocardial infarction. N Engl J
Med 2003;349:733–42.
[21] Farkouh ME, Smars PA, Reeder GS, et al. A clinical
trial of a chest-pain observation unit for patients
with unstable angina. Chest Pain Evaluation in the
Emergency Room (CHEER) Investigators. N Engl
J Med 1998;339:1882–8.
 Cardiol Clin 24 (2006) 87–102

Moving from Evidence to Practice in the Care of

Patients Who Have Acute Coronary Syndrome
Kelly L. Miller, MDa, Charles V. Pollack Jr., MA, MD, FACEPb,
Eric D. Peterson, MD, MPHc,d,*
a
Division of Cardiology, University of Maryland School of Medicine, 22 South Greene Street,
Baltimore, MD 21201, USA
b
Department of Emergency Medicine, Pennsylvania Hospital, 800 Spruce Street Philadelphia, PA 19107, USA
c
Duke University Medical Center, Box 3236, Durham, NC 27710, USA
d
Duke Clinical Research Institute, 2400 Pratt Street, Room 7009, Durham, NC, 27705 USA

The spectrum of acute coronary syndromes
(ACS), ranging from unstable angina to non–
ST-segment elevation myocardial infarction
(NSTEMI), to ST-segment elevation myocardial
infarction (STEMI), represent major causes of
morbidity and mortality for patients who have
cardiovascular disease. The field of ACS manage-
ment underwent a remarkable evolution during
the decade of the 1990s. This change was caused
primarily by a large number of randomized
clinical trials studying all aspects of the diagnosis
and treatment of ACS, including, but not limited
to, the development of cardiac biomarker assays,
early use of antithrombotic and antiplatelet drugs,
beta-blockers, angiotensin-converting enzyme
(ACE) inhibitors, and early invasive treatment
strategies. In recognition of the importance and
the breadth of these data, clinical practice guide-
lines (CPGs) have been developed by the Amer-
ican College of Cardiology (ACC) and the
American Heart Association (AHA) to provide
standards for the diagnosis and treatment of
patients who have ACS and to construct a frame-
work for clinical decision making [1–6]. These
guidelines have also been translated into user-
friendly form for emergency medicine physi-
cians [7–9]. Despite the publication and wide
* Corresponding author. Duke Clinical Research In-
stitute, Box 17969, Durham, NC 27715.
E-mail address: peter016@mc.duke.edu
(E.D. Peterson).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.008
distribution of these guidelines, their routine clin-
ical application has been slow, incomplete, and of-
ten ineffective [1,2,6].
This article discusses the spectrum of ACS but
focuses primarily on patients who have non–ST-
segment elevation (NSTE) ACS. After reviewing
the ACC/AHA guidelines for the care of NSTE
ACS patients, it discusses current practice pat-
terns as documented in several large registries
including the National Registry of Myocardial
Infarction (NRMI), the Global Registry of Acute
Coronary Events (GRACE), and the Can Rapid
risk stratification of Unstable angina patients
Suppress ADverse outcomes with Early
implementation of the ACC/AHA guidelines
(CRUSADE). Gaps in patient management be-
tween that recommended in CPGs and actual
patterns of care are discussed, and areas for impro-
vement are suggested. A discussion on means to
overcome barriers to guideline adherence through
quality improvement initiatives concludes this
article.
Emergency department care of the patient who has
acute coronary syndromes
There are more than 5 million emergency
department visits for chest pain and more than 1
million hospitalizations for acute myocardial in-
farction (AMI) annually in the United States alone
[10,11]. Despite advances in treatment, coronary
heart disease remains the leading cause of death
in the United States [12]. Patients who have
cardiology.theclinics.com
88 MILLER
a completely occluded artery are usually identified
rapidly, on their presentation to the emergency
department, because they typically present with
ST-segment elevation or a known-to-be-new left
bundle branch block on the initial ECG. In pa-
tients who have STEMI, prompt reperfusion ther-
apy, either with fibrinolytic therapy or primary
percutaneous coronary intervention (PCI), is indi-
cated; both approaches can reduce infarct size and
improve survival [13]. Up to two thirds of patients
presenting with ACS, however, do not present
with ST-segment elevation on their initial ECG.
These patients require further risk stratification,
usually by serial evaluation, to distinguish where
they fall in the spectrum of ACS, from noncoro-
nary chest pain to unstable angina to NSTEMI.
This risk stratification is a dynamic process that re-
quires time for data collection and observation of
symptoms but is critical in determining appropri-
ate management based on patient risk.
Guidelines for evidence-based care of patients who
have acute coronary syndromes
The ACC and AHA first published guidelines
for the evaluation and management of unstable
angina in 1994 [4]. The guidelines were revised in
2000, and after the publication of several trials
pertaining to the NSTE ACS population, an up-
date was published in 2002 [5,14]. These guidelines
define classes of indications for therapy based on
the strength and consistency of evidence support-
ing them (Box 1).
Prehospital factors leading to delays in treatment
For most patients who have ACS, the initial
contact with health care providers is through the
emergency department and may or may not be
preceded by prehospital evaluation and transpor-
tation by ambulance. Despite public education
initiatives, the time from symptom onset to the
initiation of therapy in large-scale clinical trials
during the last decade (Global Use of Strategies to
Open Occluded Coronary Arteries I, III, and V and
Assessment of the Safety and Efficacy of a New
Thrombolytic II and III) has remained unchanged,
with a 2.7- to 2.9-hour delay being consistently
documented. Hospital-based delays in door-to-
needle or door-to-drug time have improved during
the last decade, but patient factors (ie, delay
between onset of symptoms to activating emer-
gency medical services) have not [15]. The Rapid
Early Action for Coronary Treatment study
et al
Box 1. Definitions of indications by the
ACC/AHA [14]
Class I indication: Conditions for which
there is evidence and/or general
agreement that a given procedure or
treatment is useful and eective
Class II indication: Conditions for which
there is conflicting evidence and/or
a divergence of opinion about the
usefulness/ecacy of a procedure or
treatment
IIa: Weight of evidence/opinion is in
favor of usefulness/efficacy
IIb: Usefulness/efficacy is less well
established by evidence/opinion
Class III indication: Conditions for which
there is evidence and/or general
agreement that the procedure/
treatment is not useful/effective and
in some cases may be harmful
The weight of the evidence was ranked
highest (A) if the data were derived
from multiple, randomized clinical
trials that involved large numbers
of patients and intermediate (B) if the
data were derived from a limited
number of randomized trials
that involved small numbers of
patients or from careful analyses
of nonrandomized studies or
observational registries. A lower rank
(C) was given when expert consensus
was the primary basis for the
recommendation.
From Brunwald E, Antman EM, Beasley
JW, et al. ACC/AHA guideline update for the
management of patients with unstable angina
and non-ST–segment elevation myocardial
infarction–2002: a report of the American
College of Cardiology/American Heart Asso-
ciation Task Force on Practice Guidelines
(Committee on the Management of Patients
with Unstable Angina). Circulation 2002;106:
1893–1900.
examined factors that lead to a patient’s delay in
seeking medial attention in hopes of devising
ways to improve quality [15,16]. This study, drawn
from data in 20 communities across the country,
demonstrates that 89.4% of community survey
 CARE OF PATIENTS WHO HAVE ACS
respondents would use emergency medical services,
but only 23.2% of patients who had chest pain ac-
tually did use emergency medical services; denial of
a possible cardiac cause of the chest pain played
a critical role in the delay [16]. The disconnect be-
tween what people seem to understand is the right
course of action in a survey and actual patient be-
havior points to certain modifiable situational or
belief factors and supports a role for patient-fo-
cused education initiatives in improving delays in
seeking medical care.
To improve these delays, the National Heart,
Lung, and Blood Institute has launched several
programs to educate patients about their personal
risks of heart disease, symptoms of myocardial
infarction, and the proper action to take if they
think they are having a heart attack. The goal of
these programs is to reduce the time between the
onset of symptoms and treatment [15,17]. Patients
too must play a role in delivering timely medical care.
Initial emergency department triage
Once a patient suspected of having ACS arrives
at the emergency department, prompt and accu-
rate triage is essential. A determination should be
made in all patients who have chest discomfort as
to whether the likelihood of acute ischemia caused
by coronary artery disease is high, intermediate, or
low. Important also is the realization that many
patients, especially women and diabetics, do not
always present with classic chest pain, often
presenting instead with potentially confounding
anginal equivalents including nausea, dyspnea, or
pain in the back, neck, jaw, or epigastrium.
In the absence of ST-segment elevation on the
initial ECG, differentiating NSTEMI or unstable
angina from low-risk stable angina or noncardiac
chest pain can be challenging. This distinction,
however, is crucial in assessing risk of myocardial
infarction and death and in deciding on appro-
priate medical therapy and management strategy.
Using a composite of clinical predictors to assess
risk is an important role of the emergency de-
partment physician. Besides history (tempo and
severity of angina, including presence of ongoing
chest pain), physical examination (detection of
heart failure, other signs of vascular disease) and
the ECG (ST-segment depression or transient
elevation), cardiac biomarkers are an important
risk-stratification tool. Elevation of cardiac tro-
ponin is a marker of high risk. In fact, the degree
of troponin elevation has been correlated with in-
hospital mortality in many studies. For example,
89
in a large community study of ACS, the mortality
rate nearly doubled as serum troponin went from
normal to above the upper limit of normal and
then rose linearly thereafter [18,19].
Using cardiac troponin in addition to other
clinical variables, multivariate models of risk
assessment have been developed; the most com-
monly used and most validated are the Throm-
bolysis In Myocardial Infarction (TIMI) risk
score developed by Antman and colleagues [20].
This seven-item risk scale assigns one point for
the presence of each of the following: age greater
than 65 years, presence of more than three coro-
nary risk factors, prior angiographic coronary ob-
struction, ST-segment deviation, more than two
angina events within 24 hours, use of aspirin with-
in 7 days, and elevated cardiac markers. The risk
of developing an adverse outcomeddeath, (re)in-
farction, or recurrent severe ischemia requiring
revascularizationdranged from 5% to 41%,
increasing with the sum of the individual varia-
bles. The score was derived from data in the
TIMI 11B trial and has been validated in three ad-
ditional trialsd Efficacy and Safety of Subcutane-
ous Enoxaparin in Non-Q Wave Coronary
Events, Treat Angina with Aggrastat and Deter-
mine Cost of Therapy with an Invasive or Conser-
vative Strategy (TACTICS)-TIMI 18, and Platelet
Receptor Inhibition in Ischemic Syndrome Man-
agement in Patients Limited by Unstable Signs
and Symptoms [14,20–24]. Among patients who
have unstable angina/NSTEMI, there is a progres-
sively greater benefit from newer therapies such as
low molecular weight heparin (LMWH), platelet
glycoprotein (Gp) IIb/IIIa inhibition, and an in-
vasive strategy, with increasing risk score [20–
25]. This benefit is highlighted in the revised
ACC/AHA guidelines from 2002 [14].
American College of Cardiology/American Heart
Association recommendations for initial triage,
ECG, and biomarker assessment
Box 2 gives the ACC/AHA recommendations
for early risk stratification.
Acute medical therapies
Because of delays in or lack of bed availabil-
ity in many hospitals, a large burden of caring
for patients who have NSTE ACS is placed
on the emergency department. Therefore, it is
crucial for emergency physicians to institute
evidence-based therapy as quickly, appropriately,
and accurately as possible. Boxes 3 and 4
90 MILLER et al

Box 2. Recommendations for early risk stratification [14]

Class I
A determination should be made in all patients who have chest discomfort of the likelihood
of acute ischemia caused by coronary artery disease as high, intermediate, or low (level of
evidence: C).
Patients who present with chest discomfort should undergo early risk stratification that
focuses on anginal symptoms, physical findings, ECG findings, and biomarkers of cardiac
injury (level of evidence: B).
A 12-lead ECG should be obtained immediately (within 10 minutes) in patients who have
ongoing chest discomfort and as rapidly as possible in patients who have a history of
chest discomfort consistent with ACS but whose discomfort has resolved by the time of
evaluation (level of evidence: C).
Biomarkers of cardiac injury should be measured in all patients who present with chest
discomfort consistent with ACS. A cardiac-specific troponin is the preferred marker, and if
available it should be measured in all patients. Creatinine kinase myocardial band
(CK-MB) by mass assay is also acceptable. In patients who have negative cardiac markers
within 6 hours of the onset of pain, another sample should be drawn in the 6- to 12-hour
time frame (eg, 9 hours after the onset of symptoms) (level of evidence: C).
Class IIa
For patients who present within 6 hours of the onset of symptoms, an early marker of
cardiac injury (eg, myoglobin or CK-MB subforms) should be considered in addition to
a cardiac troponin (level of evidence: C).

Class IIb
C-reactive protein and other markers of inflammation should be measured (level of
evidence: B).
Class III
Total CK (without MB), aspartate aminotransferase, beta-hydroxybutyric dehydrogenase,
and/or lactate dehydrogenase should be the markers for the detection of myocardial injury
in patients who have chest discomfort suggestive of ACS (level of evidence: C).

From Brunwald E, Antman EM, Beasley JW, et al. ACC/AHA guideline update for the management of
patients with unstable angina and non-ST–segment elevation myocardial infarction–2002: a report of the
American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Commit-
tee on the Management of Patients with Unstable Angina). Circulation 2002;106:1893–1900.

highlight the ACC/AHA-recommended ap- with prompt cardiac catheterization, compared

proaches for anti-ischemic, antithrombotic, and
antiplatelet therapies from the 2002 update. Es-
sential also is clear and thorough communication
from emergency medical services to the emer-
gency department and from the emergency
department to the coronary care unit, catheteri-
zation laboratory, or telemetry floor, to curb er-
rors of omission.
Early invasive versus early conservative approach
Randomized clinical trial data collectively
support the use of an early invasive approach
with an initial conservative approach that reserves
cardiac catheterization for patients who develop
recurrent ischemia despite medical therapy (Box 5)
[23,25–27]. The TACTICS-TIMI-18 trial found
that catheterization within the first 48 hours af-
ter presentation was superior to an initial strat-
egy of medical management, particularly in
patients who had elevated troponin levels or el-
evated TIMI risk score [23]. Similarly, The Fast
Revascularization during Instability in Coronary
artery disease trial demonstrated a significant
reduction in long-term mortality with early
invasive management for NSTE ACS [27].
 CARE OF PATIENTS WHO HAVE ACS 91

Box 3. Recommendations for anti-ischemic therapy [14]

Class I
Bed rest with continuous ECG monitoring for ischemia and arrhythmia detection in patients
who have ongoing rest pain (level of evidence: C)
Nitroglycerine, sublingual tablet or spray, followed by intravenous administration, for the
immediate relief of ischemia and associated symptoms (level of evidence: C)
Supplemental oxygen for patients who have cyanosis or respiratory distress; finger-pulse
oximetry or arterial blood gas determination to confirm adequate arterial oxygen
saturation (SaO2 > 90%) and continued need for supplemental oxygen in the presence of
hypoxemia (level of evidence: C)
Morphine sulfate intravenously when symptoms are not immediately relieved with
nitroglycerine or when acute pulmonary congestion and/or severe agitation is present
(level of evidence: C)
A beta-blocker, with the first dose administered intravenously if there is ongoing chest pain,
followed by oral administration, in the absence of contraindications (level of evidence: B)
In patients who have continuing or frequently recurring ischemia when beta-blockers are
contraindicated, a nondihydropyridine calcium antagonist (eg, verapamil or diltiazem) as
initial therapy in the absence of severe left ventricular dysfunction or other
contraindications (level of evidence: B)
An ACE inhibitor when hypertension persists despite treatment with nitroglycerine and
a beta-blocker in patients who have left ventricular systolic dysfunction or heart failure
and in diabetic patients who have ACS (level of evidence: B)

Class IIa
Oral long-acting calcium antagonists for recurrent ischemia in the absence of
contraindications and when beta-blockers and nitrates are fully used (level of evidence: C)
An ACE inhibitor for all post-ACS patients (level of evidence: B)
Intra-aortic balloon pump counterpulsation for severe ischemia that is continuing or recurs
frequently despite intensive medical therapy or for hemodynamic instability in patients
before or after coronary angiography (level of evidence: C)

Class IIb
Extended-release form of nondihydropyridine calcium antagonists instead of a beta-blocker
(level of evidence: B)
Immediate-release dihydropyridine calcium antagonists in the presence of a beta-blocker
(level of evidence: B)
Class III
Nitroglycerine or other nitrate within 24 hours of sildenafil use (level of evidence: C)
Immediate-release dihydropyridine calcium antagonists in the absence of a beta-blocker
(level of evidence: A)

From Brunwald E, Antman EM, Beasley JW, et al. ACC/AHA guideline update for the management of
patients with unstable angina and non-ST–segment elevation myocardial infarction–2002: a report of the
American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Commit-
tee on the Management of Patients with Unstable Angina). Circulation 2002;106:1893–1900.

National acute coronary syndromes
databasesdevidence meets community practice
During the last decade, several large registries
have been created to examine the temporal and
regional management of patients who have ACS
and to provide real-world data on practice and
outcomes. This wealth of data makes it possible to
monitor trends in ACS care and to observe
evidence-based therapy only slowly being adopted
92 MILLER et al

Box 4. Recommendations for antiplatelet and anticoagulation therapy [14]

Class I
Antiplatelet therapy should be initiated promptly. Aspirin (ASA) should be administered as
soon as possible after presentation and continued indefinitely (level of evidence: A).
Clopidogrel should be administered to hospitalized patients who are unable to take ASA
because of hypersensitivity or major gastrointestinal intolerance (level of evidence: A).
In hospitalized patients in whom an early noninterventional approach is planned,
clopidogrel should be added to ASA as soon as possible on admission and administered
for at least 1 month (level of evidence: A) and for up to 9 months (level of evidence: B).
In patients for whom a PCI is planned, treatment with clopidogrel should be started and
continued for at least 1 month (level of evidence: A) and up to 9 months in patients who
are not at high risk for bleeding (level of evidence: B).
In patients taking clopidogrel in whom elective coronary artery bypass graft surgery is
planned, the drug should be withheld for 5 to 7 days before surgery (level of evidence: B).
Anticoagulation with subcutaneous LMWH or intravenous unfractionated heparin (UFH)
should be added to antiplatelet therapy with ASA and/or clopidogrel (level of evidence: A)
A platelet Gp IIb/IIIa antagonist should be administered, in addition to ASA and heparin,
to patients in whom catheterization and PCI are planned. The Gp IIb/IIIa antagonist may
also be administered just before PCI (level of evidence: A).

Class IIa
Eptifibatide or tirofiban should be administered, in addition to ASA and LMWH or UFH, to
patients who have continuing ischemia, an elevated troponin level, or with other high-risk
features in whom an invasive management strategy is not planned (level of evidence: A).
Enoxaparin is preferable to UFH as an anticoagulant in patients who have unstable angina/
NSTEMI, unless coronary artery bypass graft surgery is planned within 24 hours (level
of evidence: A).
A platelet Gp IIb/IIIa antagonist should be administered to patients already receiving
heparin, ASA, and clopidogrel in whom catheterization and PCI are planned. The Gp IIb/IIIa
antagonist may also be administered just prior to PCI (level of evidence: B).

Class IIb
Eptifibatide or tirofiban, in addition to ASA and LMWH or UFH, should be administered to
patients who do not have continuing ischemia, who have no other high-risk features, and
in whom PCI is not planned (level of evidence: A).
Class III
Intravenous fibrinolytic therapy is recommended in patients who do not have acute
ST-segment elevation, a true posterior myocardial infarction, or a presumed new left
bundle-branch block (level of evidence: A).
Abciximab should be administered to patients in whom PCI is not planned (level of
evidence: A).

From Brunwald E, Antman EM, Beasley JW, et al. ACC/AHA guideline update for the management of
patients with unstable angina and non-ST–segment elevation myocardial infarction–2002: a report of the
American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Commit-
tee on the Management of Patients with Unstable Angina). Circulation 2002;106:1893–1900.

into everyday clinical practice. During the obser-
vation period reported by these registries, there
were upward trends in the use of guideline-based
medical therapy (aspirin, beta-blockers, ACE
inhibitors, and antithrombin therapy) and
decreases in time to reperfusion. As guideline-
based therapy was adopted into practice, hospital
mortality fell (from 11.2% in 1990 to 9.4% in
1999) [28]. These registries are a reminder, how-
ever, that there are still areas for significant
 CARE OF PATIENTS WHO HAVE ACS 93

Box 5. Recommendations for early conservative versus invasive treatment strategies [14]

Class I
An early invasive strategy is indicated in patients who have unstable angina/NSTEMI and
any of the following high-risk indicators (level of evidence: A):
Recurrent angina/ischemia at rest or with low-level activities despite intensive anti-ischemic
therapy
Elevated troponin T or troponin I
New or presumably new ST-segment depression
Recurrent angina/ischemia with symptoms of chronic heart failure, an S3 gallop, pulmonary
edema, worsening rales, or new or worsening mitral regurgitation
High-risk findings on noninvasive stress testing
Depressed left ventricular systolic function (eg, ejection fraction < 0.40 on noninvasive
study)
Hemodynamic instability
Sustained ventricular tachycardia
PCI within 6 months
Prior coronary artery bypass graft surgery
In the absence of these findings, either an early conservative or an early invasive strategy
is indicated in hospitalized patients who do not have contraindications for
revascularization (level of evidence: B).
Class IIa
An early invasive strategy is indicated in patients who have repeated presentations for
ACS despite therapy and without evidence for ongoing ischemia or high risk (level of
evidence: C).
Class III
Coronary angiography is indicated in patients who have extensive comorbidities (eg, liver or
pulmonary failure, cancer), in whom the risks of revascularization are not likely to
outweigh the benefits (level of evidence: C).
Coronary angiography is recommended in patients who have acute chest pain and a low
likelihood of ACS (level of evidence: C).
Coronary angiography is indicated in patients who will not consent to revascularization
regardless of the findings (level of evidence: C).

From Brunwald E, Antman EM, Beasley JW, et al. ACC/AHA guideline update for the management of
patients with unstable angina and non-ST–segment elevation myocardial infarction–2002: a report of the
American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Commit-
tee on the Management of Patients with Unstable Angina). Circulation 2002;106:1893–1900.

improvement and many missed opportunities for
intervention in the care of patients who have
ACS. This difference between the recommenda-
tions and practice partially explains why real-
world mortality for ACS still remains much
higher than that reported in randomized clinical
trials.
National Registry of Myocardial Infarction
The NRMI was initiated in July 1990 to provide
participating hospitals a means of tracking the
characteristics, treatment, and outcome of patients
who have AMI. During the 14 years that the
NRMI has been enrolling, more 2.2 million
patients who have myocardial infarction have
been evaluated. The NRMI thus is the largest
observational registry of ACS to date [28,29].
Global Registry of Acute Coronary Events
The GRACE is an ongoing international
registry enrolling patients from 94 hospitals in
14 countries in Europe, North America, South
94 MILLER
America, Australia, and New Zealand [30].
GRACE spans a variety of care facilities, encom-
passes the full range of coronary syndromes, and
as of 2003 has enrolled 39,000 patients. GRACE
highlights the high morbidity and mortality asso-
ciated with hospitalization for any ACS. Death
occurred in 12% of patients who had STEMI, in
13% of those who had NSTEMI, and in 8% of
those who had unstable angina. Thus, among hos-
pitalized patients, the risk of death at 6 months
was just as great among patients who had ACS
presenting with NSTEMI as among those present-
ing with STEMI [30–32].
Get With The Guidelines
The Get With The Guidelines program is
a hospital-based quality improvement initiative
created by the AHA and American Stroke Asso-
ciation targeting secondary prevention efforts.
The largest quality improvement initiative, it has
targeted an initial 1800 hospitals, representing
75% of the patients in the United States who
have AMI, and currently has a database of
more than 80,000 patients [33,34]. A Web-based
management tool is used for data collection and
feedback (data are compatible with the Joint
Commission on Accreditation of Health care Or-
ganizations and NRMI) and provides real-time,
guideline-based, on-line reminders specifically ad-
dressing discharge management. Additionally the
Web tool provides communication with primary
care providers by generating a letter documenting
discharge instructions and medications to enforce
long-term compliance.
The Can Rapid Risk Stratification of Unstable
Angina Patients Suppress Adverse Outcomes
with Early Implementation of the American
College of Cardiology/American Heart
Association guidelines initiative
The national quality improvement initiative
CRUSADE was established to promote aware-
ness and widespread use of the ACC/AHA clinical
treatment guidelines specifically directed at pa-
tients who are at high risk of NSTE ACS [19,35].
Data are being collected regarding all aspects of
NSTE ACS care and will be used for reporting,
feedback, and quality improvement. More than
just a database, CRUSADE is also a quality
improvement project and will educate hospital
staff on all aspects of the ACC/AHA guidelines
before participation and offer hospitals access to
et al
educational and patient care materials (eg, stan-
dardized orders sets, risk-stratification charts, dis-
charge planning forms) with the goal of improving
both acute (emergency department–based) and
longer-term (postdischarge) management of this
population. Moreover, CRUSADE will report
back to a given institution its treatment patterns
compared with national norms, coupling these re-
ports to multispecialty educational efforts by the
steering committee, including lectures and scien-
tific publications of risk-stratification practice
and success. As of mid-2005, more than 140,000
patients from more than 400 emergency depart-
ments and medical centers in 47 states across the
United States had been analyzed.
Real-world patients versus selected trial
populations
Registry data show that the demographics of the
patients who have ACS are changing; clinicians
now more frequently are faced with higher risk and
more comorbidities in patients presenting with
NSTE ACS. For example, during the 1990s the
mean patient age in the NRMI registry increased
from 65.3 to 68.0 years [28]. Furthermore, the pro-
portion of women rose from 35.3% to 39.3%, and
there is an increased incidence of associated comor-
bidities including heart failure, stroke, diabetes,
hypertension, and previous coronary revasculariza-
tion. There was also a decrease in patients present-
ing with STEMI (from 36.4% in 1994 to 27.1% in
1999) and an increase in the prevalence of NSTEMI
(from 45% in 1994 to 63% in 1999) [28].
Although some would argue that this demo-
graphic shift makes treatment decisions more
difficult by bringing to medical attention patients
who would have less often been enrolled in clinical
trials (if not excluded altogether), it is clear from
both clinical trials and observational studies that
these high-risk patients derive the most benefit
from the early invasive approach. Also, this
demographic shift alone does not explain the clear
gaps in treatment between certain geographical
areas, institution types, and among certain patient
populations.
Real world management: risk stratification
Despite recommendations by the ACC/AHA
for prompt evaluation and obtainment of an
ECG, this guideline recommendation is met only
one third of the time according to the CRUSADE
quality improvement initiative [36]. Quality im-
provement measures focusing on the areas of
 CARE OF PATIENTS WHO HAVE ACS
timely access to care and providers may include
additional nurses or ECG technicians during
peak emergency department hours, synchronizing
emergency department and ECG machine clocks
to assess actual timing of evaluation more accu-
rately, training additional staff in ECG acquisi-
tion, and educating triage nursing staff about
atypical symptoms that could be consistent with
cardiac ischemia.
Biomarkers of cardiac injury should be mea-
sured in all patients who present with chest
discomfort consistent with ACS; a cardiac-specific
troponin is the preferred marker. It is also
recommended that biomarker results be available
to the treating physician within 30 to 60 minutes
after admission. Delays in biomarker turnaround
are unfortunately common in clinical practice. In
a study of 1340 patients at 25 CRUSADE
hospitals, the overall mean ‘‘vein to brain’’ time
(time from arrival in the emergency department as
recorded in the medical chart or ambulance
arrival data to notification of the treating physi-
cian of laboratory results, VTBT) was 115.7 G
70.1 minutes. The use of point-of-care testing
rather than a central laboratory was associated
with a significantly shorter VTBT (68.2 G 40.8
minutes) but was used in a minority of the
emergency departments. CRUSADE also found
that patients in hospitals with the shortest VTBT
were more likely to receive recommended evi-
denced-based therapies including aspirin, beta-
blockers, clopidogel, GpIIbIIIa and heparin
(Peacock WF, Roe MT, Chen AY, et al. Vein-
to-brain time: an emergency department quality
of care marker for non–ST-segment elevation
acute coronary syndromes, unpublished article).
Real world management: acute treatment
for acute coronary syndromes
Despite guideline recommendations, acute
medical therapies are often underused, especially
in populations at highest risk. Registry data
report that early in the 1990s acute therapies
were grossly underused. During the latter half of
the decade practices were becoming increasingly
concordant with guideline recommendations but
still had significant room for improvement. This
trend continues into the current decade. Fig. 1 il-
lustrates the trend in select acute therapies in the
NRMI registry (in both patients who have
STEMI and those who have NSTEMI) between
1994 and 2003. Table 1 illustrates usage rates of
acute therapies in eligible patients who have
95
NSTEMI and in-hospital mortality from the
GRACE and CRUSADE registries.
Beyond the global trends in care, these regis-
tries give insights into which patient populations
are failing to receive evidence-based care. A
disturbing trend has been noted in analysis of
many registries, namely that patients who present
with ACS and are at the highest risk are para-
doxically treated less aggressively than those who
are at the lowest risk. Additionally, there are
significant disparities in treatment based on age,
race, sex, and insurance status with the elderly,
nonwhites, women, and the uninsured least likely
to receive guideline-based therapy [37–41]. A re-
cent study that illustrates this trend examined pat-
terns of use of Gp IIb/IIIa inhibitors in the early
management of NSTEMI patients using the
NRMI 4 registry at 1189 centers across the United
States [41]. The investigators noted that of 60,770
patients who had NSTEMI and who were eligible
for early Gp IIb/IIIa inhibitor therapy, only
15,379 patients (25%) received such therapy. Pa-
tients receiving early Gp IIb/IIIa inhibitor treat-
ment were significantly younger and were more
likely to be white and male and to have private in-
surance (instead of Medicare, Medicaid, or self-
pay). Treated patients were less likely to have
had a prior myocardial infarction or most comor-
bid illnesses but were more likely to have had
a coronary revascularization procedure. The over-
all in-hospital mortality rate was 8.0% for all pa-
tients, but patients treated early with Gp IIb/IIIa
inhibitors had significantly lower unadjusted mor-
tality than did patients not receiving such therapy
(3.3% versus 9.6%; P ! .001). Patients who had
higher baseline risk were also less likely to receive
early Gp IIb/IIIa inhibitor treatment. For exam-
ple, whereas 45% of low-risk patients (expected
risk ! 2%) received early Gp IIb/IIIa inhibition,
only 9% of the high-risk patients (expected mor-
tality O 15%) received treatment. This pattern
of conservative care in high-risk patients is also
seen in other areas of medicine but is inconsistent
with maximizing absolute benefits from therapy
[41–43].
Similar patterns can be seen in the use of early
invasive catheterization in patients who have
ACS. Using patients enrolled in the CRUSADE
registry between March 2000 and September
2002, Bhatt and colleagues [44] found that the un-
adjusted incidence of in-hospital mortality was
2.0% for patients who underwent early invasive
management (within 48 hours) compared with
6.2% for patients who did not undergo early
96 MILLER et al

Fig. 1. Medication range in the first 24 hours from the NRMI registry (STEMI and NSTEMI patients. (From Gibson
MC. NRMI and current treatment patterns for ST-elevation myocardial infarction. Am Heart J 2004;148:S31; with
permission.)

invasive management. Patients who underwent patients as well as by errors of commission or

early catheterization were younger, more often failure to deliver treatments in a safe and appro-
male and white, more likely to be admitted to priate manner. Failure in any of these areas can
a cardiology service, and less likely to have lead to poor patient outcomes and high cost [46–
heart failure or renal insufficiency. Additionally, 50]. The Institute of Medicine’s report Crossing
when stratified into low, medium, and high clin- the Quality Chasm outlines six aims for
ical risk based on the modified PURSUIT risk
score (Platelet Glycoprotein IIb/IIIa in Unstable Table 1
Angina: Receptor Suppression Using Integrilin Observational studies describing care and outcomes in
NSTE ACS [1,31]
Therapy [PURSUIT Trial]), patients in all three
risk categories undergoing early invasive manage- Study
ment had a significantly lower risk of unadjusted GRACE CRUSADE
in-hospital mortality [44,45]. The patients at high- 1999–2000 2001–2002
est risk seemed to derive the greatest absolute Care/outcomea (N ¼ 2893) (N ¼ 18937)
benefit from early invasive management. Propen- Aspirin (%) 91 90
sity matching of patients by early invasive man- Beta-blockers (%) 78 76
agement status produced groups that were Heparin (%) 61 53
similarly matched for clinical, demographic, and Low molecular weight
hospital characteristics, and in this sample of pro- heparin (%) 51 36
pensity-matched pairs the frequency of in-hospital Gp IIbIIIa inhibitors (%) 20 31
mortality was lower in patients who underwent Inhospital mortality (%) 6.0 4.9
a
early invasive management (2.5% versus 3.7%; Treatments listed reflect usage patterns during hos-
P ! .001). pitalization for eligible patients who do not have possi-
ble contraindications to the given treatments at the end
of the respective evaluation periods.
Concepts of quality and quality improvement Data from Roe MT, Ohman EM, Pollack CV Jr,
et al. Changing the model of care for patients with acute
As evidenced by the data presented here and coronary syndromes. Am Heart J 2003;146:605–12; Fox
detailed in several recent prominent reports, KA, Goodman SG, Klein W, et al. Management of acute
American medicine is challenged by errors of coronary syndromes. Variations in practice and out-
omission or failure to ensure that evidence-based come; findings from the Global Registry of Acute Coro-
therapeutic and preventive measures reach nary Events (GRACE). Eur Heart J 2002;23:1177–89.
 CARE OF PATIENTS WHO HAVE ACS
improvement in the United States health care sys-
tem. For the treatment of ACS, these goals can be
translated as follows:
Timelinessdrapid risk stratification and
treatment
Effectivenessdproviding the right evidence-
based treatment
Safetydassuring that therapies, when given,
are administered correctly with an emphasis
on appropriate drug dosing and procedures
done correctly
Equitydassuring all patients, regardless of
race, sex, socioeconomic or insurance status,
receive safe and effective care
Efficiencydavoiding overtreatment in those
who do not stand to benefit
Patient centerednessdconsidering a treat-
ment’s risks and benefits in an individual pa-
tient and according to the patient’s value
system
Improving practice patterns: barriers to
evidence-based care
Cabana and colleagues [51] conducted a litera-
ture review to determine the barriers to adoption
of practice guidelines in clinical practice. The bar-
riers they identified included lack of awareness,
lack of familiarity, lack of agreement, lack of
self-efficacy, lack of outcome expectancy, inertia
of previous practice, and external barriers. Of
these barriers, lack of awareness and lack of famil-
iarity are best remedied by educational initiatives,
whereas lack of self-efficacy and lack of outcome
expectancy are best remedied by providing contin-
uous feedback on guideline adherence and patient
outcomes data, respectively. Unfortunately, inter-
ventions designed to enhance physician education
such as continuing medical education conferences
and printed materials have been shown to have lit-
tle impact on improving physician performance
[35,52,53].
Using the data from the NRMI, Bradley and
colleagues [54] attempted to identify factors present
in hospitals with higher compliance to guideline-
based therapy, specifically the use of beta-blockers
in the treatment of AMI. In an interesting and
hypothesis-generating study using open-ended
interviews of clinical staff and administrators, the
investigators noted four themes that were prevalent
among the hospitals with the greatest or increasing
use of beta-blockers (O65%) versus those with
least or declining use (!65%): (1) a high degree
97
of commitment among clinicians and support staff
regarding the use of evidence-based therapies,
(2) a substantial level of administrative support
for quality improvement, (3) strong physician lead-
ership or a physician champion, and (4) high-qual-
ity data feedback. Additional factors that have
been reported in the literature to improve adher-
ence to practice guidelines include reminder
systems such as critical care pathways or computer-
ized support programs, patient-oriented interven-
tions, and the use of local opinion leaders in the
education of physicians [55,56].
Cycle of continuous quality improvement
The process of continuous quality improve-
ment begins with the publication of CPGs, which
are generated by expert committees and based on
assimilation of clinical-trial results [1]. Data must
then be collected to determine rates of use of ther-
apies and interventions recommended by CPGs,
and performance indicators are developed to es-
tablish benchmarks for high-quality care. Next,
multidisciplinary teams led by local opinion lead-
ers must develop and implement plans to improve
adherence to practice guidelines. Finally, patient
outcomes based upon performance and adherence
to practice guidelines must be measured to en-
courage continuous improvements in patient
care (Fig. 2) [1,57]. As discussed previously, signif-
icant challenges limit the success of quality im-
provement initiatives designed to encourage the
adoption of CPGs. Clearly an active, multidisci-
plinary, and multifactorial approach is needed to
integrate CPGs into clinical practice, as demon-
strated through the various successful AMI qual-
ity improvement initiatives outlined in Table 2.
Fig. 2. The cycle of clinical therapeutics. (Adapted from
Califf RM, Peterson ED, Gibbons RJ, et al. Integrating
quality into the cycle of therapeutic development. J Am
Coll Cardiol 2002;40:1896; with permission.)
 98
Table 2
Quality improvement initiatives in patients who have myocardial infarction
Improvements
Initial Data in guideline

MILLER
Program Population Intervention Collection Re-measurement adherence Mortality
Cardiac UCLA w 300 patients Protocols and 1992–1993 1994–1995 Aspirin (78% to 92%) Reduction in death and

et al
Hospitalization per collection time standing orders Beta-blocker recurrent myocardial
Arteriosclerosis focusing on secondary Retrospective (12% to 61%) infarction from 14.8%
Management prevention at discharge data collection ACE inhibitors to 6.4%
Program Quarterly feedback (4% to 56%)
(CHAMP) [3] to clinicians Lipid-lowering therapy
(6% to 86%)
Cooperative Medicare patients in Feedback of 1992–1993 1995–1996 Thrombolytics within 10% relative reduction in
Cardiovascular four states performance 1 hour (57.1% short-term (30-day) and
Project Pilot [59] discharged with Opinion leaders to 70.8%) longer-term (1-year)
diagnosis of MI, Thrombolytics mortality
focusing on both within 30 minutes
acute and discharge (17.6% to 30.1%)
therapies Smoking cessation
counseling
(28.6% to 41%)
Aspirin (83.6% to 90.3%)
Beta-blockers
(47.5% to 68.4%)
ACE inhibitors
(48.5% to 62.2%)
Guidelines Random sample of A tool kit for 1998–1999 2001 No significant Not reported
Applied in Medicare and clinicians and difference in time
Practice non-Medicare patients included to thrombolysis
(GAP) [60] patients presenting care maps, standing Aspirin (84% to 92%)
with MI in Michigan, admission orders, Beta-blockers
focusing on acute and and discharge forms. (89% to 93%)
discharge therapies Grand rounds site ACE inhibitors
visits Physician (80% to 86%)
and nurse opinion Smoking cessation
leaders counseling (53% to 65%)
Lipid-lowering therapy
(68 to 75%)
Get With The 1800 hospitals A web-based 2000 Ongoing Significant Not reported
Guidelines representing 75% of management improvements in
(GWTG) [33,34] the MI patients. tool is used for data Smoking cessation

CARE OF PATIENTS WHO HAVE ACS

Over 80,000 patients collection and counseling
enrolled. Pilot data feedback (48% to 87%)
from 24 hospitals has and provides Lipid lowering therapy
been reported real-time, (54% to 79%)
(O1700 patients) guideline based, LDL measurement
on-line, (59% to 81%)
reminders BP control,
specifically Blood pressure less
addressing discharge than 140/90
management. (60% to 68%)
Cardiac rehabilitation
referral (34% to 73%)

99
100 MILLER
Linking quality of care to outcomes
There is strong emerging evidence that improv-
ing hospital guideline compliance is associated with
improved patient outcomes. Peterson and col-
leagues [58] used NRMI IV (2000–2002) data to re-
view the care of more than 250,000 patients who
had AMI from more than 1200 institutions. A mea-
surement of ‘‘composite quality’’ was calculated by
reviewing each patient’s care, and scores were given
for the correct application of the guideline-based
therapy for which the patient was eligible. Institu-
tions were ranked according to their composite
quality scores and divided into quartiles. In the
leading quartile (best care), the average inpatient
mortality was 8.3%. In contrast, patients treated
in the lagging quartile hospitals (worst care) had
an inpatient mortality rate of 15.3%.
This association between the degree of adher-
ence to ACC/AHA guidelines and better patient
outcomes has been duplicated using data from the
CRUSADE database [18]. Again, more than 400
hospitals in the United States were grouped into
quartiles based on reported overall adherence to
ACC/AHA guidelines. Hospitals with more than
80% adherence to these guidelines (leading quar-
tile) were compared with hospitals that demon-
strated less than 65% adherence to the
guidelines (lagging quartile). As in the NRMI,
the in-hospital mortality rate was 3.6% for lead-
ing-quartile hospitals, versus 5.9% for hospitals
in the lagging quartile. This difference in mortality
indicates that adherence to the ACC/AHA guide-
lines does indeed seem to result in improved out-
comes for patients [18].
Summary
Both acute management and secondary pre-
vention for patients presenting with the spectrum
of ACS have evolved greatly during the last
decade, as evidenced by the multitude of clinical
trials and the development of CPGs. The goal of
the next decade is to ensure the accurate, equal,
and timely application of these therapies and
management strategies in clinical practice. In the
emergency department, initiation of guideline-
based management is especially challenging given
the dynamic process of risk stratification that
must take place to ensure properly directed care.
It is clear, however, that application of such
therapies leads to improved outcomes. Lessons
learned from previous and ongoing quality im-
provement initiatives will provide the tools needed
et al
to ensure that widespread adoption of guideline-
based therapy is complete.
References
[1] Roe MT, Ohman EM, Pollack CV Jr, et al. Chang-
ing the model of care for patients with acute coro-
nary syndromes. Am Heart J 2003;146:605–12.
[2] French WJ. Trends in acute myocardial infarction
management: use of the National Registry of Myo-
cardial Infarction in quality improvement. Am J
Cardiol 2000;85:5B–12B.
[3] Fonarow GC, Gawlinski A, Moughrabi S, et al. Im-
proved treatment of coronary heart disease by
implementation of a cardiac hospitalization athero-
sclerosis management program (CHAMP). Am J
Cardiol 2001;87:819–22.
[4] Braunwald E, Jones RH, Mark DB, et al. Diagnos-
ing and managing unstable angina. Agency for
Health Care Policy and Research. Circulation
1994;90:613–22.
[5] Braunwald E, Antman EM, Beasley JW, et al.
ACC/AHA guidelines for the management of
patients with unstable angina and non-ST-seg-
ment elevation myocardial infarction: a report of
the American College of Cardiology/American
Heart Association Task Force on Practice Guide-
lines (Committee on the Management of Patients
with Unstable Angina). J Am Coll Cardiol 2000;
36:970–1062.
[6] McCarthy M. US heart-guidelines strategy makes
promising start. Lancet 2001;358:1618.
[7] Pollack CV Jr, Roe MT, Peterson ED. 2002 update
to the ACC/AHA guidelines for the management of
patients with unstable angina and non-ST-segment
elevation myocardial infarction: implications for
emergency department practice. Ann Emerg Med
2003;41:355–69.
[8] Pollack CV Jr, Diercks DB, Roe MT, et al. 2004
ACC/AHA guidelines for the management of
patients with ST-elevation myocardial infarction:
implications for ED practice. Ann Emerg Med
2005;45:363–76.
[9] Gibler WB, Cannon CP, Blomkalns AL, et al. Prac-
tical implementation of the guidelines for unstable
angina/non-ST-segment elevation myocardial in-
farction in the emergency department. Circulation
2005;111:2699–710.
[10] American Heart Association. 2002 Heart and stroke
statistical update. Dallas (TX): American Heart As-
sociation; 2001.
[11] Nouraj P. National hospital ambulatory medical
care survey: 1997 emergency department summary.
Hyattsville (MD): National Center for Health Statis-
tics; 1997.
[12] American Heart Association. Heart disease and
stroke statisticsd2005 update. Dallas (TX): Ameri-
can Heart Association; 2005.
 CARE OF PATIENTS WHO HAVE ACS
[13] Antman EM, Anbe DT, Armstrong PW, et al. ACC/
AHA guidelines for the management of patients
with ST-elevation myocardial infarction: executive
summary: a report of the ACC/AHA Task Force
on Practice Guidelines (Committee to Revise the
1999 Guidelines on the Management of Patients
With Acute Myocardial Infarction). J Am Coll Car-
diol 2004;44:671–719.
[14] Braunwald E, Antman EM, Beasley JW, et al. ACC/
AHA guideline update for the management of
patients with unstable angina and non-ST-segment
elevation myocardial infarction – 2002: a report of
the American College of Cardiology/American
Heart Association Task Force on Practice Guide-
lines (Committee on the Management of Patients
with Unstable Angina). Circulation 2002;106:
1893–900.
[15] Gibson CM. Time is myocardium and time is out-
comes. Circulation 2001;104(22):2632–4.
[16] Brown AL, Mann NC, Daya M, et al, for the Rapid
Early Action for Coronary Treatment (REACT)
study. Demographic, belief, and situational factors
influencing the decision to utilize emergency medical
services among chest pain patients. Circulation 2000;
102:173–8.
[17] Faxon D, Lenfant C. Timing is everything: motivat-
ing patients to call 9-1-1 at onset of acute myocardial
infarction. Circulation 2001;104:1210–1.
[18] Peterson ED, Roe MT, Lytle BL, et al. The associa-
tion between care and outcomes in patients with
acute coronary syndromes: national results from
CRUSADE. J Am Coll Cardiol 2004;43:406A.
[19] Ohman EM, Roe MT, Smith SC Jr, et al. Care of the
non-ST-segment elevation patients: insights from
the CRUSADE national quality improvement ini-
tiative. Am Heart J 2004;148:S34–9.
[20] Antman EM, Cohen M, Bernink PJ, et al. The TIMI
risk score for unstable angina/non-ST elevation MI:
a method for prognostication and therapeutic deci-
sion making. JAMA 2000;284:835–42.
[21] Antman EM, McCabe CH, Gurfinkel EP, et al.
Enoxaparin prevents death and cardiac ischemic
events in unstable angina/non-Q-wave myocardial
infarction: results of the Thrombolysis In Myocar-
dial Infarction (TIMI) 11B trial. Circulation 1999;
100:1593–601.
[22] Cohen M, Demers C, Gurfinkel EP, et al, for the Ef-
ficacy and Safety of Subcutaneous Enoxaparin in
Non-Q-Wave Coronary Events Study Group. A
comparison of low-molecular-weight heparin with
unfractionated heparin for unstable coronary artery
disease. N Engl J Med 1997;337:447–52.
[23] Cannon CP, Weintraub WS, Demopoulos LA, et al.
TACTICS (Treat Angina with Aggrastat and Deter-
mine Cost of Therapy with an Invasive or Con-
servative Strategy)–Thrombolysis in Myocardial
Infarction 18 Investigators. Comparison of early in-
vasive and conservative strategies in patients with
unstable coronary syndromes treated with the
101
glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J
Med 2001;344(25):1879–87.
[24] Platelet Receptor Inhibition in Ischemic Syndrome
Management in Patients Limited by Unstable Signs
and Symptoms (PRISM-PLUS) Study investigators.
Inhibition of the platelet glycoprotein IIb/IIIa recep-
tor with tirofiban in unstable angina and non-Q-
wave myocardial infarction. N Engl J Med 1998;
338:1488–97.
[25] The TIMI IIIB Investigators. Effects of tissue plas-
minogen activator and a comparison of early inva-
sive and conservative strategies in unstable angina
and non-Q-wave myocardial infarction: results of
the TIMI IIIB trial. Thrombolysis In Myocardial Is-
chemia. Circulation 1994;89:1545–56.
[26] Fragmin and Fast Revascularisation during Insta-
bility in Coronary artery disease investigators. Inva-
sive compared with non-invasive treatment in
unstable coronary-artery disease: FRISC II prospec-
tive randomised multicentre study. Lancet 1999;354:
708–15.
[27] Wallentin L, Lagerqvist B, Husted S, et al. FRISC II
investigators. Outcome at 1 year after an invasive
compared with a non-invasive strategy in unstable
coronary-artery disease: the FRISC II invasive rand-
omised trial. Lancet 2000;356:9–16.
[28] Rogers WJ, Canto JG, Lambrew CT, et al. Tempo-
ral trends in the treatment of over 1.5 million
patients with myocardial infarction in the US from
1990 through 1999: the National Registry of Myo-
cardial Infarction 1, 2 and 3. J Am Coll Cardiol
2000;36(7):2056–63.
[29] Gibson MC. NRMI and current treatment patterns
for ST-elevation myocardial infarction. Am Heart J
2004;148:S29–33.
[30] GRACE investigators. Rationale and design of the
GRACE (Global Registry of Acute Coronary
Events) project: a multinational registry of patients
hospitalized with acute coronary syndromes.
Am Heart J 2001;141:190–3.
[31] Fox KA, Goodman SG, Klein W, et al. Manage-
ment of acute coronary syndromes. Variations in
practice and outcome; findings from the Global
Registry of Acute Coronary Events (GRACE).
Eur Heart J 2002;23:1177–89.
[32] Eagle KA, Goodman SG, Avezum A, et al. Practice
variation and missed opportunities for reperfusion
in ST–segment-elevation myocardial infarction:
findings from the Global Registry of Acute Coro-
nary Events (GRACE). Lancet 2002;359:373–7.
[33] LaBresh KA, Ellrodt AG, Gliklich R, et al. Get With
the Guidelines for cardiovascular secondary preven-
tion: pilot results. Arch Intern Med 2004;164(2):
203–9.
[34] Smaha LA. The American Heart Association Get
With the Guidelines Program. Am Heart J 2004;
148:S46–8.
[35] Hoekstra JW, Pollack CV Jr, Roe MT, et al. Im-
proving the care of patients with non-ST-elevation
102 MILLER
acute coronary syndromes in the emergency depart-
ment: the CRUSADE initiative. Acad Emerg Med
2002;9(11):1146–55.
[36] Diercks DB, Roe MT, Chen AY, et al. Disparities by
sex in timing of initial electrocardiogram for patients
presenting with acute coronary syndromes. J Am
Coll Cardiol 2005;45(3):221A.
[37] Shahi CN, Rathore SS, Wang Y, et al. Quality of
care among elderly patients hospitalized with unsta-
ble angina. Am Heart J 2001;142:263–70.
[38] Scirica BM, Moliterno DJ, Every NR, et al. Differ-
ences between men and women in the management
of unstable angina pectoris (the GUARANTEE reg-
istry). Am J Cardiol 1999;84:1145–50.
[39] Stone PH, Thompson B, Anderson HV, et al. Influ-
ence of race, sex, and age on management of unsta-
ble angina and non–Q-wave myocardial infarction:
the TIMI III registry. JAMA 1996;275:1104–12.
[40] Giugliano RP, Camargo CA Jr, Lloyd-Jones DM,
et al. Elderly patients receive less aggressive medical
and invasive management of unstable angina: poten-
tial impact of practice guidelines. Arch Intern Med
1998;158:1113–20.
[41] Peterson ED, Pollack CV Jr, Roe MT, et al.
National Registry of Myocardial Infarction
(NRMI) 4 investigators. Early use of glycoprotein
IIb/IIIa inhibitors in non-ST-elevation acute myo-
cardial infarction: observations from the
National Registry of Myocardial Infarction 4.
J Am Coll Cardiol 2003;42(1):45–53.
[42] Batchelor WB, Peterson ED, Mark DB, et al. A
comparison of US and Canadian cardiac catheteri-
zation practices in detecting severe coronary artery
disease after myocardial infarction: efficiency, yield
and long-term implications. J Am Coll Cardiol
1999;34:12–9.
[43] Pilote L, Califf RM, Sapp S, et al, for the GUSTO-I
investigators. Regional variation across the United
States in the management of acute myocardial in-
farction. N Engl J Med 1995;333:565–72.
[44] Bhatt DL, Roe MT, Peterson ED, et al, for the
CRUSADE investigators. Utilization of early inva-
sive management strategies for high-risk patients
with non-ST-segment elevation acute coronary syn-
dromes: results from the CRUSADE Quality
Improvement Initiative. JAMA 2004;292(17):
2096–104.
[45] Boersma E, Pieper KS, Steyerberg EW, et al, for the
PURSUIT investigators. Predictors of outcome in
patients with acute coronary syndromes without
persistent ST-segment elevation: results from an in-
ternational trial of 9461 patients. Circulation 2000;
101:2557–67.
[46] Zeidel ML. Improving the quality of health care in
America: what medical schools, leading medical
et al
journals and federal funding agencies can do. Am J
Med 2002;112(2):165–7.
[47] Chassin MR. Improving the quality of care. N Engl J
Med 1996;335:1060–3.
[48] Kohn LT, Corrigan JM, Donaldson MS. To err is
human: building a safer health system. Washington
(DC): National Academy Press; 2000.
[49] Allison JJ, Kiefe CI, Weissman NW, et al. Relation-
ship of hospital teaching status with quality of care
and mortality for Medicare patients with acute MI.
JAMA 2000;284:1256–62.
[50] Institute of Medicine. Crossing the quality chasm:
a new health system for the 21st century. Washington
(DC): National Academy Press; 2001.
[51] Cabana MD, Rand CS, Powe NR, et al. Why don’t
physicians follow clinical practice guidelines? A
framework for improvement. JAMA 1999;282:
1458–65.
[52] Davis DA, Thomson MA, Oxman AD, et al. Chang-
ing physician performance: a systematic review of
the effect of continuing medical education. JAMA
1995;274:700–5.
[53] Grol R. Improving the quality of medical care.
Building bridges among professional pride, payer
profit, and patient satisfaction. JAMA 2001;284:
2578–85.
[54] Bradley EH, Holmboe ES, Mattera JA, et al. A qual-
itative study of increasing beta-blocker use after
myocardial infarction: why do some hospitals suc-
ceed? JAMA 2001;285(20):2604–11.
[55] Soumerai SB, McLaughlin TJ, Gurwitz JH, et al.
Effect of local medical opinion leaders on quality
of care for acute myocardial infarction: a ran-
domized controlled trial. JAMA 1998;279:1358–63.
[56] Cooperative Cardiovascular Project Best Practices
Working Group. Improving care for acute myocar-
dial infarction: experience from the Cooperative
Cardiovascular Project. Jt Comm J Qual Improv
1998;24:480–90.
[57] Califf RM, Peterson ED, Gibbons RJ, et al. Inte-
grating quality into the cycle of therapeutic develop-
ment. J Am Coll Cardiol 2002;40:1895–901.
[58] Peterson ED, Parsons LS, Pollack CV, et al. Varia-
tion in AMI care quality across 1,085 hospitals and
its association with hospital mortality rates. Circula-
tion 2002;106:II-722.
[59] Marciniak TA, Ellerbeck EF, Radford MJ, et al. Im-
proving the quality of care for Medicare patients
with acute myocardial infarction: results from the
Cooperative Cardiovascular Project. JAMA 1998;
279:1351–7.
[60] Mehta RH, Montoye CK, Gallogly M, et al. Im-
proving the quality of care for acute myocardial in-
farction: the Guidelines Applied in Practice (GAP)
initiative. JAMA 2002;287:1269–76.
 Cardiol Clin 24 (2006) 103–114

Evaluation and Management of the Patient Who Has

Cocaine-associated Chest Pain
Judd E. Hollander, MDa,*, Timothy D. Henry, MDb
a
Department of Emergency Medicine, University of Pennsylvania, Ground Floor, Ravdin Building,
3400 Spruce Street Philadelphia, PA 19104-4283, USA
b
Minneapolis Heart Institute Foundation, University of Minnesota, 920 East 28th Street,
Suite 40, Minneapolis, MN 55407, USA

Erythroxylon coca, the shrub from which co-
caine is naturally derived, grows indigenously in
South America. Cocaine was first identified as
the active alkaloid in the coca leaf in 1857. As
far back as the twelfth century, Incas used co-
caine-filled saliva as local anesthesia for ritual
trephinations [1]. In 1884, it was recognized med-
ically as a local anesthetic [2]. In the early twenti-
eth century, cocaine was used briefly as an
ingredient in Coca-Cola. In 1906, the United
States began to control cocaine use, and in 1914
the Harrison Narcotic Act labeled cocaine as
a narcotic. It became a schedule II drug in the
1970s. During the last several decades, recreation-
al cocaine use has increased, and reports of side
effects have grown exponentially. As of 2003,
34.9 million citizens of the United States
(14.7%) have used cocaine at least once, with
2.3 million citizens using cocaine within the past
month [3].
Pharmacology
Cocaine is absorbed through application to the
mucosa, ingestion, inhalation, and direct intrave-
nous injection. Effects from nasal insufflation
begin rapidly with peak concentrations typically
reached within 30 to 60 minutes. Intravenous
and inhalational routes of cocaine use produce
near-immediate distribution throughout the cir-
culation. ‘‘Crack’’ is the direct precipitate of
* Corresponding author.
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.003
free-base cocaine that results from alkalinization
of aqueous cocaine hydrochloride.
The relative contributions of cocaine and its
metabolites to the clinical effects remain some-
what unclear. Cocaine is hydrolyzed rapidly by
liver and plasma esterases to ecgonine methylester
(EME), which accounts for 30% to 50% of the
parent product. Nonenzymatic hydrolysis results
in the formation of the other major metabolite,
benzoylecgonine (approximately 40% of the par-
ent product). The biologic half-life of cocaine is
0.5 to 1.5 hours; Benzoylecgonine and EME, the
major metabolites of cocaine, have half-lives of 5
to 8 hours and 3.5 to 6 hours, respectively [4].
Minor metabolites, norcocaine and ecgonine,
account for the majority of the other degradation
products. Early studies suggested that cocaine and
norcocaine accounted for majority of the vascular
effects of cocaine [5]. Recent studies, however,
demonstrate an active role for many of the metab-
olites. Most studies suggest that cocaine and nor-
cocaine are the most potent vasoconstrictors;
benzoylecgonine and ecgonine have less of an ef-
fect. Some studies suggest that EME may result
in mild cerebral vasodilation [6,7]. Sodium-chan-
nel antagonist effects occur with cocaine and nor-
cocaine [8]. Sodium-channel antagonist effects do
not occur with benzoylecgonine or EME [8].
Cocaethylene is a unique metabolite that results
from the combined use of alcohol and cocaine [9].
In clinical studies, cocaethylene produces hemody-
namic effects comparable to those of cocaine. Co-
caethylene has a direct myocardial depressant
effect [10] that is independent of any coronary
artery vasoconstriction [11]. The permeability of
cardiology.theclinics.com
104 HOLLANDER & HENRY
human endothelial cells to low-density lipopro-
teins is increased by both cocaine and cocaethy-
lene, potentially suggesting a mechanism for the
accelerated atherosclerosis seen with cocaine [12].
Pathophysiology
Cocaine has diverse actions in humans. It
directly blocks fast sodium channels, stabilizing
the axonal membrane, with a resultant local
anesthetic effect. Blockade of myocardial fast
sodium channels causes cocaine to have type I
antidysrhythmic properties [13,14]. Cocaine inter-
feres with the uptake of neurotransmitters at the
nerve terminal. Cocaine functions as a vasocon-
strictive agent. These three properties account
for most of the toxicity seen in the clinical setting.
The initial effect on the cardiovascular system
is a transient bradycardia, secondary to stimula-
tion of the vagal nuclei. Tachycardia typically
ensues, predominantly from increased central
sympathetic stimulation. Cocaine has a cardiosti-
mulatory effect through sensitization to epineph-
rine and norepinephrine. It prevents neuronal
reuptake of these catecholamines and increases
the release of norepinephrine from adrenergic
nerve terminals, leading to enhanced sympathetic
effects. The vasopressor effects of cocaine are
mostly mediated by norepinephrine of sympa-
thetic neural origin, and the tachycardiac effects
of cocaine are mostly mediated by epinephrine of
adrenal medullary origin [15].
The pathophysiology of cocaine-associated
myocardial ischemia is multifactorial. Chronic
cocaine users develop left ventricular hypertrophy,
premature atherosclerosis, and coronary aneur-
ysms and ectasia [16]. Acutely, cocaine causes cor-
onary arterial vasoconstriction, in situ thrombus
formation, platelet aggregation, and increased
myocardial oxygen demand. The combination of
increased myocardial oxygen demand (from hy-
pertension, tachycardia, and left ventricular hy-
pertrophy) in the setting of decreased blood flow
(from atherosclerosis, platelet aggregation, and
thrombus formation) and coronary vasoconstric-
tion results in myocardial ischemia.
Cocaine produces coronary vasoconstriction of
the epicardial coronaries that can be reversed by
phentolamine, an alpha-adrenergic antagonist [17]
and exacerbated by propranolol, a beta-adrenergic
antagonist [18]. Cocaine-induced vasoconstriction
occurs in both diseased and nondiseased coronary
artery segments; however, the magnitude of the
effect is more pronounced in the diseased segments
[19]. Tobacco smoking induces coronary artery
vasoconstriction through an alpha-adrenergic me-
chanism similar to that of cocaine [20]. Most
cocaine-using patients are also cigarette smokers,
and the combination of near-simultaneous to-
bacco and cocaine use has a synergistic effect on
epicardial coronary artery vasoconstriction [21].
Cocaine also impedes microvascular blood flow
in the heart [22].
Platelet activation and thrombus formation
also occur secondary to cocaine. Cocaine activates
platelets directly [23] and indirectly through an al-
pha-adrenergic–mediated increase in platelet ag-
gregability [24]. Adenosine diphosphate–induced
platelet aggregation is enhanced [25], and tissue
plasminogen activator inhibitor is increased [26]
in the presence of cocaine. Thrombus formation
can occur in patients who have cocaine-associated
myocardial infarction (MI) whether or not under-
lying coronary artery disease is present [2,22].
Chronic users of cocaine may be prone to early
atherosclerosis, as demonstrated by autopsy stud-
ies of young cocaine users [27–29]. Although ini-
tial reports of cocaine-associated MI emphasized
a lower-than-expected prevalence of coronary ar-
tery disease, many, if not most, patients who
have cocaine-associated MI have underlying coro-
nary artery disease. Large clinical series have
found that 31% to 67% of patients who have co-
caine-associated MI have atherosclerotic coronary
artery disease [2,16,22,27]. Coronary artery dis-
ease is less common in patients who have MI asso-
ciated with cocaine than in cocaine-free patients;
however, it is still more common than in controls
[22]. Chronic cocaine use has also recently been
associated with coronary artery ectasia and aneur-
ysms [16]. In a consecutive series of 112 cocaine
users who underwent angiography, 30% had cor-
onary ectasia and aneurysms. MI had occurred in
65% of the patients who had coronary ectasia and
aneurysms, including patients who did not have
significant coronary artery stenosis [16]. Epicar-
dial [30] and intramyocardial coronary artery dis-
ease [31] is also present in patients who have chest
pain in the absence of MI.
Cocaine use has also been identified as a pos-
sible cause of aortic dissection. One study in an
inner-city setting reported 38 cases of acute aortic
dissection over a 20-year period, 14 of which
(37%) were associated with cocaine use [32]. Al-
though the International Registry for Aortic Dis-
section suggests that cocaine use is involved in less
than 1% of aortic dissections, this diagnosis may
 COCAINE-ASSOCIATED CHEST PAIN
be underappreciated and must be considered in
patients presenting with cocaine-associated chest
pain [33,34].
Initial approach to the patient in the emergency
department
Patients who have potential cocaine toxicity
should receive a complete evaluation including
a history of cocaine use, recognition of signs and
symptoms consistent with sympathetic nervous
system excess, and evaluation of organ-specific
complaints. It is imperative to determine whether
signs and symptoms are caused by cocaine itself,
underlying unrelated structural abnormalities, or
cocaine-induced structural abnormalities.
The differential diagnosis of cocaine-associated
chest pain is similar to the differential diagnosis of
chest pain unrelated to cocaine except that the
likelihood of a patient having a serious event in
the absence of traditional risk factors for that
particular disease is increased. Potentially serious
causes of chest pain are cardiovascular, pulmo-
nary, and vascular in origin and include myocar-
dial ischemia and infarction, aortic dissection,
pulmonary embolism, pneumothorax, pneumo-
mediastinum, and noncardiogenic pulmonary
edema [35]. Less serious causes of chest pain fol-
lowing cocaine use include traumatic injury and
rhabdomyolysis.
The risk of MI is increased 24-fold in the hour
following cocaine use [36]. Approximately 6% of
patients who have cocaine-associated chest pain
are actually having an acute MI [37,38], and an-
other 15% have an acute coronary syndrome
[37]. The classic patient who has cocaine-associat-
ed MI is a young, male tobacco smoker with a his-
tory of repetitive cocaine use and few other
cardiac risk factors [2,37–50]; however, cocaine-
associated MI has been reported in patients over
60 years old [2,37,39]. Demographic or historical
factors do not reliably predict or exclude acute
MI [37]. Likewise, location of chest pain, duration
of chest pain, quality of chest pain, and symptoms
associated with chest pain are not predictive of
MI [37].
The duration of time for which a patient is at risk
for MI or ischemia following cocaine use is un-
known, because it has been postulated that cocaine-
associated acute coronary syndromes (myocardial
ischemia and infarction) can also occur secondary
to cocaine withdrawal [2,37,45,47]. Spontaneous
episodes of ST-elevation myocardial infarction
105
(STEMI) have been documented for up to 6 weeks
after withdrawal of cocaine [47]. These patients of-
ten do not have classic chest pain syndromes. The
chest pain can be delayed for hours to days after
their most recent use of cocaine, although most co-
caine-related MIs occur within 24 hours of last use
[36,37,39,43,50]. Both STEMI and non-STEMI
can occur [2,40,51]. Chronic cocaine use is also as-
sociated with premature atherosclerosis and coro-
nary ectasia; these patients can present with MI
that is not temporally related to cocaine use.
Electrocardiography
Interpretation of the EKG in patients who
have cocaine-associated chest pain can be difficult,
because patients can have nondiagnostic EKGs in
the setting of ischemia as well as abnormal EKGs
in the absence of ischemia. MI occurs in patients
who have normal or nonspecific EKGs [37,48]. In
one series, patients who had MI were as likely to
present with normal or nonspecific EKGs as
with ischemic EKGs, thereby, resulting in the
emergency department release of 15% of patients
who have MI [37]. The sensitivity of the EKG for
MI is less than in other patients who have acute
MI (approximately 36%) [37].
Conversely, EKGs are abnormal in 56% to
84% of patients who have cocaine-associated
chest pain [37,40,50,52], and up to 43% of pa-
tients who do not have MI may meet standard
EKG criteria for use of reperfusion therapy [50].
J-point and ST-segment elevation secondary to
early repolarization or left ventricular hypertro-
phy makes the identification of ischemia more dif-
ficult in these patients [50,53].
Cardiac markers
Cardiac markers are elevated in MI, but false
elevations of creatine kinase-MB fraction are
common [48,54,55]. Elevations in creatine kinase
and creatine kinase-MB occur in the absence of
MI [54,55] because of cocaine-induced skeletal
muscle injury and rhabdomyolysis. Following
the use of cocaine, approximately 50% of patients
have elevations in serum creatine kinase with or
without myocardial injury [37]. Rising enzyme
patterns are more likely to occur in patients who
have MI [37,48], whereas initial elevations that
rapidly decline less commonly indicate infarction
[37,48]. Additionally, in the setting of an elevated
absolute creatine kinase-MB, reliance should not
be placed entirely on the creatine kinase-MB
106 HOLLANDER & HENRY
relative index, because the creatine kinase-MB rel-
ative index may be falsely low when concurrent
MI and skeletal muscle rhabdomyolysis occur.
Cardiac troponin I and T are more specific than
creatine kinase-MB for myocardial injury when
concomitant skeletal muscle injury exists. Use of
cardiac troponin I or T may enhance the diagnos-
tic accuracy of MI in patients who have cocaine-as-
sociated ischemia and therefore is preferred
[54,55].
Urine drug testing
Patients may initially deny cocaine use. As
a result, urine drug testing can be helpful.
Relatively little cocaine is excreted unchanged in
the urine [56]. Because of a long elimination half-
life, assays for cocaine and cocaine metabolites
generally will detect benzoylecgonine for up to
48 to 72 hours after use.
Other diagnostic tests
Laboratory evaluation may include a complete
blood cell count, electrolytes, glucose, blood urea
nitrogen, creatinine, arterial blood gas analysis,
urinalysis, creatine kinase, and cardiac marker
determinations. Excess sympathetic stimulation
may result in hyperglycemia and hypokalemia.
Patients who have acute cocaine toxicity can have
severe acid-based disturbances as well as rhabdo-
myolysis [57]. A chest radiograph should be ob-
tained in patients who have cardiopulmonary
complaints. It may help demonstrate noncardiac
causes for the chest pain.
Initial disposition decision
Patients who have acute STEMI should receive
immediate reperfusion therapy. Direct percutane-
ous coronary intervention (PCI) is preferred in
light of the frequently complex clinical presenta-
tion and risk of thrombolytics. Patients who have
cocaine-associated MI who are likely to develop
complications can be identified with a high degree
of accuracy during the initial 12 hours of hospi-
talization [58]. Patients who have cocaine-associ-
ated chest pain and do not have infarction have
an extremely low frequency of delayed complica-
tions [37,39,50]. Cost-effective evaluation strate-
gies, such as 9- to 12-hour observation periods,
are appropriate for many patients who have co-
caine-associated chest pain, because these patients
seem to have a low incidence of cardiovascular
complications. Weber and colleagues [59]
demonstrated the safety of a 12-hour observation
protocol in 302 patients who had cocaine-associ-
ated chest pain.
EKG evidence of ischemia, elevated cardiac
markers, and cardiovascular complications occur-
ring before or within 12 hours of arrival predict
late complications [58]. These patients should be
admitted to monitored beds. Other patients who
have cocaine-associated chest pain can be evalu-
ated safely in a 12-hour observation unit [59].
Initial treatment considerations
The initial treatment should focus on airway,
breathing, and circulation. Specific treatments are
based on the specific sign, symptom, or organ
system affected. Because of the direct relationship
between the neuropsychiatric and other systemic
complications, management of neuropsychiatric
manifestations affects the systemic manifestations
of cocaine toxicity.
Patients who have suspected cocaine-induced
ischemia or MI should be treated similarly to
those with traditional acute coronary syndromes,
with some notable exceptions. Aspirin, nitroglyc-
erin, and heparin remain important initial thera-
pies. Intravenous benzodiazepines should be
provided as early management [2,39,44,60–62].
They will decrease the central stimulatory effects
of cocaine, thereby indirectly reducing the cardio-
vascular toxicity of cocaine. Beta antagonists are
contraindicated, because they may exacerbate co-
caine-induced coronary artery vasoconstriction
[39,44,61,62]. Another management difference be-
tween the management of cocaine-using patients
and those who have STEMI is a preference for
PCI over fibrinolysis [39,44,51,61,62]. Finally,
there are anecdotal reports of the safety and ef-
ficacy of phentolamine, an alpha antagonist, for
treatment of cocaine-associated acute coronary
syndromes [39,44,61–62].
Studies in the cardiac catheterization labora-
tory have largely provided the evidence-based
approach to patients who have cocaine-associated
coronary vasoconstriction. In these studies, adult
patients who had not previously used cocaine
were given a low dose of intranasal cocaine (2 mg/
kg). Patients developed an increase in heart rate,
blood pressure, and coronary vascular resistance
with the coronary arterial diameter narrowed by
13% [17]. With administration of phentolamine,
the coronary arterial diameter returned to base-
line [17]. This finding suggests that phentolamine
 COCAINE-ASSOCIATED CHEST PAIN
may be useful for treatment of cocaine-induced is-
chemia. Based on these data, case reports, and an-
ecdotal experience, the International Guidelines
for Emergency Cardiovascular Care recommend
alpha-adrenergic antagonists (phentolamine) for
the treatment of cocaine-associated acute coro-
nary syndrome [61,62].
Case series and one randomized, controlled
trial show that nitroglycerin relieves cocaine-asso-
ciated chest pain [60,63]. Cardiac catheterization
studies demonstrate that nitroglycerin reverses co-
caine-induced vasoconstriction [64]. Benzodiaze-
pines have a salutary effect on the hyperdynamic
effects of cocaine and relieve chest pain [60]. Ben-
zodiazepines are similar to nitroglycerin with re-
spect to effects on pain relief, cardiac dynamics,
and left ventricular function in patients who have
cocaine-associated chest pain [60].
The role of calcium-channel blockers for the
treatment of cocaine-associated chest pain re-
mains ill defined. Pretreatment of cocaine-intoxi-
cated animals with calcium-channel blockers has
had variable results with respect to survival,
seizures, and cardiac dysrhythmias [65–71]. In car-
diac catheterization studies, verapamil reverses
cocaine-induced coronary artery vasoconstriction
[72]. Large-scale multicenter clinical trials in pa-
tients who have acute coronary syndromes unre-
lated to cocaine have not demonstrated any
beneficial effects of calcium-channel blockers on
important outcomes such as survival. Thus, the
role of calcium-channel blockers in patients who
have cocaine-induced acute coronary syndrome
has not yet been defined.
Cocaine induced coronary artery vasoconstric-
tion is clearly exacerbated by the administration of
propranolol [18]. An unopposed alpha-adrenergic
effect may occur, which leads to vasoconstriction
and an increased blood pressure [73–75]. Multiple
experimental models have shown that beta-adren-
ergic antagonists lead to decreased coronary blood
flow, increased seizure frequency, and high fatality
rates [71,76–79]. The use of short-acting beta-
adrenergic antagonists such as esmolol has re-
sulted in significant increases in blood pressure
in up to 25% of patients [80,81]. Therefore,
the use of beta-adrenergic antagonists for the
treatment of cocaine toxicity is contraindicated
[39,44,61,62].
Labetalol does not seem to offer any advan-
tages over propranolol. It has substantially more
beta-adrenergic antagonism than alpha-adrener-
gic antagonist effects [82]. Labetalol increases the
risk of seizure and death in animal models of
107
cocaine toxicity [71] and does not reverse coronary
artery vasoconstriction in humans [83]. Nitroglyc-
erin or phentolamine is considered a better option
to achieve vasodilation [39,44,61,62].
Cocaine injures the vascular endothelium, in-
creases platelet aggregation, and impairs normal
fibrinolytic pathways [84]. As a result, the use of
antiplatelet and antithrombin agents makes theo-
retical sense [39,44,61,62,85]. Fibrinolytic admin-
istration poses problems, however. Many young
patients have benign early repolarization, and
only a small percentage of patients who have co-
caine-associated chest pain syndromes and J-point
elevation are actually in the midst of an acute MI
[50,53]. Patients are frequently hypertensive, and
aortic dissection must be considered. Several
case reports document adverse outcomes follow-
ing fibrinolytic administration in cocaine-using
patients [86–88]. Patients who have cocaine-asso-
ciated STEMI have a low mortality; therefore,
the benefit of early fibrinolytic administration is
less than in patients who have STEMI unrelated
to cocaine, and the risk is higher. Therefore, fibri-
nolytic therapy should be reserved for patients
who are definitely having STEMI and cannot re-
ceive primary PCI [36,60,61,85]. With the increas-
ing availability of primary PCI including transfer,
this situation should rarely occur. More aggres-
sive antiplatelet therapy with glycoprotein IIb/
IIIa antagonists and clopidogrel may be useful
in patients who have cocaine-associated acute cor-
onary syndromes, but they have not been well
studied in this patient population [89].
Hypertension and tachycardia alone rarely
require specific treatment but may need to be
addressed in a patient who has definite acute
coronary syndromes. In a patient with chest pain
of unclear cause, hypertension and tachycardia
alone should be treated conservatively. Resolution
of anxiety, agitation, and ischemia often lead to
resolution of the hypertension and tachycardia.
When necessary, treatment directed toward the
central effects of cocaine, such as the benzodiaze-
pines, usually reduces blood pressure and heart
rate. When sedation is unsuccessful, hypertension
can be managed with sodium nitroprusside, ni-
troglycerin, or intravenous phentolamine [17,64].
Most atrial arrhythmias respond to sedative
hypnotics. When they do not, verapamil or diltia-
zem may be indicated. The treatment of ventricular
arrhythmias depends upon the time between co-
caine use, arrhythmia onset, and treatment. Ven-
tricular arrhythmias occurring immediately after
cocaine use should be presumed to occur from the
108 HOLLANDER & HENRY
local anesthetic (sodium-channel) effects on the
myocardium. They may respond to the adminis-
tration of sodium bicarbonate, similar to arrhyth-
mias associated with other type IA and type IC
agents [14,90]. In addition, one animal model sug-
gested that lidocaine exacerbates cocaine-induced
seizures and arrhythmias as a result of similar
effects on sodium channels [91]; however, this find-
ing has not been confirmed in other animal models
[14,92,93]. Bicarbonate therapy may be preferable
and has been used effectively [94].
Ventricular arrhythmias that develop several
hours after the last use of cocaine often occur as
a result of ischemia. Standard management for
ventricular arrhythmias, including lidocaine, is
indicated and seems to be safe [95]. There are no
data concerning the efficacy of amiodarone in
clinical cocaine intoxication. Torsades de pointes
is a rare complication of cocaine use [96] and
should be managed with intravenous magnesium
sulfate and overdrive pacing.
Other cardiovascular effects of cocaine
Cocaine also causes significant cardiovascular
conditions besides those that result in chest pain.
Cocaine has a direct myocardial-depressant effect
[13,97]. Chronic cocaine use leads to a dilated car-
diomyopathy, possibly from recurrent or diffuse
ischemia with subsequent ‘‘stunned’’ myocardium
[98]. Alternatively, it may a direct effect on myo-
cardial contractility. Direct infusion of cocaine
into human coronary arteries increases left ven-
tricular end-diastolic pressures and end-systolic
volume and decreases left ventricular ejection frac-
tion [99]. Left ventricular function may improve
when cocaine use is halted [100].
Intravenous cocaine use increases the risk of
bacterial endocarditis, even more than intrave-
nous heroin use, presumably because of the
increased frequency of injection to sustain effects
[101,102]. Direct effects of cocaine on endovascu-
lar tissues and the immune system may also play
a role [102].
Aortic dissection is a rare but life-threatening
complication of cocaine abuse and must be
considered in the differential diagnosis. In a con-
secutive series, 37% of acute aortic dissections in
an inner-city hospital were associated with co-
caine use. This patient cohort was younger than
expected with a high percentage of African
American males with untreated hypertension [32].
Higher doses of cocaine are associated with
virtually all types of tachyarrhythmias. Atrial
fibrillation, atrial flutter, supraventricular tachy-
cardias, ventricular premature contractions, ac-
celerated idioventricular rhythms, ventricular
tachycardia, torsades de pointes, and ventricular
fibrillation may occur as a result of cocaine. High
doses of cocaine lead to infranodal and intraven-
tricular conduction delays and lethal ventricular
arrhythmias secondary to prolonged QRS and QT
intervals [103,104]. Prolonged QT intervals have
been noted in patients who have recently used co-
caine and who have not had arrhythmias [53].
These effects are probably mediated by the local
anesthetic sodium-channel blockade. In addition
to the local anesthetic effects, arrhythmias may
also occur as a result of cocaine-induced acute
coronary syndrome [13,95]. Low doses of cocaine
can result in a transient bradycardia.
In-hospital management
There is limited specific information available
regarding the in-hospital management of patients
who have cocaine-related cardiovascular disease.
Therefore, as a general rule, treatment guidelines
follow those recommended for patients who have
acute coronary syndromes not associated with
cocaine Table 1 [105,106].
ST-elevation myocardial infarction
The diagnosis of STEMI can be more chal-
lenging in patients who have acute cocaine toxic-
ity because of the atypical presentation and
challenges with interpretation of the EKG, in-
cluding left ventricular hypertrophy and early
repolarization. Patients who have acute cocaine
toxicity frequently are younger than expected and
are hypertensive, and aortic dissection must be
considered in the differential diagnosis. Therefore
cardiac catheterization with direct PCI is the
preferred method of reperfusion, and fibrinolytic
therapy should be reserved for patients when
cardiac catheterization is not available. With the
development of transfer systems for STEMI,
fibrinolytic therapy should rarely be required.
The method of revascularization and subsequent
management should follow guidelines for treat-
ment of patients who do not have a history of
cocaine abuse, with a few exceptions. There are no
data available regarding the use of drug-eluting
stents in patients who abuse cocaine, but they
would be expected to decrease target lesion re-
vascularization compared with bare metal stents.
 COCAINE-ASSOCIATED CHEST PAIN 109

Patients with ongoing cocaine abuse may have
poor compliance with the required chronic anti-
platelet regimen of aspirin and clopidogrel, which
could potentially increase the risk of subacute
thrombosis. Therefore each patient’s potential for
drug rehabilitation and compliance history needs
to be considered in the stent choice. Patients who
have accelerated atherosclerosis, coronary aneur-
ysms, and ectasia require aggressive risk factor
modification and antiplatelet therapy. Long-term
clopidogrel should be considered in addition to
aspirin. Patients who have left ventricular dys-
function should receive angiotensin-converting
enzyme inhibitor therapy. Although patients
who have STEMI would be expected to benefit
from long-term beta blockade, caution should be
used in patients expected to have continued
exposure to cocaine.
High-risk unstable angina/non–ST-segment
elevation myocardial infarction
Patients who have elevated cardiac enzymes or
abnormal EKGs are at higher risk for subsequent
events and benefit from an early invasive ap-
proach with cardiac catheterization and revascu-
larization [107]. Although no specific data exist
for cocaine-related unstable angina/non-STEMI,
it is reasonable to believe these patients would
benefit from a similar approach. It is important
to use drug rehabilitation as well as aggressive
risk factor modification in these patients, because

Table 1
Treatment summary for specific cocaine-related medical conditions
Medical Condition
Cardiovascular complications
Dysrhythmias
Sinus tachycardia
Supraventricular tachycardia
Ventricular dysrhythmias
Acute coronary syndrome
Hypertension
Pulmonary edema
Adapted from Hollander JE and Hoffman RS. Cocaine. In Godlfrank LR, Flomenbaum NE, Lewin NA, et al,
editors. Goldfrank’s toxicologic emergency. 6th edition. Stamford (CT): Appleton and Lange,; 1998. p. 1071–89.
Treatments
observation oxygen
diazepam 5 mg IV or lorazepam 2–4 mg IV titrated to effect
oxygen
diazepam 5 mg IV or lorazepam 2–4 mg IV
consider diltiazem 20 mg IV or verapamil 5 mg IV
adenosine 6 mg or 12 mg IV
cardioversion if hemodynamically unstable
oxygen
sodium bicarbonate 1–2 meq/kg
lidocaine 1.5 mg/kg IV bolus followed by 2 mg/min infusion
defibrillation if hemodynamically unstable
diazepam 5 mg IV or lorazepam 2–4 mg IV
oxygen
diazepam 5–10 mg IV or lorazepam 2–4 mg IV
soluble aspirin 325 mg
nitroglycerin 1/150 sublingual  three every 5 minutes followed by a infusion titrated
to a mean arterial pressure reduction of 10% or relief of chest pain.
morphine sulfate 2 mg IV every 5 minutes titrated to pain relief
phentolamine 1 mg IV; repeat in 5 minutes
verapamil 5–10 mg IV
heparin or enoxaparin
percutaneous intervention (angioplasty and stent placement)
glycoprotein IIb/IIIa inhibitors
observation
diazepam 5–10 mg IV or lorazepam 2–4 mg IV titrated to effect
phentolamine 1 mg IV; repeat in 5 minutes
nitroglycerin or nitroprusside continuous infusion titrated to effect
lasix 20–40 mg IV
morphine sulfate 2 mg IV every 5 minutes titrated to pain relief or respiratory status
nitroglycerin infusion titrated to blood pressure
consider phentolamine or nitroprusside
110 HOLLANDER & HENRY

there is a high likelihood of recurrence. In one re- factor modification is indicated in patients who
view of patients who had cocaine-associated MI have MI or evidence of premature atherosclerosis,
followed for median of 4.5 months, 12 of 24 pa- coronary artery aneurysm, or ectasia. This risk
tients had recurrent ischemic pain; 8 of whom sus- factor modification includes smoking cessation,
tained a second MI, suggesting that these patients hypertension control, diabetes control, and ag-
are at high risk for subsequent events [2]. gressive lipid-lowering therapy with a target low-
density lipoprotein level below 70. Although these
strategies have not been tested specifically for pa-
Low-risk unstable angina tients who have cocaine-associated chest pain,
they are standard of care for patients who have
Most patients who have cocaine-related chest
underlying coronary artery disease.
pain have low-risk unstable angina and can be
Patients who have evidence of MI or athero-
treated satisfactorily with a 12-hour observation
sclerosis should receive long-term antiplatelet
protocol [59].
therapy with aspirin. In addition to aspirin,
Patients who have cocaine-associated chest
clopidogrel should be given for at least 1 month
pain have a 1-year survival rate of 98% and an
with bare metal stents or for 3 to 6 months with
incidence of late MI of only 1% [49]. Most deaths
currently available drug-eluting stents. Long-term
occur because of concurrent medical problems
combination antiplatelet therapy with aspirin and
(such as HIV disease). Because patients who
clopidogrel may be beneficial in patients who have
have cocaine-associated chest pain are not at
extensive atherosclerosis, coronary ectasia, or
high risk for MI or death during the ensuing
aneurysm, but this possibility has not been
year, urgent cardiac evaluation is probably not
studied. The role of nitrates and calcium-channel
necessary for patients in whom MI is ruled out.
blockers remains speculative and should be used
Evaluation for possible underlying coronary ar-
for symptomatic relief. The use of beta-adrenergic
tery disease may be accomplished on a more elec-
antagonists, although useful in patients who have
tive basis. Continued cocaine usage, however, is
had a previous MI and cardiomyopathy, needs
associated with an increased likelihood of recur-
special consideration in the setting of cocaine
rent chest pain, and therefore aggressive drug re-
abuse. Because recidivism is high in patients who
habilitation may be useful [49].
have cocaine-associated chest pain (60% admit to
The appropriate diagnostic evaluation for
cocaine use within the next year [49]), beta-blocker
these patients remains unclear and therefore
therapy probably should be avoided in certain
should follow general principles for risk stratifi-
patients.
cation in patients who have coronary artery
disease. In light of the underlying EKG abnor-
malities, most patients would benefit from imag- Summary
ing with stress testing, either echocardiography or
Patients who have chest pain following the use
nuclear. Many patients who have cocaine-related
of cocaine have become more common in emer-
chest pain have pain related to coronary vasocon-
gency departments throughout the United States,
striction or increased myocardial demand and
therefore have negative stress evaluations. There with approximately 6% of these patients sustain-
ing an acute MI. The authors have described the
have been remarkable advances in cardiac imag-
rationale for recommending aspirin, benzodiaze-
ing in the last few years. Both cardiac MRI and
CT angiography have theoretical advantages over pines, and nitroglycerin as first-line treatments
and calcium-channel blockade or phentolamine as
conventional stress imaging for detection of pre-
possible second-line therapies and have summa-
mature atherosclerosis and coronary artery aneur-
rized the controversies surrounding the use of
ysms, but there are currently no data available
fibrinolytic agents. Admission for observation is
specific to cocaine-related chest pain.
one reasonable approach to the management of
the low-risk cohort. Evaluation for underlying
coronary artery disease is reasonable, particularly
Secondary prevention
in patients who have acute MI. Patients who do
Cessation of cocaine use is the hallmark of not have infarction can undergo evaluation for
secondary prevention. Recurrent chest pain is less possible coronary artery disease on an outpatient
common and MI and death are rare in patients basis. Routine interventions for secondary pro-
who discontinue cocaine use [49]. Aggressive risk phylaxis as well as cocaine rehabilitation should
 COCAINE-ASSOCIATED CHEST PAIN
be used in this patient population, because the
long-term prognosis seems somewhat dependent
upon the ability of the patient to discontinue
cocaine use.
References
[1] Haddad LM. 1978: Cocaine in perspective. JACEP
1979;8:374–6.
[2] Hollander JE, Hoffman RS. Cocaine induced myo-
cardial infarction: an analysis and review of the
literature. J Emerg Med 1992;10:169–77.
[3] Substance Abuse and Mental Health Services Ad-
ministration. 2003 National Survey on Drug Use
and Health. Available at: http://www.samhsa.gov/
oas/oasftp.htm. Accessed March 5, 2005.
[4] Hollander JE, Hoffman RS. Cocaine. In:
Goldfrank LR, Flomenbaum NE, Lewin NA,
et al, editors. Goldfrank’s toxicologic emergency.
7th edition. New York: McGraw Hill; 2002. p.
1004–19.
[5] Borne RF, Bedford JA, Buelke JL, et al. Biological
effects of cocaine, derivatives I: improved synthesis
and pharmacologic evaluation of norcocaine.
J Pharm Sci 1977;66:119–29.
[6] Schreiber MD, Madden JA, Covert RF, et al.
Effects of cocaine, benzoylecgonine and cocaine
metabolites on cannulated pressurized fetal sheep
cerebral arteries. J Appl Physiol 1994;77:834–9.
[7] Madden J, Powers R. Effect of cocaine and cocaine
metabolites on cerebral arteries in vitro. Life Sci
1990;47:1109–14.
[8] Crumb WJ, Clarkson CW. Characterization of so-
dium channel blocking properties of the major
metabolites of cocaine in single cardiac myocytes.
J Pharmacol Exp Ther 1992;261:910–7.
[9] Brookoff D, Rotondo MF, Shaw LM, et al. Coca-
ethylene levels in patients who test positive for co-
caine. Ann Emerg Med 1996;27:316–20.
[10] Henning RJ, Wilson LD, Glauser JM. Cocaine plus
ethanol is more cardiotoxic than cocaine or ethanol
alone. Crit Care Med 1994;22:1896–906.
[11] Pirwitz MJ, Willard JE, Landau C, et al. Influence
of cocaine, ethanol, or their combination on epicar-
dial coronary arterial dimensions in humans. Arch
Intern Med 1995;155:1186–91.
[12] Kolodgie FD, Wilson PS, Mergner WJ, et al. Co-
caine induced increase in the permeability function
of human vascular endothelial cell monolayers.
Exp Mol Pathol 1999;66:109–22.
[13] Bauman JL, Grawe JJ, Winecoff AP, et al. Co-
caine-related sudden cardiac death: a hypothesis
correlating basic science and clinical observations.
J Clin Pharmacol 1994;34:902–11.
[14] Winecoff AP, Hariman RJ, Grawe JJ, et al. Rever-
sal of the electrocardiographic effects of cocaine
by lidocaine. Part 1. Comparison with sodium
111
bicarbonate and quinidine. Pharmacotherapy
1994;14:698–703.
[15] Tella SR, Schindler CW, Goldberg SR. Cocaine:
cardiovascular effects in relation to inhibition of
peripheral neuronal monoamine uptake and cen-
tral stimulation of the sympathoadrenal system.
J Pharmacol Exp Ther 1993;267:153–62.
[16] Satran A, Bart BA, Henry CR, et al. Increased
prevalence of coronary artery aneurysms among
cocaine users. Circulation 2005;111:2424–9.
[17] Lange RA, Cigarroa RG, Yancy CW, et al. Co-
caine-induced coronary-artery vasoconstriction.
N Engl J Med 1989;321:1557–61.
[18] Lange RA, Cigarroa RG, Flores ED, et al. Poten-
tiation of cocaine-induced coronary vasoconstric-
tion by beta-adrenergic blockade. Ann Intern
Med 1990;112:897–903.
[19] Flores ED, Lange RA, Cigarroa RC, et al. Effect of
cocaine on coronary artery dimensions in athero-
sclerotic coronary artery disease: enhanced vaso-
constriction at sites of significant stenosis. J Am
Coll Cardiol 1990;16:74–9.
[20] Winniford MD, Wheelan KR, Kremers MS, et al.
Smoking-induced coronary vasoconstriction in
patients with atherosclerotic coronary artery dis-
ease: evidence for adrenergically mediated altera-
tions in coronary artery tone. Circulation 1986;
73:662–7.
[21] Moliterno DJ, Willard JE, Lange RA, et al. Coro-
nary artery vasoconstriction induced by cocaine,
cigarette smoking, or both. N Engl J Med 1994;
330:454–9.
[22] Weber JE, Hollander JE, Murphy SA, et al. Quan-
titative comparison of coronary artery flow and
myocardial perfusion in patients with acute myo-
cardial infarction in the presence and absence of
recent cocaine use. J Thromb Thrombolysis 2003;
14(3):239–45.
[23] Togna G, Tempesta E, Togna AR, et al. Platelet re-
sponsiveness and biosynthesis of thromboxane and
prostacyclin in response to in vitro cocaine treat-
ment. Haemostasis 1985;15:100–7.
[24] Schnetzer GW III. Platelets and thrombogenesisd
current concepts. Am Heart J 1972;83:552–64.
[25] Rezkalla S, Mazza JJ, Kloner RA, et al. The effect
of cocaine on human platelets. Am J Cardiol 1993;
72:243–6.
[26] Moliterno DJ, Lange RA, Gerard RD, et al. Influ-
ence of intranasal cocaine on plasma constituents
associated with endogenous thrombosis and
thrombolysis. Am J Med 1994;96:492–6.
[27] Mittleman RE, Wetli CV. Cocaine and sudden
‘‘natural’’ death. J Forensic Sci 1987;32:11–9.
[28] Dressler FA, Malekzadeh S, Roberts WC. Quanti-
tative analysis of amounts of coronary arterial nar-
rowing in cocaine addicts. Am J Cardiol 1990;65:
303–8.
[29] Tardiff K, Gross E, Wu J, et al. Analysis of cocaine
positive fatalities. J Forensic Sci 1989;34:53–63.
112 HOLLANDER & HENRY
[30] Om A, Warner M, Sabri N, et al. Frequency of
coronary artery disease and left ventricular
dysfunction in cocaine users. Am J Cardiol 1992;
69:1549–52.
[31] Majid PA, Patel B, Kim HS, et al. An angiographic
and histologic study of cocaine induced chest pain.
Am J Cardiol 1990;65:812–4.
[32] Hsue PY, Salina CL, Bolger AF, et al. Acute aortic
dissection related to crack cocaine. Circulation
2002;105:1592–5.
[33] Eagle KA, Isselbacher EM, DeSanctis RW. Co-
caine-related aortic dissection in perspective. Cir-
culation 2002;105:1529.
[34] Hagan PG, Nienaber CA, Isselbacher EM, et al.
The International Registry of Acute Aortic Dissec-
tion (IRAD); new insights into an old disease.
JAMA 2000;283:897–903.
[35] Thadani PV. NIDA conference report on cardio-
pulmonary complications of crack cocaine used
clinical manifestations and pathophysiology. Chest
1996;110:1072–6.
[36] Mittleman MA, Mintzewr D, Maclure M, et al.
Triggering of myocardial infarction by cocaine.
Circulation 1999;99:2737–41.
[37] Hollander JE, Hoffman RS, Gennis P, et al. Pro-
spective multicenter evaluation of cocaine associ-
ated chest pain. Acad Emerg Med 1994;1:330–9.
[38] Weber JE, Chudnofsky C, Wilkerson MD, et al.
Cocaine associated chest pain: how common is
myocardial infarction? Acad Emerg Med 2000;7:
873–7.
[39] Hollander JE. Management of cocaine associated
myocardial ischemia. N Engl J Med 1995;333:
1267–72.
[40] Amin M, Gabelman G, Karpel J, et al. Acute myo-
cardial infarction and chest pain: syndromes after
cocaine use. Am J Cardiol 1990;66:1434–7.
[41] Del Aguila C, Rosman H. Myocardial infarction
during cocaine withdrawal [letter]. Ann Intern
Med 1990;112:712.
[42] Hollander JE, Carter WC, Hoffman RS. Use of
phentolamine for cocaine induced myocardial
ischemia [letter]. N Engl J Med 1992;327:361.
[43] Hollander JE, Hoffman RS, Burstein J, et al. Co-
caine associated myocardial infarction. Mortality
and complications. Arch Intern Med 1995;155:
1081–6.
[44] Lange RA, Hillis RD. Cardiovascular complica-
tions of cocaine use. N Engl J Med 2001;345:351–8.
[45] Levine MAH, Nishakawa J. Acute myocardial in-
farction associated with cocaine withdrawal. Can
Med Assoc J 1991;144:1139–40.
[46] Minor RL, Scott BD, Brown DD, et al. Cocaine in-
duced myocardial infarction in patients with nor-
mal coronary arteries. Ann Intern Med 1991;115:
797–806.
[47] Nademanee K, Gorelick DA, Josephson MA, et al.
Myocardial ischemia during cocaine withdrawal.
Ann Intern Med 1989;111:876–80.
[48] Tokarski GF, Paganussi P, Urbanski R, et al. An
evaluation of cocaine-induced chest pain. Ann
Emerg Med 1990;19:1088–92.
[49] Hollander JE, Hoffman RS, Gennis P, et al. Co-
caine associated chest pain: one year follow-up.
Acad Emerg Med 1995;2:179–84.
[50] Gitter MJ, Goldsmith ER, Dunbar DN, et al. Co-
caine and chest pain: clinical features and outcome
of patients hospitalized to rule out myocardial in-
farction. Ann Intern Med 1991;115:277–82.
[51] Hollander JE, Burstein JL, Shih RD, et al. Cocaine
associated myocardial infarction: clinical safety of
thrombolytic therapy. Chest 1995;107:1237–41.
[52] Zimmerman JL, Dellinger RP, Majid PA. Cocaine
associated chest pain. Ann Emerg Med 1991;20:
611–5.
[53] Hollander JE, Lozano M Jr, Fairweather P, et al.
‘‘Abnormal’’ electrocardiograms in patients with
cocaine-associated chest pain are due to ‘‘normal’’
variants. J Emerg Med 1994;12:199–205.
[54] Hollander JE, Levitt MA, Young GP, et al. The ef-
fect of cocaine on the specificity of cardiac markers.
Am Heart J 1998;135:245–52.
[55] McLaurin M, Apple FS, Henry TD, et al. Cardiac
troponin I and T concentrations in patients with
cocaine-associated chest pain. Ann Clin Biochem
1996;33:183–6.
[56] Jatlow PI. Drug of abuse profile: cocaine. Clin
Chem 1987;33:66b–71b.
[57] Drake TR, Henry T, Marx J, et al. Severe acid-
base abnormalities associated with cocaine abuse.
J Emerg Med 1990;8:331–4.
[58] Hollander JE, Hoffman RS, Burstein J, et al, for the
Cocaine Associated Myocardial Infarction Study
Group. Cocaine associated myocardial infarction:
complications and morbidity. Arch Intern Med
1995;155:1081–6.
[59] Weber JE, Shofer FS, Larkin GL, et al. Validation
of a brief observation period for patients with co-
caine associated chest pain. N Engl J Med 2003;
348:510–7.
[60] Baumann BM, Perrone J, Hornig SE, et al. Ran-
domized controlled double blind placebo con-
trolled trial of diazepam, nitroglycerin or both for
treatment of patients with potential cocaine associ-
ated acute coronary syndromes. Acad Emerg Med
2000;7:878–85.
[61] The American Heart Association in collaboration
with the International Liaison Committee on Re-
suscitation (ILCOR). Guidelines for cardiopulmo-
nary resuscitation and emergency cardiovascular
care. Circulation 2000;102:89.
[62] Albertson TE, Dawson A, de Latorre F, et al.
TOX-ACLS: toxicologic-oriented advanced car-
diac life support. Ann Emerg Med 2001;37:S78–90.
[63] Hollander JE, Hoffman RS, Gennis P, et al. Nitro-
glycerin in the treatment of cocaine associated chest
pain: clinical safety and efficacy. J Toxicol Clin
Toxicol 1994;32:243–56.
 COCAINE-ASSOCIATED CHEST PAIN
[64] Brogan WC, Lange RA, Kim AS, et al. Alleviation
of cocaine-induced coronary vasoconstriction by
nitroglycerin. J Am Coll Cardiol 1991;18:581–6.
[65] Derlet RW, Albertson TE. Diazepam in the preven-
tion of seizures and death in cocaine-intoxicated
rats. Ann Emerg Med 1989;18:542–6.
[66] Billman GE, Hoskins RS. Cocaine-induced ventric-
ular fibrillation: protection afforded by the calcium
antagonist verapamil. FASEB J 1988;2:2990–5.
[67] Nahas G, Trouve R, Demus JF, et al. A calcium
channel blocker as antidote to the cardiac effects
of cocaine intoxication. N Engl J Med 1985;313:
519 [Letter.].
[68] Trouve R, Nahas GG, Maillet M. Nitrendipine as
an antagonist to the cardiac toxicity of cocaine.
J Cardiovasc Pharmacol 1987;9:S49–53.
[69] Derlet RW, Albertson TE. Potentiation of cocaine
toxicity with calcium channel blockers. Am J
Emerg Med 1989;7:464–8.
[70] Hale SL, Alker KJ, Rezkalla SH, et al. Nifedipine
protects the heart from the acute deleterious effects
of cocaine if administered before but not after
cocaine. Circulation 1991;83:1437–43.
[71] Smith M, Garner D, Niemann JT. Pharmacologic
interventions after an LD50 cocaine insult in
a chronically instrumented rat model: are beta
blockers contraindicated? Ann Emerg Med 1991;
20:768–71.
[72] Negus BH, Willard JE, Hillis LD, et al. Alleviation
of cocaine induced coronary vasoconstriction with
intravenous verapamil. Am J Cardiol 1994;73:
510–3.
[73] Ramoska E, Sacchetti AD. Propranolol-induced
hypertension in treatment of cocaine intoxication.
Ann Emerg Med 1985;14:112–3.
[74] Rappolt RT, Gay G, Inaba DS. Use of inderal
(propranolo-Ayerst) in 1-a (early stimulative) and
1-b (advanced stimulative) classification of cocaine
and other sympathomimetic reactions. Clin Toxi-
col 1978;13:325–32.
[75] Rappolt TR, Gay G, Inaba DS, et al. Propranolol
in cocaine toxicity [letter]. Lancet 1976;2:640–1.
[76] Catravas JD, Waters IW. Acute cocaine intoxica-
tion in the conscious dog: studies on the mecha-
nism of lethality. J Pharmacol Exp Ther 1981;217:
350–6.
[77] Guinn MM, Bedford JA, Wilson MC. Antagonism
of intravenous cocaine lethality in nonhuman pri-
mates. Clin Toxicol 1980;16:499–508.
[78] Spivey WH, Schoffstall JM, Kirkpatrick R, et al.
Comparison of labetalol, diazepam, and haloperi-
dol for the treatment of cocaine toxicity in a swine
model. Ann Emerg Med 1990;19:467–8.
[79] Vargas R, Gillis RA, Ramwell PW. Propanolol
promotes cocaine induced spasm of porcine coro-
nary artery. J Pharmacol Exp Ther 1991;257:644–6.
[80] Pollan S, Tadjziechy M. Esmolol in the manage-
ment of epinephrine and cocaine induced cardio-
vascular toxicity. Anesth Analg 1989;69:663–4.
113
[81] Sand IC, Brody SL, Wrenn KD, et al. Experience
with esmolol for the treatment of cocaine associated
cardiovascular complications. Am J Emerg Med
1991;9:161–3.
[82] Sybertz EJ, Sabin CS, Pula KK, et al. Alpha and
beta adrenoreceptor blocking properties of labeta-
lol and its R, R-isomer, SCH 19927. J Pharmacol
Exp Ther 1981;218:435–43.
[83] Boehrer JD, Moliterno DJ, Willard JE, et al. Influ-
ence of labetalol of cocaine-induced coronary vaso-
constriction in humans. Am J Med 1993;94:608–10.
[84] Rinder HM, Ault KA, Jatlow PI, et al. Platelet
alpha granule release in cocaine users. Circulation
1994;90:1162–7.
[85] Hoffman RS, Hollander JE. Thrombolytic therapy
in cocaine-induced myocardial infarction [edi-
torial]. Am J Emerg Med 1996;14:693–5.
[86] Bush HS. Cocaine associated myocardial infarc-
tion: a word of caution about thrombolytic thera-
py. Chest 1988;94:878.
[87] Hollander JE, Wilson LD, Leo PJ, et al. Complica-
tions from the use of thrombolytic agents in pa-
tients with cocaine associated chest pain. J Emerg
Med 1996;14:731–6.
[88] LoVecchio F, Nelson L. Intraventricular bleeding
after the use of thrombolytics in a cocaine user.
Am J Emerg Med 1996;14:663–4.
[89] Franogiannis NG, Farmer JA, Lakkis NM. Tirofi-
ban for cocaine induced coronary artery thrombo-
sis. A novel therapeutic approach. Circulation
1999;100:1939.
[90] Beckman KJ, Parker RB, Hariman RJ, et al. He-
modynamic and electrophysiological actions of co-
caine: effects of sodium bicarbonate as an antidote
in dogs. Circulation 1991;83:1799–807.
[91] Derlet RW, Albertson TE, Tharratt RS. Lidocaine
potentiation of cocaine toxicity. Ann Emerg Med
1991;20:135–8.
[92] Grawe JJ, Hariman RJ, Winecoff AP, et al. Rever-
sal of the electrocardiographic effects of cocaine by
lidocaine, 2. Concentration-effect relationships.
Pharmacotherapy 1994;14:704–11.
[93] Heit J, Hoffman RS, Goldfrank LR. The effects of
lidocaine pretreatment on cocaine neurotoxicity
and lethality in mice. Acad Emerg Med 1994;1:
438–42.
[94] Kerns W, Garvey L, Owens J. Cocaine induced
wide complex dysrhythmia. J Emerg Med 1997;
15:321–9.
[95] Shih RD, Hollander JE, Hoffman RS, et al. Clinical
safety of lidocaine in cocaine associated myocardial
infarction. Ann Emerg Med 1995;26:702–6.
[96] Schrem SS, Belsky P, Schwartzman D, et al. Co-
caine-induced torsades de pointes in a patient
with idiopathic long QT syndrome. Am Heart J
1990;120:980–4.
[97] Hale SL, Alker KJ, Rezkalla S, et al. Adverse
effects of cocaine on cardiovascular dynamics,
myocardial blood flow, and coronary artery
114 HOLLANDER & HENRY
diameter in an experimental model. Am Heart J
1989;118:927–33.
[98] Weiner RS, Lockhart JT, Schwartz RG. Dilated
cardiomyopathy and cocaine abuse. Am J Med
1986;81:699–701.
[99] Pitts WR, Vongpatannasin W, Cigoarroa JE, et al.
Effects of intracoronary infusion of cocaine on left
ventricular systolic and diastolic function in
humans. Circulation 1998;97:1270–3.
[100] Chokshi SK, Moore R, Pandian NG, et al. Revers-
ible cardiomyopathy associated with cocaine intox-
ication. Ann Intern Med 1989;111:1039–40.
[101] Chambers HF, Morris DL, Tauber MG, et al. Co-
caine use and the risk for endocarditis in intrave-
nous drug users. Ann Intern Med 1987;106:833–6.
[102] Weiss SH. Links between cocaine and retroviral in-
fection. JAMA 1989;261:607–8.
[103] Parker RB, Beckman KJ, Hariman RJI, et al. The
electrophysiologic and arrhythmogenic effects
of cocaine [abstract]. Pharmacotherapy 1989;9:176.
[104] Schwartz AB, Janzen D, Jones RT, et al. Electro-
cardiographic and hemodynamic effects of intrave-
nous cocaine in the awake and anesthetized dogs.
J Electrocardiol 1989;22:159–66.
[105] Braunwald E, Antman EM, Beasley JW, et al.
ACC/AHA 2002 guideline update for the manage-
ment of patients with unstable angina and non-ST-
segment elevation myocardial infarctiondsummary
article: a report of the American College of Cardiol-
ogy/American Heart Association task force on
practice guidelines (Committee on the Management
of Patients With Unstable Angina). J Am Coll Car-
diol 2002;40:1366–74.
[106] Antman EM, Anbe DT, Armstrong PW, et al.
ACC/AHA guidelines for the management of
patients with ST-elevation myocardial infarctiond
executive summary: a report of the American Col-
lege of Cardiology/American Heart Association
Task Force on Practice Guidelines (Writing Com-
mittee to Revise the 1999 Guidelines for the Man-
agement of Patients With Acute Myocardial
Infarction). Circulation 2004;110:588–636.
[107] Cannon CP, Weintraub WS, Demopoulos LA,
et al. Comparison of early invasive and conserva-
tive strategies in patients with unstable coronary
syndromes treated with the glycoprotein IIb/IIIa
inhibitor tirofiban. N Engl J Med 2001;344:
1879–87.
 Cardiol Clin 24 (2006) 115–123

Acute Congestive Heart Failure in the Emergency

Department
Robert L. Rogers, MDa, Erika D. Feller, MDb,
Stephen S. Gottlieb, MDb,*
a
Division of Emergency Medicine, Department of Medicine, University of Maryland School of Medicine,
110 South Paca Street, Sixth Floor, Suite 200, Baltimore, MD 21201, USA
b
Division of Cardiology, Department of Medicine, University of Maryland School of Medicine,
22 South Greene Street, Baltimore, MD 21201, USA

Acute congestive heart failure (CHF) and
pulmonary edema is a clinical entity commonly
encountered in the emergency department. It is
estimated that more than 5 million people in the
United States have CHF, and it is expected that as
the population ages the incidence of CHF and
emergency department visits for acutely decom-
pensated CHF and pulmonary edema will rise
[1,2]. As survival rates for acute myocardial in-
farction continue to increase, the incidence of
heart failure is expected to increase as well. The
estimated prevalence of CHF in adults over the
age of 75 years is 10%, with a lifetime risk of al-
most 20% [3]. In the Acute Decompensated Heart
Failure National Registry (ADHERE), a large,
national database of demographic, clinical, and
outcomes data for patients hospitalized for de-
compensated CHF, the emergency department is
the initial site of care for more than 78% of pa-
tients who have acute symptomatic heart failure
[4].
The syndrome of CHF is most commonly
defined as a state in which cardiac abnormalities
cause cardiac dysfunction so that the heart is
unable to meet the circulatory demands of the
body or does so with elevated filling pressures.
Clinically, this syndrome causes symptoms of
reduced exercise tolerance or signs of fluid re-
tention. Congestive heart failure commonly is
* Corresponding author.
E-mail address: sgottlie@medicine.umaryland.edu
(S.S. Gottlieb).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.004
a result of systolic dysfunction but can also occur
in the setting of normal systolic dysfunction.
The presentation of decompensated CHF is
variable and ranges from mild dyspnea on exer-
tion to acute, severe pulmonary edema. Critically
ill patients who have acute, cardiogenic pulmo-
nary edema pose the greatest clinical challenge.
The primary role of the emergency physician is to
perform a rapid assessment of the patient, develop
an initial differential diagnosis for entities that
could have led to the decompensation, and de-
termine what therapies are indicated. Patients in
extremis from acute pulmonary edema require the
most aggressive care. In the emergency depart-
ment, treatment strategies are tailored to the
acuity and severity of the CHF exacerbation.
Evaluation and management in the emergency
department
The emergency physician’s role in the stabili-
zation, evaluation, and treatment of the patient
who has decompensated CHF is critical. Al-
though no data exist for a ‘‘golden hour’’ in
treating CHF, a thorough workup, triage, and
treatment strategy initiated by the emergency
physician is likely to have a significant impact
on patient morbidity and mortality. To emphasize
this point, a study by Sacchetti and colleagues [5]
has shown that pharmacologic interventions
started in the emergency department reduce the
need for ICU admission and endotracheal intuba-
tion. Thus, emergency department treatment of
cardiology.theclinics.com
116 ROGERS
the CHF patient has the potential to save money
and lives.
A careful analysis of past medical history and
chief complaint is crucial for accurately diagnos-
ing acute CHF and its potential cause. A clear
understanding of the categories and specific
causes of cardiomyopathies is essential. Broad
categories include dilated, hypertrophic, restric-
tive, and arrythmogenic right ventricle. Specific
causes are ischemic, valvular, hypertensive, in-
flammatory, metabolic, toxic, peripartum, genetic,
and idiopathic (Box 1).
The interview of a patient who has potential
heart failure should include several crucial ques-
tions. The interviewer can quickly and accurately
discover whether the patient has a history of CHF
or risk factors for the development of heart
failure: coronary artery disease, diabetes, hy-
pertension, arrhythmias, valvular disease. It is
also important to verify whether the patient has
symptoms consistent with angina, which could
indicate an acute coronary syndrome as an in-
citing event triggering CHF. In addition, a survey
in search of inciting factors for acute decompen-
sation is warranted. Common factors leading to
CHF include myocardial ischemia or infarct, new-
onset arrhythmias (especially atrial fibrillation),
Box 1. Differential diagnosis of acute
CHF/pulmonary edema
Coronary artery disease
Acute myocardial infarction
Myocardial ischemia
Mechanical complications of acute
myocardial infarction (papillary
muscle rupture)
Valvular disease
Aortic stenosis
Aortic regurgitation
Mitral regurgitation
Mitral stenosis
Myocardial disease
Hypertrophic cardiomyopathy
Dilated cardiomyopathy (eg,
idiopathic, familial)
Myocarditis
Hypertension
Peripartum
Toxic and metabolic (eg, alcohol,
cocaine)
et al
medical noncompliance, and dietary indiscretion.
Less common but certainly well-known inciting
factors include infection and diuretic resistance.
Evaluation of the patient who has acute CHF
(pulmonary edema) begins with a thorough as-
sessment of the patient’s airway, respiratory
status, and circulation. An assessment of the
patient’s airway should be the first step in
management, because patients who have hypox-
emia or altered mental status may require im-
mediate endotracheal intubation before further
workup can proceed. Patients who are compro-
mised but not in need of emergent intubation
can be treated with noninvasive means of venti-
lation such as continuous positive airway pres-
sure (CPAP) or bilevel positive airway pressure
(BiPAP) as a temporizing measure. Large-bore in-
travenous catheters should be placed. Once it has
been established that the airway is secure, an as-
sessment of the patient’s breathing and circulation
can begin.
Acute medical therapy for heart failure should
be considered as distinct from treatment for
chronic heart failure. In the acute treatment of
decompensated heart failure the goals are three-
fold: to stabilize the patient clinically, to normal-
ize filling pressures, and to optimize perfusion to
vital organs. Diuretics, vasodilators, and positive
inotropic agents can be used to achieve these
goals. In severely compromised patients mechan-
ical support with an intra-aortic balloon pump
may be warranted. In contrast, the goals for the
chronic management of heart failure include
improvement of morbidity and mortality.
Stabilization measures, diagnostic testing, and
medical management begin in the emergency
department. Vital signs should be assessed imme-
diately, supplemental oxygen should be adminis-
tered, and the patient should be placed on
a cardiac monitor. All patients should have
a 12-lead ECG obtained upon arrival. A complete
blood cell count, renal function, and electrolytes
should be obtained on all patients. Patients who
have an unclear cause of their dyspnea should
have a B-type natriuretic peptide (BNP) drawn,
because this test may help differentiate CHF from
other pulmonary disease. Patients who have the
potential for ischemia should have cardiac bio-
markers drawn in the emergency department to
assess for myocardial infarction as the precipitant
of decompensated heart failure. If available,
bedside transthoracic echocardiography can be
performed to assess left ventricular function and
to evaluate for entities such as pericardial effusion
 ACUTE CONGESTIVE HEART FAILURE
and valvular dysfunction. There are limited data
on emergency physician–performed bedside ultra-
sound to estimate left ventricular function. A
study by Randazzo and colleagues [6] evaluated
the use of echocardiography performed by emer-
gency physicians. It was found that after a small
amount of focused training emergency physicians
could assess left ventricular ejection fraction accu-
rately in the emergency department. After stabili-
zation of the patient, a search for the cause and
precipitant of CHF/pulmonary edema should be
undertaken, because detection of specific cause
might have a significant impact on treatment.
Currently there are no guidelines for the
management of acutely decompensated heart
failure in the emergency department. DiDomenico
and colleagues [7] published a set of guidelines in
2004 for the initial therapy of CHF in the emer-
gency department. The proposed algorithm relies
on a rapid assessment of the patient’s overall vol-
ume status with the initial treatment approach
aimed at alleviating pulmonary congestion and
improving cardiac output. Patients with milder
degrees of volume overload may respond to intra-
venous diuretics, whereas patients who have low
cardiac output and moderate to severe volume
overload need an approach that combines volume
management, preload and afterload reduction,
and inotropic support.
Noninvasive airway management
In many instances, patients do not require
immediate control of their airway by endotracheal
intubation but might need additional assistance to
decrease their work of breathing. It has been
established that noninvasive ventilation (CPAP or
BiPAP) is an effective means of providing venti-
latory support for critically ill patients who have
acutely decompensated CHF and pulmonary
edema [8]. Noninvasive ventilatory strategies us-
ing modes like CPAP work by limiting decline in
functional residual capacity, improving respiratory
mechanics and oxygenation, and decreasing left
ventricular preload and afterload [9–11]. Two ran-
domized studies have also found beneficial physi-
ologic effects by showing a significant reduction in
endotracheal intubation rates in patients who had
acute cardiogenic pulmonary edema [12,13].
Previously, studies examining noninvasive ven-
tilation have been performed in the ICU when
respiratory failure was already present. In a study
by Nava and colleagues [11], noninvasive pressure
117
support ventilation (NPSV) was compared with
conventional oxygen therapy in the treatment of
acute cardiogenic pulmonary edema. In this mul-
ticenter emergency department study, 130 patients
who had acute respiratory failure secondary to
cardiogenic pulmonary edema were randomly as-
signed to traditional medical therapy including
oxygen (65 patients) or NPSV (65 patients). The
main outcome measured was the need for intuba-
tion. NPSV provided faster improvement in the
ratio of arterial oxygen saturation to inspired ox-
ygen concentration (PaO2/FIO2), respiratory rate,
and dyspnea. Although the rates of intubation,
hospital mortality, and duration of hospitaliza-
tion were similar in the two groups, a subgroup
of hypercapneic patients did benefit from the ther-
apy and had a decreased intubation rate (2 of 33
versus 9 of 31). This study lends credence to the
anecdotal evidence that noninvasive ventilation
is effective in relieving the work of breathing and
in some cases in preventing endotracheal intuba-
tion [11].
B-type natriuretic peptide as a diagnostic tool in
the emergency department
Among the tools available to the emergency
physician for the assessment of the patient who
has undifferentiated dyspnea or suspected heart
failure, no test has been proven as useful as BNP.
This peptide is released into the circulation when
ventricular stress is present and has been shown
to assist in the diagnosis of CHF and to help
differentiate it from other syndromes [14]. The
ability to measure this peptide to aid diagnosis
and to use as a prognostic indicator represents
a major advance in the diagnosis and management
of CHF. Some studies have shown that measure-
ment of BNP can reduce hospitalization rates, re-
duce length of stay, and may help resource use
and possibly even improve survival [15].
There are some limitations of using BNP as
a diagnostic aid in the emergency department.
Interpretation of the test should be used in
conjunction with other clinical data and judgment
and as an adjunct only [16]. Also, a number of
other conditions have been shown to elevate
BNP levels. It should never be assumed that ele-
vated BNP levels alone indicate CHF. Box 2 re-
views the reasons for a falsely elevated BNP (in
the absence of heart failure) and the reasons for
a falsely low BNP.
118 ROGERS
Box 2. Conditions that affect BNP levels
Conditions that can elevate BNP levels
Right-sided heart failure
Pulmonary embolism
Myocardial infarction
Advanced age
Renal failure
Conditions that cause
lower-than-expected BNP levels
Obesity
Acute pulmonary edema
Mitral stenosis
Acute mitral regurgitation
Several studies have investigated the use of
BNP as a diagnostic test. The Breathing Not
Properly (BNP) trial showed that, in patients
presenting to the emergency department with
acute dyspnea, the diagnostic accuracy of BNP
measurement was 81% for a BNP level greater
than 100 pg/mL compared with an accuracy if
74% for clinical judgment [17]. In fact, a BNP
level of 100 pg/mL or greater provides a sensitivity
of 90%, specificity of 76%, positive predictive val-
ue of 79%, and a negative predictive value of
89%. The overall accuracy in the study was deter-
mined to be 81%. A BNP level of 500 pg/mL or
greater has been shown to indicate a 95% proba-
bility of heart failure [18]. Recent results from the
Rapid Emergency Department Heart Failure Out-
patient Trial (REDHOT) showed that BNP levels
might also be useful in the assessment of disease
severity and prognosis. This study evaluated 464
patients presenting to the emergency department
with dyspnea and BNP levels greater than 100
pg/dL. BNP was found to be a predictor of events
and mortality [19].
Volume management
Diuretics are the first-line therapeutic modality
to consider in acute treatment of CHF. Although
diuretics have no proven mortality benefit, they
effectively relieve symptoms of congestion, pul-
monary edema, extremity swelling, and hepatic
congestion. The acute effect of diuretics in pa-
tients who have heart failure–related volume
overload is to reduce left ventricular filling pres-
sures. There is no acute increase in cardiac output.
et al
Many patients who have chronic heart failure
are taking stable doses of loop diuretics as out-
patients. Therefore, it is important in the acute
setting to administer a dose of intravenous di-
uretic to achieve the desired therapeutic effect. In
general, with normal renal function, two times the
oral dose is given intravenously in the acute
setting. With abnormal renal function, two and
half times the oral dose is generally required to
achieve the desired affect.
It should be obvious within 2 hours whether
a patient will respond to initial intravenous di-
uretic therapy. If not, and if diuretic resistance is
suspected, it is likely that the patient will require
hospital admission for other fluid removal ther-
apy. These measures may include escalating doses
and combination diuretic therapy, ultrafiltration,
or parenteral therapy.
Vasodilator therapy
For the patient who presents to the emergency
department with acutely decompensated CHF,
vasodilator therapy should be initiated to achieve
reduction in preload and afterload. The use of
intravenous vasodilators to treat acute heart
failure and pulmonary edema makes sound phys-
iologic sense, because the underlying mechanism
of dyspnea and respiratory distress relates to
elevated filling pressures. Several different forms
of vasodilators are currently available. Most act
to reduce filling pressures and systemic vascular
resistance, thereby increasing cardiac function.
Therapy should be individualized [20,21].
Nitroglycerin
Nitroglycerin traditionally has been the vasodi-
lator of choice in the treatment of acutely decom-
pensated heart failure and pulmonary edema.
Nitroglycerin acts primarily to lower preload by
increasing venous capacitance. The lowered pre-
load in turn reduces ventricular filling pressure
and volume and leads to a decrease in myocardial
oxygen consumption. Nitrates also cause coronary
vasodilatation, which may be beneficial if ischemia
is the underlying precipitant of acute heart failure.
Nitroglycerin’s effect on the arterial side is seen
mainly when high doses are used, in excess of 30 mg
per minute. In cases of acutely decompensated
heart failure and pulmonary edema, nitroglycerin
can be given sublingually while an intravenous
drip is prepared [22]. Use of nitroglycerin should
be considered for any patient who presents with
 ACUTE CONGESTIVE HEART FAILURE
acutely decompensated CHF, particularly in cases
of pulmonary edema and respiratory distress.
Drawbacks to its use include its contraindication
in patients taking phosphodiesterase-5 inhibitors
for erectile dysfunction (sildenafil, tadalafil, varde-
nafil) and side effects such as headache.
Angiotensin-converting enzyme inhibitors
The use of angiotensin-converting enzyme
(ACE) inhibitors to treat acutely decompensated
CHF and pulmonary edema has been a controver-
sial topic among cardiologists and emergency
physicians. There are anecdotal reports of rapid
improvements in patients who had acute cardio-
genic pulmonary edema after of sublingual or
intravenous administration of ACE inhibitors. Is
there, however, any evidence in the literature to
support these reports?
In a study of 24 patients who had acute
cardiogenic pulmonary edema, Haude and col-
leagues [23] showed that significant hemodynamic
changes were induced by the acute administration
of sublingual captopril. Captopril caused an in-
crease in stroke volume and was also found to de-
crease systemic vascular resistance. Although no
firm conclusions can be drawn from this study,
it did indicate that ACE inhibitors could affect pa-
rameters shown to be problematic in patients who
have acute pulmonary edema, namely elevated
peripheral vascular resistance. Hamilton and
colleagues [24] studied the effects of adding an
ACE inhibitor to the usual therapy of oxygen, ni-
trates, morphine, and diuretics (furosemide) in the
treatment of acute pulmonary edema. He found
that the addition of an ACE inhibitor produced
more rapid clinical improvement than the stan-
dard treatment. Again, this study was small and
evaluated only 48 patients. In a placebo-con-
trolled, randomized, double-blind trial of intra-
venous enalapril in patients who had acute
cardiogenic pulmonary edema, Annane and col-
leagues [25] showed that early administration of
intravenous enalapril was effective and well toler-
ated in patients who had acutely decompensated
heart failure. Of course, vasodilation may have
been the cause of the benefit seen in these small
studies. Acute ACE inhibition can cause renal
dysfunction in patients who are intravascularly
depleted, however, and therefore should not be
the vasodilator of choice in patients who have un-
known renal function or intravascular volume
status. To date, there has not been a large, ran-
domized study of ACE inhibitor use in acute
119
pulmonary edema or comparisons with other
vasodilators.
Nitroprusside
Nitroprusside is a potent arterial and venous
vasodilator and can be used to treat acutely
decompensated heart failure in specific circum-
stances. Nitroprusside was one of the first vaso-
dilators used in the management of acute heart
failure. Patients who have acute mitral or aortic
regurgitation may benefit from the use of nitro-
prusside. In addition, the drug is useful in patients
who have cardiogenic pulmonary edema in the
setting of severely elevated blood pressure. Cau-
tion should be exercised, however, if ischemia is
a possible underlying mechanism, because nitro-
prusside administration has been shown to induce
a ‘‘coronary steal’’ phenomenon and worsen
cardiac ischemia. Drawbacks of nitroprusside
include the need for invasive hemodynamic mon-
itoring and the side effects of accumulation of
cyanide and thiocyanate.
Nesiritide
Nesiritide is a vasodilator that has been shown
to decrease pulmonary artery and pulmonary
capillary wedge pressures in patients who have
heart failure [26]. It has also been advocated as an
acute treatment in the emergency department with
the goal of decreasing hospitalization rates. The
argument is that starting active therapy earlier
will lead to quicker resolution of symptoms and
shorter hospital stays.
Unfortunately, beneficial outcomes of nesiri-
tide have never been documented in a randomized,
blinded, controlled study. Indeed, recent studies
have suggested that nesiritide may actually de-
crease survival [27] and increase the rate of renal
dysfunction [28]. Despite a common belief that
renal function and urine output improve with
nesiritide, this improvement has never been docu-
mented in heart failure patients receiving currently
used doses [29].
It is clear that the vasodilation caused by
nesiritide can lead to clinically significant hypo-
tension. Thus, it is contraindicated in patients
who have a systolic blood pressure less than
90 mm Hg. It should be used very cautiously in
patients who have ischemic heart disease, espe-
cially those suspected of having myocardial in-
farction, because hypotension can be particularly
detrimental in these patients.
120 ROGERS
As a vasodilator, nesiritide can clearly decrease
symptoms associated with increased volume and
left ventricular filling pressures. Diuretics and
other vasodilators have the same potential ac-
tions, however. Analyses of large databases have
suggested better outcomes with nesiritide [30], but
these uncontrolled studies with many potential
biases should not be misused to support conclu-
sions that can be provided only by well-designed
investigations. Indeed, the findings of one analysis
that patients not receiving nesiritide were more
likely to be discharged to extended care facilities
suggests that the patients receiving nesiritide
were healthier and different from patients who re-
ceived other care.
At present, nesiritide can be used in patients
who have adequate blood pressure and symptom-
atic heart failure until the effects of more definitive
therapy can take hold. It is expensive, however,
and physicians should not assume that it improves
outcomes or affects renal function.
Inotropes
Inotropic therapy is commonly used to treat
the sickest patients. Although its potential adverse
effects are now well accepted, the mainstay of
treatment of patients who have worsening renal
function or suggestion of other end-organ damage
continues to include dobutamine or milrinone.
This use arises mostly from lack of other options
and has not been supported by studies; inotropic
therapy which increases cAMP by receptor stim-
ulation (dobutamine) or phosphodiesterase inhi-
bition (milrinone) has never been shown to be
beneficial. Fortunately, studies of newer interven-
tions are being undertaken and may provide
pharmacologic options for the sickest patients.
The few randomized studies of inotropic ther-
apy have been disappointing. Chronic therapy has
been shown to be detrimental [31,32], and in-hos-
pital use has also been shown to be of no benefit.
The Outcomes of a Prospective Trial of Intrave-
nous Milrinone for Exacerbations of Chronic
Heart Failure (OPTIME-CHF) study tested the
hypothesis that milrinone given to patients hospi-
talized because of heart failure would lead to
shorter hospitalizations and improved outcomes
as compared with placebo [33]. The study demon-
strated no improvement in patients receiving mil-
rinone, however, and mortality, arrhythmia, and
myocardial infarction rates tended to be worse.
Furthermore, adverse events and the incidence
of sustained hypotension were statistically worse
et al
in patients receiving active drug. OPTIME-CHF
clearly demonstrated that milrinone should not
be used routinely in patients hospitalized for heart
failure.
The applicability of OPTIME-CHF to sicker
patients, however, is uncertain. Investigators did
not randomize patients who were thought to need
inotropic therapy; these patients were given active
drug. Thus, the impact of milrinone in patients
who have worsening renal function or refractory
symptoms is unknown. Although the randomized
studies of chronic inotropic use and the OPTIME-
CHF study led to concern that inotropic therapy
may be detrimental, the OPTIME-CHF data
cannot be extrapolated to the sickest patients.
There are few studies of these patients. Neverthe-
less, retrospective data also raise concern about
the utility of conventional inotropic therapy. Of
course, such data are limited by differences
between groups of patients that cannot be con-
trolled for by any statistical analysis.
Recent controlled data do support the concept
that adrenergic agents and phosphodiesterase
inhibitors may be harmful. Levosimendan is
a novel agent that increases calcium sensitivity.
In one study, the comparison of levosimendan
and dobutamine showed improved survival with
levosimendan [34]. Whether this improved survival
reflects benefits of levosimendan or harm of do-
butamine (or both) is not known. Some studies
demonstrate better outcome with levosimendan
than with placebo and provide hope that levosi-
mendan will prove to increase survival and de-
crease symptoms [35]. The composite outcome
in the ongoing Randomized, Multicenter Evalu-
ation of Intravenous Levosimendan Efficacy
Versus Placebo in the Short Term Treatment
of Decompensated Chronic Heart Failure (RE-
VIVE) study may help illuminate whether levosi-
mendan is truly beneficial.
One problem regarding analyses of older
studies of inotropic therapy is that concomitant
therapies were very different. Those studies were
not performed with patients taking beta-adrener-
gic blocking agents or even ACE inhibitors. It is
certainly possible that the effects of an inotropic
agent will be different in patients receiving modern
therapy. For this reason studies of enoximone are
in process and may lead to a better understanding
of the effects of inotropic therapy.
Despite these concerns, with present knowl-
edge and available agents, physicians appropri-
ately still find it necessary to use dobutamine and
milrinone in selected patients. When these agents
 ACUTE CONGESTIVE HEART FAILURE
are used, a few factors should be considered. First,
because many patients arrive at the emergency
department taking chronic b-blocking therapy, an
agent that is still effective may be preferable. The
effects of dobutamine are more likely to be
impacted by b-blocking agents [36], and milrinone
may be the preferred drug in these patients.
Tolerance to these drugs should also be con-
sidered in patients who have received them for
a prolonged period. Decreased expression of
receptors may lead to decreased contractility (as
compared with before initiation of dobutamine) if
the drug is abruptly stopped. Therefore, any
patient who has been taking dobutamine for
more than 1 day should be weaned off it slowly.
In the sickest patients, weaning may take many
days. Although changes in receptors do not affect
the efficacy of milrinone, drug weaning may also
be needed in patients receiving this drug.
Beta-adrenergic blockers
With the multiple studies showing marked
benefit when beta-adrenergic blockers are given
chronically to patients who have heart failure,
these drugs are occasionally prescribed for acutely
decompensated patients. These patients often have
a tachycardia, which is tempting to treat. Beta-
blockers are negative inotropic, however, and will
decrease contractility. Although their chronic
effect is to improve cardiac function, a dose of
a beta-blocker will impair cardiac performance. In
a decompensated patient, they are likely to lead to
deterioration and should not be used.
A common question is what to do with beta-
adrenergic blockers in patients who present with
worsening heart failure. Although their negative
inotropic properties can certainly decrease con-
tractility in compromised patients, it is also
known that abrupt withdrawal of these agents
can lead to adverse consequences. Furthermore, if
the drugs are not given for a prolonged period,
retitration may take weeks or months. Unfortu-
nately, there are no studies indicating how to deal
with this situation.
Each case must be evaluated individually. A
patient who presents with fluid overload and an
anticipation of rapid improvement with diuresis
probably does not need to have the beta-blocker
withheld. Conversely, giving a beta-blocker to
someone in cardiogenic shock will undoubtedly
make the situation worse. At times, halving the
dose may provide acute help while making it
121
simple to titrate back to a therapeutic dose when
the patient stabilizes.
Of course, someone who deteriorates with
initiation or increasing titration of a beta-blocker
(usually occurring approximately 1 week after the
change) [37] may be helped by decreasing the
dose. Even in some of these patients, however,
all that is needed is diuresis and time to accommo-
date to the new dose.
Some patients present to the emergency room
with primary tachycardia [38]. Atrial fibrillation
with a rapid ventricular response or a supraven-
tricular tachycardia in a patient who has poor car-
diac function may be particularly difficult to treat.
Although a slower rate may improve hemody-
namic parameters, agents that are negative ino-
tropic (such as calcium-channel blockers or
beta-blockers) might cause deterioration. In such
patients, the risk of these agents must be remem-
bered. This situation is ideal for the use of esmo-
lol, which can be tried but discontinued with
immediate reversal of its effects. If improvement
is seen in clinical status and heart rate, a longer-
acting beta-blocker can be given. In contrast, de-
terioration can be easily reversed. If the primary
problem is cardiac dysfunction, calcium-channel
blockers such as diltiazem must be used cautiously,
if at all.
Another agent that can be considered in these
patients is amiodarone, remembering that the
intravenous formulation is a vasodilator and
that blood pressure must be followed carefully.
Although it might be difficult to know if the
tachycardia is the cause of the heart failure or
its consequence, cardioversion should always be
considered in a compromised patient who is
presumed to have primary atrial fibrillation or
supraventricular tachycardia.
Vasoconstrictors
In patients who have heart failure, low blood
pressure is usually the consequence of a decreased
cardiac output. Increasing the blood pressure with
a vasoconstrictor results in further lowering of
cardiac output and worsening of the primary
problem. Thus, vasoconstrictors should be used
only in a patient whose blood pressure is clearly
affecting organ systems, particularly the brain. A
patient who has chronic heart failure without
symptoms of dizziness or lightheadedness rarely
needs a vasoconstrictor. If a higher blood pressure
is clearly needed, vasopressin may increase blood
pressure without directly affecting the heart.
122 ROGERS
When hypotension is present, it is necessary to
see if other problems might be leading to the
deterioration. Sepsis should be considered, and
volume must be assessed. In a patient who has
heart failure, however, the routine administration
of large volumes of fluid for hypotension may
exacerbate the heart failure without increasing
blood pressure. Volume should be given judiciously
and in small boluses to ensure a positive response.
References
[1] Feinglass J, Martin GJ, Lin E, et al. Is heart failure
survival improving? Evidence from 2323 elderly
patients hospitalized between 1989–2000. Am Heart
J 2003;146:111–4.
[2] Ansari M, Massie BM. Heart failure: how big is the
problem? Who are the patients? What does the
future hold? Am Heart J 2003;146:1–4.
[3] Lloyd-Jones DM, Larson MG, Leip EP, et al. Life-
time risk for developing congestive heart failure:
the Framingham Heart Study. Circulation 2002;
106:3068–72.
[4] Acute Decompensated Heart Failure National
Registry. 3rd quarter 2003 national benchmark
report. Available at: http://www.adhereregistry.
com/national_BMR/Q3_2003_ADHERE_National_
BMR.pdf. Accessed January 7, 2005.
[5] Sacchetti A, Ramoska E, Moakes ME, et al. Effect
of ED management on ICU use in acute pulmonary
edema. Am J Emerg Med 1999;17:571–4.
[6] Randazzo MR, Snoey ER, Levitt MA, et al. Accuracy
of emergency physician assessment of left ventricular
ejection fraction and central venous pressure using
echocardiography. Acad Emerg Med 2003;10(9):
973–7.
[7] DiDomenico RJ, Park HY, Southworth MR, et al.
Guidelines for acute decompensated heart failure
treatment. Ann Pharmacother 2004;38:649–60.
[8] Yan AT, Bradley D, Liu PP. The role of continuous
positive airway pressure in the treatment of conges-
tive heart failure. Chest 2001;120:1675–85.
[9] Katz JA, Marks JD. Inspiratory work with and
without continuous positive airway pressure in pa-
tients with acute respiratory failure. Anesthesiology
1985;63:598–607.
[10] Katz J, Kraemer RW, Gjerde GE. Inspiratory work
and airway pressure with continuous positive airway
pressure delivery systems. Chest 1985;4:519–26.
[11] Nava S, Carbone G, DiBattista N, et al. Noninva-
sive ventilation in cardiogenic pulmonary edema:
a multicenter randomized trial. Am J Respir Crit
Care Med 2003;168(12):1432–7.
[12] Bernsten AD, Holt AW, Vedig AE, et al. Treatment
of severe cardiogenic pulmonary edema with contin-
uous positive airway pressure delivered by face
mask. N Engl J Med 1991;325:1825–30.
et al
[13] Lin M, Yang YF, Chiang HT, et al. Reappraisal of
continuous positive airway pressure therapy in acute
cardiogenic pulmonary edema. Short-term results
and long-term follow-up. Chest 1995;107:1379–86.
[14] McCullough PA, Nowak RM, McCord J, et al.
B-type natriuretic peptide and clinical judgment in
emergency diagnosis of heart failure. Circulation
2002;106:416.
[15] Mueller C, Scholer A, Laule-Killian K, et al. Use of
B-type natriuretic peptide in the evaluation and
management of acute dyspnea. N Engl J Med 2004;
350:647–54.
[16] Maisel A. B-type natriuretic peptide measurements
in diagnosing congestive heart failure in the dyspneic
emergency department patient. Rev Cardiovasc
Med 2002;3(Suppl 4):S10–7.
[17] Maisel AS, Krishnaswamy P, Nowak RM, et alfor
the Breathing Not Properly Multinational Study
Investigators. Rapid measurement of B-type natri-
uretic peptide in the emergency diagnosis of heart
failure. N Engl J Med 2002;347:161–7.
[18] Silver MA, Maisel A, Yancy CW, et al. BNP
Consensus Panel 2004: a clinical approach for the
diagnostic, prognostic, screening, treatment moni-
toring, and therapeutic roles of natriuretic peptides
in cardiovascular diseases. Congest Heart Fail
2004;10:1–30.
[19] Maisel A, Hollander JE, Guss D, et al. Primary
results of the Rapid Emergency Department Heart
Failure Outpatient Trial (REDHOT): a multicenter
study of B-type natriuretic peptide levels, emergency
department decision making, and outcomes in pa-
tients presenting with shortness of breath. J Am
Coll Cardiol 2004;44:1328–33.
[20] Nohria A, Lewis E, Stevenson LW. Medical man-
agement of advanced heart failure. JAMA 2002;
287:628–40.
[21] Johnson W, Omland T, Collins CM, et al. Neuro-
hormonal activation rapidly decreases after intrave-
nous therapy with diuretics and vasodilators for
class IV heart failure. J Am Coll Cardiol 2002;39:
1623–9.
[22] Abrams J. Beneficial actions of nitrates in cardiovas-
cular disease. Am J Cardiol 1996;77:31C–7C.
[23] Haude M, Steffen W, Erbel R, et al. Sublingual ad-
ministration of captopril versus nitroglycerin in
patients with severe congestive heart failure. Int J
Cardiol 1990;27:351–9.
[24] Hamilton RJ, Carter WA, Gallagher J. Rapid im-
provement of acute pulmonary edema with sublin-
gual captopril. Acad Emerg Med 1996;3:205–12.
[25] Annane D, Bellissant E, Pussard E, et al. Placebo-
controlled, randomized, double-blind study of in-
travenous enalaprilat efficacy and safety in acute
cardiogenic pulmonary edema. Circulation 1996;
94:1316–24.
[26] Publication Committee for the VMAC Investigators
(Vasodilatation in the Management of Acute CHF).
Intravenous nesiritide vs nitroglycerin for treatment
 ACUTE CONGESTIVE HEART FAILURE
of decompensated congestive heart failure: a ran-
domized controlled trial. JAMA 2002;287:1531–40.
[27] Sackner-Bernstein JD, Kowalski M, Fox M, et al.
Short-term risk of death after treatment with nesiri-
tide for decompensated heart failure: a pooled anal-
ysis of randomized controlled trials. JAMA 2005;
293:1900–5.
[28] Sackner-Bernstein JD, Skopicki HA, et al. Risk of
worsening renal function with nesiritide in patients
with acutely decompensated heart failure. Circula-
tion 2005;111:1487–91.
[29] Wang DJ, Dowling TC, Meadows D, et al. Nesiri-
tide does not improve renal function in patients
with chronic heart failure and worsening serum cre-
atinine. Circulation 2004;110(12):1620–5.
[30] Peacock F, Emerman CL, Wynne J, for the
ADHERE Scientific Advisory Committee and
Investigators and the ADHERE Study Group. Early
use of nesiritide in the emergency department is asso-
ciated with improved outcome: An ADHERE regis-
try analysis. Ann Emerg Med 2004;44(Suppl 4):S78.
[31] Packer M, Carver JR, Rodeheffer RJ, et al, for the
PROMISE Study Research Group. Effect of oral
milrinone on mortality in severe chronic heart fail-
ure: the PROMISE Study Research Group. N Engl
J Med 1991;325:1468–75.
[32] Cohn JN, Goldstein SO, Greenberg BH, et al. A
dose-dependent increase in mortality with vesnari-
none among patients with severe heart failure. Ves-
narinone Trial Investigators. N Engl J Med 1998;
339:1810–6.
123
[33] Cuffe MS, Califf RM, Adams KFJ, et al, for the Out-
comes of a Prospective Trial of Intravenous Milri-
none for Exacerbations of Chronic Heart Failure
(OPTIME-CHF) investigators. Short-term intrave-
nous milrinone for acute exacerbation of chronic
heart failure: a randomized controlled trial. JAMA
2002;287:1541–7.
[34] Follath F, Cleland JG, Just H, et al. Efficacy and
safety of intravenous levosimendan compared with
dobutamine in severe low-output heart failure
(the LIDO study): a randomized double blind trial.
Lancet 2002;360:196–202.
[35] Moiseyev VS, Poder P, Andrejevs N, et alfor the
RUSSLAN Study Investigators. Safety and efficacy
of a novel calcium sensitizer, levosimendan, in pa-
tients with left ventricular failure due to an acute
myocardial infarction. A randomized, placebo-
controlled, double-blind study (RUSSLAN). Eur
Heart J 2002;23:1422–32.
[36] Lowes BD, Tsvetkova T, Eichhorn EJ, et al. Milri-
none versus dobutamine in heart failure subjects
treated chronically with carvedilol. Int J Cardiol
2001;81:141–9.
[37] Gottlieb SS, Fisher ML, Kjekshus J, et al, on behalf
of the MERIT-HF Investigators. Tolerability of
beta-blocker initiation and titration in MERIT-
HF. Circulation 2002;105:1182–8.
[38] Shinbane JS, Wood MA, Jensen DN, et al. Tachy-
cardia-induced cardiomyopathy: a review of animal
models and clinical studies. J Am Coll Cardiol 1997;
29:709–15.
 Cardiol Clin 24 (2006) 125–133
Management of Arrhythmias in the Emergency
Department
R.E. Hood, MD, Stephen R. Shorofsky, MD, PhD*
Division of Cardiology, Department of Medicine, University of Maryland School of Medicine,
22 South Greene Street, Baltimore, MD 21201, USA
Little raises the anxiety level of a physician more
than a call to the emergency department about
a patient who has an arrhythmia. Despite devoting
significant time in training to the mechanisms and
treatment of cardiac rhythm disturbances, most
physicians are uncomfortable dealing with them.
This article strives to provide practical and concise
advice on initial diagnosis and management of
arrhythmias that present to the emergency depart-
ment. There has been no attempt to be exhaustive in
the descriptions. Common pitfalls and concerns are
addressed. The article is divided into a discussion of
tachyarrhythmias, both wide and narrow complex,
bradyarrhythmias, and management of arrhythmia
devicesdpacemakers and defibrillators. Finally,
no account of emergency department arrhythmias
can be complete without some mention of syncope.
The acute treatment of cardiac rhythm dis-
turbances is quite simple if one follows a few basic
rules (Box 1). First, if the presenting arrhythmia is
fast, and the patient is hemodynamically unstable,
regardless of the arrhythmia mechanism, perform
a direct current transthoracic cardioversion to re-
store normal sinus rhythm. The patient should ei-
ther be unconscious or sedated before delivery of
the energy. Second, if the patient presents with
a bradycardia that is hemodynamically unstable,
pace the heart either transvenously or transtho-
racically. The former is a more reliable method.
For all other arrhythmias in which the patient is
hemodynamically stable, obtain a 12-lead ECG;
there is time to think before acting. By following
* Corresponding author.
E-mail address: sshorofsky@medicine.umaryland.edu
(S.R. Shorofsky).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.005
these two simple rules, all acute arrhythmic emer-
gencies can be handled appropriately.
Tachycardias
Tachycardias can be divided into two catego-
ries based on the width of the QRS complexes,
narrow complex (QRS duration % 120 milli-
seconds) or wide complex (QRS duration O 120
milliseconds) tachycardias. When determining the
QRS width, it is important to use at least two
orthogonally placed lead systems. A single-lead
rhythm strip is often inadequate. In any given
single-lead recording, a wide complex tachycardia
may appear narrow (Fig. 1). If the arrhythmia has
a narrow QRS complex, it is by definition a supra-
ventricular tachycardia (SVT). These arrhythmias
are usually benign, and often the patient can be
treated and discharged from the emergency de-
partment to complete the evaluation as an outpa-
tient. If the tachycardia has a wide QRS complex,
it is either a ventricular tachycardia (VT) or, ex-
traordinarily rarely, a SVT with aberrant conduc-
tion or pre-excitation. In contrast to the narrow
complex tachycardias, these tachycardias are usu-
ally malignant and require hospitalization for fur-
ther treatment.
The differential diagnosis between a SVT with
aberration and VT has fascinated physicians for
years. Several algorithms have been proposed to
differentiate between these arrhythmias based on
the 12-lead ECG pattern [1–3]. It seems as though
physicians try to demonstrate their diagnostic
acumen by diagnosing SVT with aberration in pa-
tients who present with wide complex tachycar-
dias. The reality, however, is that in attempting
to prove their diagnostic skills, physicians who
cardiology.theclinics.com
126 HOOD & SHOROFSKY

diagnose a wide complex tachycardia as SVT with Supraventricular tachycardias

aberration are wrong most of the time [4]. The
Many articles and reviews have been written
correct diagnosis is important, because medica-
about the diagnosis and management of SVT. The
tion given for the treatment of SVT may be harm-
reader is referred to the American College of
ful or fatal to a patient in VT [5–7].
Cardiology/American Heart Association/European
In addition to characterizing the tachycardia
Society of Cardiology guidelines for a complete
by its QRS morphology, other important clinical
review [13]. This article focuses on a simplified,
and ECG information can be helpful in diagnos-
practical approach to diagnosing and treating
ing the arrhythmia and treating the patient.
these common arrhythmias.
Dissociation between the atrium and ventricle is
When approaching a patient who has SVT, it is
diagnostic of VT but is recognized only about
helpful to have in mind the different arrhythmias
25% of tracings [4]. The converse, however, is not
that cause narrow complex tachycardias. An easy
true, because ventriculo-atrial association is com-
framework on which to base a differential di-
mon during VT. Other morphologic features that
agnosis is to think anatomically, beginning where
suggest VT include a QRS width greater than 160
the heartbeat starts at the sinus node and ending
milliseconds, the QRS axis in the frontal plane
at the atrioventricular (AV) node. SVT can be
(far left or right), concordance of the QRS com-
caused by sinus tachycardias (appropriate, inap-
plexes across the precordium, and a time from
propriate, or reentrant), atrial tachycardias, atrial
the onset of the R wave to the nadir of the S
flutter, atrial fibrillation, junctional tachycardia,
wave in any precordial lead of more than 100
AV nodal reentry, or AV reciprocating tachy-
milliseconds [1,2,8,9]. Clinical data are also useful
cardia using a bypass tract. It is important to
for establishing the correct diagnosis. A history of
diagnose the arrhythmia mechanism, because
a previous myocardial infarction or structural
treatment is specific for individual arrhythmias.
heart disease is highly suggestive that the wide
For example, appropriate sinus tachycardia is
complex tachycardia is VT [8,10,11]. A careful his-
treated by identifying the cause for the arrhyth-
tory and physical examination often will be suffi-
mia, such as fever, hypovolemia, or hyperadrener-
cient to establish the presence or absence of
gic states rather than simply slowing the heart
structural heart disease in the patient. If this infor-
rhythm with beta-blockers.
mation is insufficient, a transthoracic echocardio-
Patients who have SVT usually present with
gram will define the patient’s cardiac structure
palpitations. They notice a rapid heart rate and
and function. Lastly, hemodynamic stability dur-
seek medical attention if it persists. They usually
ing the wide complex tachycardia is often assumed
give a history of paroxysmal symptoms that have
to be evidence that the tachycardia is caused by
been present for years. Initial evaluation in the
SVT with aberration or pre-excitation. This as-
emergency department includes an ECG, history,
sumption is clearly incorrect. Steinman and col-
and complete physical examination. If symptoms
leagues [12] demonstrated that VT is the correct
persist, a Valsalva maneuver or carotid sinus
diagnosis in 85% of patients who present with
massage, provided there is no carotid disease,
a wide complex tachycardia without significant
might be useful in either terminating the arrhyth-
hemodynamic compromise.
mia or causing the loss of ventricular capture for
a single beat to allow diagnosis of the arrhythmia
Box 1. Rules for the acute management mechanism [14]. If the arrhythmia is not termi-
of arrhythmias nated by these bedside maneuvers, adenosine (6 or
12 mg intravenously) will usually cause AV nodal
If heart rhythm is fast and the patient is block, either terminating the arrhythmia or allow-
hemodynamically unstable, shock the ing the visualization of the underlying mechanism.
patient as quickly as possible and This drug has a rapid onset and short duration of
keep shocking until normal sinus action with a half-life of less than 10 seconds. It is
rhythm is restored. important to administer the adenosine rapidly and
If the heart rhythm is slow and the to follow the infusion with a flush of saline to af-
patient is symptomatic, pace the heart fect a result. The usual side effects of flushing,
with either a transthoracic or dyspnea, and chest discomfort are usually short
transvenous pacemaker. lived [15,16]. Severe complications are extraordi-
narily rare, but case reports of prolonged asystole,
 MANAGEMENT OF ARRHYTHMIAS
Fig. 1. Rhythm strips from a patient. (A) A rhythm strip from lead 2, which has a QRS duration of 70 milliseconds.
(B) The simultaneous rhythm strip from lead V1. The QRS duration in this strip is 140 milliseconds.
VT and fibrillation, and bronchospasm have
been published. Resuscitation capability must be
immediately available. There are additional con-
siderations to remember about adenosine. The
methylxanthines, such as theophylline and caf-
feine, are competitive antagonists. Dipyridamole
may potentiate adenosine’s effects. The dener-
vated, transplanted heart is supersensitive to ad-
enosine [17].
Other intravenous AV nodal blocking agents
such as beta-blockers or calcium-channel blockers
(verapamil or diltiazem) have been used to treat
SVT acutely. These drugs, however, have longer
half-lives than adenosine and cause other prob-
lems such as hypotension (verapamil) or severe
bradycardia (beta-blockers). Because of these po-
tential complications, they have been largely
replaced by adenosine [18].
Most SVTs are benign and can be treated in
the emergency department, and the patient can be
discharged home. The exception is atrial fibrilla-
tion in a patient who has Wolff-Parkinson-White
syndrome who has rapid conduction in the bypass
tract from the atrium to the ventricle. This
arrhythmia presents as an irregularly irregular
arrhythmia with QRS complexes of variable
duration (Fig. 2). This arrhythmia may present
with hemodynamic compromise if the ventricular
127
rate is sufficiently fast. As mentioned earlier, any
SVT with hemodynamic compromise should be
treated by electrical cardioversion. This presenta-
tion is also the one case when calcium-channel
blocking drugs and adenosine should be avoided.
These drugs often increase conduction through
the bypass tract while decreasing it through the
AV node, thus increasing the ventricular response
rate and causing hemodynamic compromise [5].
The ideal drug for this condition is intravenous
procainamide, which slows the ventricular rate
by blocking conduction through the bypass tract.
Most SVTs are caused by a reentrant circuit
involving the AV node or use the AV node to
conduct to the ventricle. Therefore, long-term
treatment is directed at decreasing conduction
through the AV node. Beta-blockers and calcium-
channel blockers are particularly useful for
treating these arrhythmias. Once patients have
been stabilized in the emergency department,
they can usually be discharged on either a long-
acting beta-blocker or calcium-channel blocker
to complete their evaluation in the outpatient
clinic. If patients have recurrent episodes des-
pite these medications, a referral to the electro-
physiologist for consideration of an ablation
is warranted. Most SVTs can be cured with an
ablation [18].
128 HOOD & SHOROFSKY

Fig. 2. Pre-excited atrial fibrillation.

Atrial flutter and atrial fibrillation Of special consideration is the patient who
presents with atrial fibrillation and an acute
Atrial flutter and atrial fibrillation represent
myocardial infarction. Although the diagnosis is
a special challenge to the emergency department
commonly considered, few patients who present
physician. Both arrhythmias may present as in-
with new-onset atrial fibrillation are having an
cidental findings in a patient who presents to the
acute myocardial infarction. In fact, when atrial
emergency department for other complaints. These
fibrillation occurs during an acute myocardial
arrhythmias are common in the general popula-
infarction, it is an indication of significant myo-
tion, with the incidence increasing with advancing
cardial damage and pulmonary congestion. The
age [19,20]. In the emergency department, the same
treatment of choice is diuresis and hemodynamic
rules apply as for other supraventricular arrhyth-
support rather than rate control with AV nodal
mias. Cardioversion is the treatment of choice if
blocking agents [28,29].
the patient is hemodynamically compromised.
Ibutilide, an intravenous class III anti-arrhyth-
If the patient is hemodynamically stable, control
mic, has been used successfully for conversion of
the ventricular response rate with AV nodal block-
atrial fibrillation and atrial flutter, particularly
ing medications such as beta-blockers and cal-
when they are of short duration, and can be
cium-channel blockers. Digoxin is a third-line
considered for acute cardioversion in the emer-
drug, which may be useful in patients who have
gency department. Because of its potential side
left ventricular dysfunction. The AV nodal block-
effects, it has been recommended that ibutilide be
ing properties of digoxin are caused by an increase
limited to patients who have an ejection fraction
in vagal tone and thus are not useful for a physically
greater than 30%. There is an increased risk of
active patient. It is important to begin anticoa-
torsades de pointes with lower ejection fractions,
gulation in patients who present with atrial
which may be minimized by pretreatment with
fibrillation or atrial flutter because these arrhyth-
magnesium [30–32].
mias predispose the patient to thromboembolic
events [21–25]. Anticoagulation recommendations
are well summarized and have been recently up-
Ventricular tachycardia
dated [26]. Recently, a consensus agreement has
been proposed on a treatment scheme for these ar- Again, the management of VT should not
rhythmias in both acute and chronic settings [27]. be frightening to the emergency department
 MANAGEMENT OF ARRHYTHMIAS
physician. Most patients who present with these
arrhythmias have underlying heart disease. If the
patient is hemodynamically unstable, perform an
electrical cardioversion as quickly as possible. The
rate of survival depends on the speed of cardio-
version [33]. Support the patient hemodynami-
cally following advanced cardiac life support
protocols. In the acute phase of treatment, the in-
travenous drug of choice is amiodarone. When
given to patients in the ambulance, amiodarone
improves survival to admission to the hospital
when compared with lidocaine [34]. Amiodarone
should be administered as a 150-mg bolus, fol-
lowed by an infusion at 1 mg/minute for 6 hours,
decreasing to 0.5 mg/minute subsequently. One
complication that is specific to intravenous amio-
darone is hypotension, secondary to the detergent,
polysorbate 80, used to dissolve the drug [35].
Other medications that can be used if more readily
available are lidocaine and procainamide, the
latter being useful when atrial fibrillation with
pre-excitation is suspected. Polymorphic VT or
torsades de pointes responds to intravenous mag-
nesium (1–5 g), and VT storm often responds to
intravenous beta-blockers and often with amio-
darone supplementation [36–39]. Once stabilized,
all patients who have VT should be admitted to
the hospital for further evaluation and treatment.
Cardioversion
Cardioversion refers to the process of terminat-
ing arrhythmias either by pharmacologic agents,
direct current, or a combination of both. Cardio-
version is used to restore normal sinus rhythm to
relieve symptoms and reduce the heart rate. Stroke
risk must be assessed before every cardioversion.
Hemodynamic stability provides opportunity to
choose therapies, drug versus electrical. Pharma-
cologic cardioversion has the advantage of not
requiring sedation, which may a particular concern
in patients who have severe respiratory disease.
Drugs commonly used for cardioversion are those
already mentioned for the treatment of VT and
SVT. Direct current cardioversion, however, is the
most efficacious means for restoring normal
rhythm. In the past, investigative efforts directed
toward the optimal vector, pad/paddle size, and
energy requirement have been the subject of much
discussion. The one substantial improvement has
been the introduction of biphasic shocks. There
continues to be controversy over the ideal wave-
form, but from a practical perspective, the current,
commercially available external defibrillators are
129
all highly effective. Skin erythema is common.
Tachyarrhythmias, bradyarrhythmias, and pulmo-
nary edema have been rarely reported complica-
tions. If cardioversion is not successful, repeating
with increased power, changing the vector, use of
paddles with manual compression, or some com-
bination often is successful.
Syncope
Syncope is the sudden, transient loss of con-
sciousness and postural tone with spontaneous
recovery, most often caused by generalized cere-
bral hypoperfusion [40]. Loss of consciousness
must involve either both cerebral hemispheres or
the reticular activating system of the brainstem.
It is the result of decrease in cardiac output or
loss of vascular tone or both.
The differential diagnoses of altered mental
status include syncope, seizure, obtundation, de-
lirium, dementia, coma, change in postural tone,
and drop attacks. It is not regional hypoperfusion,
which may present as transient ischemic attack or
stroke. It may be confused with dizziness, vertigo,
disequilibrium, or lightheadedness. Syncope is
rarely caused by acute coronary syndrome or
myocardial infarction with the possible exception
of the very elderly or when an arrhythmia occurs
[41,42]. Syncope is not caused by hypoglycemia.
Causes for cardiovascular syncope include (1)
circulatory obstruction such as that occurring
with critical aortic stenosis, hypertrophic obstruc-
tive cardiomyopathy, and pulmonary embolism;
(2) arrhythmia such as VT, atrial flutter with
1:1 conduction, or occasionally bradycardia;
(3) orthostatic intolerance associated with the
autonomic nervous system including neurocardio-
genic syncope, postural orthostatic tachycardia
syndrome, and primary and secondary autonomic
failure [43].
The patient’s history provides the best means of
elucidating the origin of a syncopal spell [44]. A de-
scription of a prodrome such as palpitations,
blurred vision, nausea, warmth, diaphoresis, or
lightheadedness is important, as are postevent
symptoms such as nausea, warmth, diaphoresis,
and fatigue. Activity, position, recurrence, family
history of syncope, sudden death, unexplained acci-
dents including drowning and motor vehicle acci-
dents, and epilepsy are all relevant [44]. The
physical examination should be directed to ortho-
static vital signs, murmurs, prominent P2, gallops,
and other evidence of compromised ventricular
function. The 12-lead ECG should be examined
130 HOOD & SHOROFSKY
for evidence of heart block, ectopy, prior myocar-
dial infarction, and the QT interval. Very rarely,
classic findings of a delta wave, epsilon wave, or
Brugada syndrome will suggest a cause. The selec-
tive use of cardiac diagnostic tests should be consid-
ered. These include echocardiography, tilt table
testing, ambulatory monitoring, or electrophysio-
logic studies. Without specific neurologic findings,
extensive neurologic testing has a low yield and gen-
erally is not warranted. Another pressing question
is when to hospitalize a patient who has syncope.
The Task Force on Syncope from the European So-
ciety of Cardiology recommended hospitalization
of those with (1) known or suspected significant
heart disease, (2) ECG abnormalities suspected of
arrhythmic syncope, (3) syncope during exercise,
(4) syncope causing severe injury, and (5) a family
history of sudden death and recommends that occa-
sionally others be admitted to the hospital for fur-
ther evaluation [45].
The outcome of patients who have syncope
is largely dependent on the cause. Those with
a cardiac cause fare the worst. Patients who
have a noncardiac cause have an intermediate
risk, and those with neurocardiogenic syncope
have a mortality rate of nearly nil. Ambulatory
outcome data from the Framingham study were
collected on 2336 men and 2873 women whose
ages were between 30 and 62 years at entry. A 26-
year surveillance revealed that 3.0% of men and
3.5% of women had at least one syncopal episode
In 79% of men and 88% of women who experi-
enced isolated syncope, there was no excess risk of
stroke or myocardial infarction, and there was no
excess all-cause or cardiovascular mortality [46].
Quinn and colleagues [47] described 684 visits to
San Francisco emergency departments for syn-
cope. Predictors of serious outcome (defined as
death, myocardial infarction, arrhythmia, pulmo-
nary embolism, stroke, subarachnoid hemor-
rhage, or repeat emergency department visit with
hospitalization at day 7) include abnormal ECG,
dyspnea, hematocrit below 30%, systolic blood
pressure below 90 mm Hg, and history of conges-
tive heart failure.
In certain populations, syncope can also be
harbinger of increased mortality. Kapoor and
colleagues [48] reported on 204 patients who pre-
sented with syncope to the emergency depart-
ments, hospitals, or clinics. Causes of mortality
at 1 year were 30% cardiovascular, 12% noncar-
diovascular, and 6.4% of unknown categories.
He further reported on mortality at 24 months
in a 2  3 matrix of elderly (60–90 years) and
young (15–59 years) persons. Mortality from
cardiovascular syncope was 38.1% and 32.5%,
respectively. Noncardiovascular syncope had
a mortality of 21.6% and 4.7%, respectively,
and unknown causes had 20.4% and 2.5%, re-
spectively [49]. Using a risk score which assigned
one point each for age above 65 years, history of
cardiovascular disease, syncope without pro-
drome, and abnormal ECG, Colivicchi and col-
leagues [50] reported that the mortality of 270
patients who presented to the emergency depart-
ment was 0.8%, 20%, 35%, and 57% for patients
assigned 1 through 4 points, respectively [50].
Implanted cardiac devices
Besides patients who have active arrhythmias,
the other large group that presents to the emergency
department is those who have implanted cardiac
devices. They may have unique complaints related
to their hardware. Three types of devices are
discussed here: pacemaker, implantable cardi-
overter defibrillators (ICDs), and insertable loop
recorders. Devices are most commonly found in the
left pectoral fossa. The right pectoral fossa and
abdomen are alternate sites. Very rarely infra-
mammary or inguinal placement is used. At the
time of implantation, the device manufacturer
sends the patient an identification card that pro-
vides basic information such as manufacturer,
model numbers, serial numbers, and telephone
numbers for contact. All patients are strongly
encouraged to carry this card with them at all times.
Device-related implant problems that might
appear in the acute setting include pocket hema-
toma, wound dehiscence, upper extremity deep
venous thrombosis, pneumothorax, and lead dis-
lodgement. Chronic device complications include
battery exhaustion and catastrophic, random
component failure. Time-independent problems
include oversensing which results in underpacing
or inappropriate shocks, undersensing producing
overpacing and failure of ICD therapy, infection,
and erosion.
Chief complaints that are possibly device related
include syncope, near syncope, palpitations, and
infections. Of course there are a plethora of chief
complaints that are not device related but still
impact care. For instance, patient who has an
implanted cardiac device cannot undergo a MRI
scan at present. CT scans are safe, but the metal
may produce some scatter in chest compro-
mising image quality. If emergent surgical proce-
dures are contemplated, the device may require
 MANAGEMENT OF ARRHYTHMIAS
reprogramming or magnet placement to use elec-
trocautery safely. An American Heart Association
Scientific Advisory was published in 2004 on ICD
management for the non-electrophysiologist [51].
Permanent pacemakers at a minimum pace at
least one chamber although typically they have
many more functions. Pacemakers are conven-
tionally described by the NBG code. (The acro-
nym NBG stands for the North American Society
for Pacing and Electrophysiology [NASPE] and
British Pacing and Physiology Group [BPEG]
Generic.) An abbreviated description of the code
identifies the chamber(s) paced (first letter), cham-
ber(s) sensed (second letter), pacemaker response
(third letter), and rate responsiveness (fourth
letter). A is for atrium; V is ventricle; D is dual
chamber or dual function that is both triggered
and inhibited; I equals inhibited, and R is for rate
responsiveness. Thus a VVIR pacemaker paces
the ventricle and senses in the ventricle; its output
is inhibited by a sensed event, and it is rate
responsive, that is, the motion sensor is active.
The most common designations seen are DDD,
DDDR, VVI, and VVIR.
When a patient who has a pacemaker presents
to the emergency department, the first concern is
whether pacemaker malfunction could account
for the presenting symptoms. If the patient has the
device checked regularly, and the ECG does not
demonstrate a problem, the likelihood of a pace-
maker malfunction is quite small. The reliability
of pacemakers is such that, if the above conditions
are met, the device is almost certainly functioning
as programmed. It is certainly reasonable to
interrogate the pacemaker, but one must prepare
for alternative causes.
All ICDs have defibrillation capability (shock
therapy) as well as ventricular anti-bradycardia
pacing. They often have additional features such
as dual-chamber pacing, biventricular pacing, and
atrial defibrillation capabilities as well a variety of
diagnostic recording capabilities. If a patient
presents to the emergency department with a car-
diac arrest, advanced cardiac life support proto-
cols should be followed as if the patient does not
have a defibrillator. If the device has not termi-
nated a ventricular arrhythmia, external defibril-
lation should be performed. The only caveat is to
avoid placing the pads or paddles directly over the
pulse generator. Contact with the patient as an
internal shock is delivered by the ICD will not
cause harm the emergency department personnel.
The most disconcerting situation for both pa-
tient and physician is that of a patient who presents
131
with multiple ICD shocks. The physician needs to
assess whether the shocks were appropriate or
spurious. Repetitive, appropriate shocks indicate
electrical storm. One must search for a cause such as
active ischemia, electrolyte abnormalities, or de-
terioration of left ventricular function. This de-
termination must be concurrent with management
as described above for ventricular storm. Spurious
shocks could occur as the result of rapid, non-
malignant rhythms such as sinus tachycardia or
rapid atrial fibrillation or a malfunctioning device
(noise on the rate sensing lead). In this situation, it is
important to deactivate or inhibit ICD therapy.
Deactivation is done with a magnet placed over the
generator or by reprogramming the device. In
addition to the cardiac issues, the psychologic stress
of multiple shocks should be addressed.
The insertable loop recorder is a leadless device
implanted subcutaneously in the left pectoral
fossa. The purpose of this device is to record
spontaneous arrhythmias and triggered events.
It has service life in excess of 1 year and no
therapeutic capability. Except for recognizing that
it is neither a pacemaker nor a defibrillator, it is
not discussed here.
Magnets and devices
Each manufacturer produces a doughnut-
shaped magnet for use with devices. Pacemakers
and defibrillators have different responses to a
magnetic field. The typical response of a pace-
maker is asynchronous pacing, that is, VOO or
DOO. It does not sense and does not track or
inhibit; it simply paces. Defibrillators’ anti-tachy-
cardia therapies (shocks and anti-tachycardia
pacing) are inhibited by a magnet. The defibrilla-
tor’s anti-bradycardia pacing function is not
affected by a magnet.
Making the situation more confusing, some
models of pacemakers and defibrillators can have
their typical magnet responses disabled, in which
case the effects described above will not occur.
It is uncommon to program a device with the
magnet function disabled. Recently, however,
one manufacturer has recommended that the
magnet function on several models of defibrillator
be disabled as a means for dealing with a
manufacturing defect.
References
[1] Brugada P, Brugada J, Mont L, et al. A new ap-
proach to the differential diagnosis of a regular
132 HOOD & SHOROFSKY
tachycardia with a wide QRS complex. Circulation
1991;83:1649–59.
[2] Wellens HJJ, Bar FWHM, Lie KI. The value of the
electrocardiogram in the differential diagnosis of
a tachycardia with a widened QRS complex. Am J
Med 1978;64:27–33.
[3] Marriott HJL. Differential diagnosis of subraven-
tricular and ventricular tachycardia. Geriatrics
1970;25:91–101.
[4] Akhtar M, Shenasa M, Jazayeri M, et al. Wide QRS
complex tachycardia. Reappraisal of a common clin-
ical problem. Ann Intern Med 1988;109:905–12.
[5] Stewart RB, Bardy GH, Greene HL. Wide complex
tachycardia: misdiagnosis and outcome after emer-
gent therapy. Ann Intern Med 1986;104:766–71.
[6] Buxton AE, Marchlinski FE, Doherty JU. Hazards
of intravenous verapamil for sustained ventricular
tachycardia. Am J Cardiol 1987;59:1107–10.
[7] Dancy M, Camm AJ, Ward D. Misdiagnosis of
chronic recurrent ventricular tachycardia. Lancet
1985;2:320–3.
[8] Wellens HJJ. Ventricular tachycardia; diagnosis of
a broad QRS complex tachycardia. Heart 2001;86:
579–85.
[9] Steurer G, Gursoy S, Frey B, et al. The differential
diagnosis on the electrocardiogram between ventric-
ular tachycardia and preexcited tachycardia. Clin
Cardiol 1994;17:306–8.
[10] Griffith MJ, de Belder MA, Linker NJ, et al. Multi-
variate analysis to simplify the differential diagnosis
of broad complex tachycardia. Br Heart J 1991;66:
166–74.
[11] Tchou P, Young P, Mahmud R, et al. Useful clinical
criteria for the diagnosis of ventricular tachycardia.
Am J Med 1988;84:53–6.
[12] Steinman RT, Herrera C, Schuger CD, et al. Wide
QRS tachycardia in the conscious adult. Ventricular
tachycardia is the most frequent cause. JAMA 1989;
261:1013–6.
[13] Blomstrom-Lundqvist C, Scheinman MM, Aliot
EM, et alfor the European Society of Cardiology
Committee, NASPE-Heart Rhythm Society. ACC/
AHA/ESC guidelines for the management of
patients with supraventricular arrhythmiasdexecu-
tive summary. A report of the American College of
Cardiology/American Heart Association task force
on practice guidelines and the European Society of
Cardiology committee for practice guidelines (writ-
ing committee to develop guidelines for the manage-
ment of patients with supraventricular arrhythmias)
developed in collaboration with NASPE-Heart
Rhythm Society. J Am Coll Cardiol 2003;42(8):
1493–531.
[14] Mehta D, Wafa S, Ward DE, et al. Relative efficacy
of various physical manoeuvres in the termination of
junctional tachycardia. Lancet 1988;1:1181–5.
[15] Lerman BB, Belardinelli L. Cardiac electrophysio-
logy of adenosine: basic and clinical concepts. Circu-
lation 1991;83:1499–509.
[16] Camm AJ, Garratt CJ. Adenosine and supraventric-
ular tachycardia. N Engl J Med 1991;325:1621–9.
[17] Ellenbogen KA, Thames MD, DiMarco JP, et al.
Electrophysiological effects of adenosine in the
transplanted human heart. Evidence of supersensi-
tivity. Circulation 1990;81(3):821–8.
[18] Ganz LI, Friedman PL. Supraventricular tachycar-
dia. N Engl J Med 1995;332:162–73.
[19] Camm AJ, Obel OA. Epidemiology and mechanism
of atrial fibrillation and atrial flutter. Am J Cardiol
1996;78(8A):3–11.
[20] Feinberg WM, Blackshear JL, Laupacis A, et al.
Prevalence, age distribution, and gender of patients
with atrial fibrillation. Analysis and implications.
Arch Intern Med 1995;155:469–73.
[21] Stroke Prevention in Atrial Fibrillation Study. Final
results. Circulation 1991;84:527–39.
[22] The Boston Area Anticoagulation Trial for Atrial
Fibrillation Investigators. The effect of low-dose
warfarin on the risk of stroke in patients with non-
rheumatic atrial fibrillation. N Engl J Med 1990;
323:1505–11.
[23] Connolly SJ, Laupacis A, Gent M, et al. Canadian
Atrial Fibrillation Anticoagulation (CAFA) Study.
J Am Coll Cardiol 1991;18:349–55.
[24] Petersen P, Boysen G, Godtfredsen J, et al. Placebo-
controlled, randomised trial of warfarin and aspirin
for prevention of thromboembolic complications
in chronic atrial fibrillation. The Copenhagen
AFASAK study. Lancet 1989;1:175–8.
[25] Ezekowitz MD, Bridgers SL, James KE, et al. War-
farin in the prevention of stroke associated with
nonrheumatic atrial fibrillation. Veterans Affairs
Stroke Prevention in Nonrheumatic Atrial Fibril-
lation Investigators. N Engl J Med 1992;327:
1406–12.
[26] Singer DE, Albers GW, Dalen JE, et al. Antithrom-
bic therapy in atrial fibrillation: the seventh ACCP
Conference on Antithrombic and Thrombolytic
Therapy. Chest 2004;126:429S–56S.
[27] Fuster V, Ryden LE, Asinger RW, et al. ACC/AHA/
ESC guidelines for the management of patients with
atrial fibrillation: executive summary. A report of
the American College of Cardiology/ American
Heart Association Task Force on Practice Guide-
lines and the European Society of Cardiology
Committee for Practice Guidelines and Policy Con-
ferences (Committee to Develop Guidelines for the
Management of Patients With Atrial Fibrillation:
developed in collaboration with the North American
Society of Pacing and Electrophysiology. J Am Coll
Cardiol 2001;38(4):1231–66.
[28] Asanin M, Perunicic J, Mrdovic I, et al. Prognostic
significance of new atrial fibrillation and its relation
to heart failure following acute myocardial infarc-
tion. Eur J Heart Fail 2005;7(4):671–6.
[29] Lehto M, Snapinn S, Dickstein K, et al. for the OPTI-
MAAL investigators. Prognostic risk of atrial fibril-
lation in acute myocardial infarction complicated by
 MANAGEMENT OF ARRHYTHMIAS
left ventricular dysfunction: the OPTIMAAL experi-
ence. Eur Heart J 2005;26(4):350–6.
[30] Oral H, Souza JJ, Michaud GF, et al. Facilitating
transthoracic cardioversion of atrial fibrillation
with ibutilide pretreatment. N Engl J Med 1999;
340(24):1849–54.
[31] Kalus JS, Spencer AP, Chung J, et al. Does magne-
sium prophylaxis alter ibutilide’s therapeutic efficacy
in atrial fibrillation patients? [absract]. Circulation
2002;106(19 Suppl II):634.
[32] Kalus JS, Spencer AP, Tsikouris JP, et al. Impact of
prophylactic i.v. magnesium on the efficacy of ibuti-
lide for conversion of atrial fibrillation or flutter. Am
J Health Sys Pharm 2003;60(22):2308–12.
[33] De Maio VJ, Stiell IG, Wells GA, et al, for the
Ontario Prehospital Advanced Life Support Study
Group. Optimal defibrillation response intervals
for maximum out-of-hospital cardiac arrest survival
rates. Ann Emerg Med 2003;42(2):242–50.
[34] Dorian P, Cass D, Schwartz B, et al. Amiodarone as
compared with lidocaine for shock-resistant ventric-
ular fibrillation. N Engl J Med 2002;346:884–90.
[35] Munoz A, Karila P, Gallay P, et al. A randomized
hemodynamic comparison of intravenous amiodar-
one with and without Tween 80. Eur Heart J 1988;
9(2):142–8.
[36] Roden DM. A practical approach to torsade de
pointes. Clin Cardiol 1997;20(3):285–90.
[37] Credner SC, Klingenheben T, Mauss O, et al. Elec-
trical storm in patients with transvenous implantable
cardioverter-defibrillators: incidence, management
and prognostic implications. J Am Coll Cardiol
1998;32(7):1909–15.
[38] Tsagalou EP, Kanakakis J, Rokas S, et al. Suppres-
sion by propranolol and amiodarone of an electrical
storm refractory to metoprolol and amiodarone. Int
J Cardiol 2005;99(2):341–2.
[39] Nademanee K, Taylor R, Bailey WE, et al. Treating
electrical storm: sympathetic blockade versus ad-
vanced cardiac life support–guided therapy. Circula-
tion 2000;102:742–7.
[40] Henderson MC, Prabhu SD. Syncope: current diag-
nosis and treatment. Curr Probl Cardiol 1997;22(5):
242–96.
[41] Link MS, Lauer EP, Homoud MK, et al. Low yield
of rule-out myocardial infarction protocol in
133
patients presenting with syncope. Am J Cardiol
2001;88(6):706–7.
[42] Bayer AJ, Chandha JS, Farag RR, et al. Changing
presentation of myocardial infarction with increas-
ing old age. J Am Geriatr Soc 1986;34(4):263–6.
[43] Grubb BP. Neurocardiogenic syncope and related
disorders of orthostatic intolerance. Circulation
2005;111(22):2997–3006.
[44] Calkins H, Shyr Y, Frumin H, et al. The value of the
clinical history in the differentiation of syncope due
to ventricular tachycardia, atrioventricular block,
and neurocardiogenic syncope. Am J Med 1995;
98(4):365–73.
[45] Brignole M, Alboni P, Benditt D, et al. Guidelines
on management (diagnosis and treatment) of
syncopedupdate 2004. Europace 2004;6:467–537.
[46] Savage DD, Corwin L, McGee DL, et al. Epidemio-
logic features of isolated syncope: the Framingham
Study. Stroke 1985;16(4):626–9.
[47] Quinn JV, Stiell IG, McDermott DA, et al. Deriva-
tion of the San Francisco syncope rule to predict
patients with short-term serious outcomes. Ann
Emerg Med 2004;43(2):224–32.
[48] Kapoor WN, Karpf M, Wieand S, et al. A prospec-
tive evaluation and follow-up of patients with syn-
cope. N Engl J Med 1983;309(4):197–204.
[49] Kapoor W, Snustad D, Peterson J, et al. Syncope in
the elderly. Am J Med 1986;80(3):419–28.
[50] Colivicchi F, Ammirati F, Melina D, et al, for the
OESIL (Osservatorio Epidemiologico sull Sincope
nel Lazio) Study Investigators. Development and
prospective validation of a risk stratification system
for patients with syncope in the emergency depart-
ment: the OESIL risk score. Eur Heart J 2003;
24(9):811–9.
[51] Stevenson WG, Chaitman BR, Ellenbogen KA,
et al, for the Subcommittee on Electrocardiography
and Arrhythmias of the American Heart Association
Council on Clinical Cardiology. Heart Rhythm
Society. Clinical assessment and management of
patients with implanted cardioverter-defibrillators
presenting to nonelectrophysiologists. Clinical as-
sessment and management of patients with
implanted cardioverter-defibrillators presenting to
nonelectrophysiologists. Circulation 2004;110(25):
3866–9.
 Cardiol Clin 24 (2006) 135–146
Hypertensive Crisis: Hypertensive Emergencies
and Urgencies
Monica Aggarwal, MD, Ijaz A. Khan, MD*
Division of Cardiology, University of Maryland School of Medicine, 22 South Greene Street,
Baltimore, MD 21201, USA
Hypertension affects an estimated 50 million
people in the United States, and it contributed to
more than 250,000 deaths in the year 2000 because
of end-organ damage [1]. Normal blood pressure is
defined as a systolic blood pressure of less than
120 mm Hg and diastolic blood pressure of less
than 80 mm Hg. Hypertension is defined as a sys-
tolic blood pressure of 140 mm Hg or higher or
a diastolic blood pressure of 90 mm Hg or higher.
A systolic blood pressure of 120 to 139 mm Hg or
a diastolic blood pressure of 80 to 89 mm Hg is
considered prehypertension, because people in
this range of blood pressure have higher tendency
to develop hypertension over time. There is a con-
tinuous, graded relationship between hypertension
and cardiovascular risk; even a slightly elevated
blood pressure increases risk for cardiovascular
disease. The maximum blood pressure as well as
the duration of elevated pressure determines the
outcome [2,3]. Most patients who have chronically
uncontrolled hypertension suffer end-organ dam-
age over time. Patients with previously untreated
or inadequately treated high blood pressures are
most prone to acute rises in their blood pressures
[4,5]. Patients with secondary causes of hyperten-
sion are at higher risk of acute rises of blood pres-
sure than patients who have essential hypertension
[6,7]. The terms ‘‘malignant hypertension,’’ ‘‘hy-
pertensive emergency,’’ and ‘‘hypertensive urgency’’
were instituted to describe these acute rises in
blood pressure and resulting end-organ damage.
Hypertensive crisis includes hypertensive emer-
gencies and urgencies. Hypertensive emergency is
* Corresponding author.
E-mail address: ikhan@medicine.umaryland.edu
(I.A. Khan).
0733-8651/06/$ - see front matter Ó 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2005.09.002
defined as severe hypertension with acute end-
organ damage, such as aortic dissection, heart
failure, papilledema, or stroke. Although there is
no blood pressure threshold for the diagnosis of
hypertensive emergency, most end-organ damage
is noted with diastolic blood pressures exceeding
120 to 130 mm Hg. In these patients, immediate
but monitored reduction, often accomplished with
parenteral medications, is essential in preventing
long-term damage. Hypertensive urgency, on the
other hand, describes significantly elevated blood
pressure but without evidence of acute end-organ
damage. These patients also need reductions in
their blood pressures; but these reductions can be
achieved over a period of days, with oral medi-
cations and usually without an intensive monitor-
ing setting.
Historical perspective
Physicians have noticed the effects of hyper-
tension and hypertensive crises for decades. Vol-
hard and Fahr [8] were the first to notice the acute
changes in blood pressure and the differences in
pathophysiology of these changes from the chron-
ically elevated blood pressure. They noted that
patients who had severe hypertension had fundo-
scopic changes such as retinopathy and papille-
dema along with renal insufficiency and fibroid
necrosis of the renal arterioles. Also, they noticed
that patients who had acute elevations in blood
pressure were more prone to papilledema and to
acute changes in their kidneys. In 1914, they de-
fined the term ‘‘malignant hypertension’’ as an el-
evation in blood pressure with the sign of acute
end-organ damage. Subsequently, in 1921, Keith
and Wagener [9] described a similar finding of
cardiology.theclinics.com
136 AGGARWAL & KHAN
papilledema and severe retinopathy in patients
who had severe hypertension but who did not
have renal insufficiency. They then realized that
the end-organ eye and kidney damage were not mu-
tually exclusive to acute hypertensive episodes and
therefore broadened the definition of malignant
hypertension by stating that renal insufficiency
was not a necessary requirement for acute hyper-
tensive damage. Keith and Wagener [9] also used
the term ‘‘accelerated hypertension,’’ which they
defined as a syndrome with severe elevations in
blood pressure in the presence of retinal hemor-
rhages and exudates but without papilledema.
Later studies have shown that retinal hemorrhages
and exudates are important in malignant hyperten-
sion and are associated with decreased survival.
Notably, however, the presence or absence of
papilledema is not associated with decreased
survival [10,11].
In 1928, Oppenheimer and Fishberg [12] were
the first to use the term ‘‘hypertensive encephalop-
athy’’ when they noted malignant hypertension
associated with headaches, convulsions, and neu-
rologic deficits in a 19-year-old student. Currently,
the terms ‘‘malignant hypertension’’ and ‘‘acceler-
ated hypertension’’ are used infrequently and
have been replaced by terms such as ‘‘hypertensive
crisis,’’ ‘‘hypertensive emergency,’’ and ‘‘hyperten-
sive urgency’’ (Table 1).
Demographics
The prevalence of hypertension has increased,
partially because of the stringent definition of
hypertension. There are notable demographic
trends in the prevalence of hypertension. Hyper-
tension is more common in older age groups and is
more common in men than in women [13]. It is 1.5
to two times more prevalent in black Americans.
An analysis of data from the 1999 to 2000 National
Health and Nutrition Examination Survey has
shown that the combined prevalence of prehyper-
tension and hypertension has increased to 60% of
American adults (67% of men and 50% of women),
and 27% of American adults have established
hypertension. The combined prevalence of prehy-
pertension and hypertension is 40% in the 18- to
39-years age group and is 88% in the greater-
than-60-years age group [14]. This survey showed
certain risk predictors of hypertension. Education
level was a notable factor. The combined preva-
lence of prehypertension and hypertension in-
creased from 54% in the high-school–educated
group to 65% in the non–high-school–educated
Table 1
Definitions
Term Definition
Prehypertension Systolic blood pressure
120–139 mm Hg and diastolic
blood pressure 80–89 mm Hg
Hypertension Systolic blood pressure O
140/90 mm Hg
Hypertensive Hypertensive urgency or
crisis emergency
Hypertensive Acute rise in blood pressure
urgency without acute end-organ
damage; diastolic blood
pressure usually O 120 mm Hg
Hypertensive Acute rise in blood pressure with
emergency acute end-organ damage;
diastolic blood pressure usually
O 120 mm Hg
group. Obesity was also an important risk predic-
tor in this analysis; 75% of overweight individuals
were prehypertensive or hypertensive, but only
47% of the non-overweight group qualified as
such.
Patients who were noted to have prehyperten-
sion were also noted to have other risk factors for
stroke and cardiovascular disease, such as hyper-
cholesterolemia, obesity, and diabetes. These risk
factors were less prevalent in people who had
normal blood pressures. The percentage of people
who had more than one risk factor for cardiovas-
cular disease was higher in the prehypertensive
group than in patients who had normal blood
pressure [14]. This cross-sectional analysis also
evaluated patients who, when made aware of their
hypertension, followed dietary, lifestyle, and med-
ication changes. It was noted that 7% of patients
did not adopt any lifestyle changes, and 15% of
patients would not take any antihypertensive
medications. The problem was more notable in
younger patients and in Mexican-American pa-
tients. Of the patients taking anti-hypertensive
medications, 54% had their hypertension con-
trolled. Men and patients with higher education
were more likely to have their blood pressures
well controlled.
Of the estimated 50 million Americans with
hypertension, less than 1% will have a hyperten-
sive crisis [15]. In a study by Zampaglione and as-
sociates [16], hypertensive crises were found to
account for more than 25% of all patient visits
to a medical section of an emergency department.
One third of those patients were noted to have
hypertensive emergencies. In the years when
 HYPERTENSIVE CRISIS
treatment of hypertensive crises was difficult, be-
cause of inadequate monitoring and lack of paren-
teral medications, survival was only 20% at 1 year
and 1% at 5 years [17]. Before antihypertensive
agents became available, thoracolumbar sympa-
thectomy prolonged survival to 40% at 6.5 years.
With the advent of ganglionic blocking medica-
tions, the 5-year survival rates increased to 50%
to 60% in 1960 [18]. During the past 2 decades,
with the increased focus on blood pressure control
and emphasis on compliance, the 10-year survival
rates have approached 70% [19].
Cause
Ninety-five per cent of patients who have
hypertension have no obvious underlying cause.
As such, hypertension without secondary causes is
defined as essential hypertension. The remaining
5% of patients have an underlying cause for their
elevated blood pressures, of which certain groups
have higher chances of presenting with a hyper-
tensive crisis (Box 1). Use of recreational drugs,
such as cocaine, has become a frequent cause of
hypertensive crisis. Cocaine amphetamines, phen-
cyclidine hydrochloride, and diet pills are sympa-
thomimetic and thus may cause severe acute
hypertension. Patients taking monoamine oxidase
inhibitors along with tricyclic antidepressants,
antihistamines, or food with tyramine are prone
to hypertensive crises. Withdrawal syndromes
from drugs such as clonidine or beta-blockers
may also precipitate hypertensive crises [20]. Pheo-
chromocytoma is a rare cause of hypertensive cri-
ses. Patients with spinal cord disorders, such as
Guillain Barré syndrome, are also at a higher risk
for hypertensive crises. These patients are prone
to autonomic overactivity syndrome manifested
by severe hypertension, bradycardia, headache,
and diaphoresis. The syndrome is triggered by
stimulation of dermatomes or muscles innervated
by nerves below the spinal cord lesions.
Pathophysiology
Normal mechanisms to regulate blood pressure
Blood pressure regulation is a critical action
that allows perfusion to vital organs of the body.
This action is based on a balance between pe-
ripheral vascular resistance and cardiac output and
is dependent on the integrated actions of
the cardiovascular, renal, neural, and endocrine
systems. This interdependence allows a back-up
137
Box 1. Causes of secondary
hypertension
Medications
Oral contraceptive pills
Cocaine hydrochloride
Phencyclidine hydrochloride
Monoamine oxidase inhibitors
Sympathomimetic diet pills
Nonsteroidal anti-inflammatory drugs
Amphetamines
Cyclosporin
Steroids
Acute glomerulonephritis
Renal parenchymal disease
Renal artery stenosis
Hyperaldosteronism
Cushing disease
Pheochromocytoma
Pregnancy
Sleep apnea
Coarctation of aorta
Spinal cord injuries
system so that the body can cope with internal and
external stresses such as thirst, fear, infection, and
trauma. Multiple intrinsic systems are activated in
the body in response to external and internal
stressors [21]. The renin-angiotensin-aldosterone
system is thought to be critically responsible for
blood pressure changes. Renin is released from
the juxtaglomerular apparatus in response to low
sodium intake, underperfusion of the kidney, and
increased sympathetic activity. Renin is responsi-
ble for converting angiotensinogen to angiotensin,
which is not metabolically active. The angiotensin
is subsequently converted to angiotensin II in the
lungs by the angiotensin-converting enzyme. An-
giotensin II is a potent vasoconstrictor, which leads
to increases in blood pressure. Besides its intrinsic
vasoconstrictive effects, angiotensin II also causes
aldosterone release, which further increases blood
pressure by causing salt and water retention. Stud-
ies in rats support the role of the renin-angiotensin-
aldosterone system in blood pressure elevation.
When rats were given the Ren-2 gene, which
activates the renin-angiotensin-aldosterone sys-
tem, they developed severe hypertension [22]. Fur-
ther support comes from therapeutic methods,
such as using angiotensin-converting enzyme
inhibitors or angiotensin receptor blockers or
138 AGGARWAL & KHAN
surgical removal of an ischemic kidney, which can
prevent blood pressure elevations [23].
The renin-angiotensin-aldosterone system is
not considered solely responsible for changes in
blood pressure. Black Americans, for instance,
often have low renin, angiotensin II, and aldoste-
rone levels and yet have a notably higher incidence
of hypertension. Therefore, they are less responsive
to medications blocking the renin-angiotensin-
aldosterone system. Theoretically, patients who
have low renin states might have noncirculatory
local renin-angiotensin paracrine or epicrine sys-
tems. These systems have been found in the kidney,
arterial tree, and the heart; and are probably
responsible for local control of blood pressure [24].
The sympathetic nervous system also affects
blood pressure, especially in times of stress and
exercise. The sympathetic nervous system can
cause arterial vasoconstriction and can raise car-
diac output. It is thought that initially the sympa-
thetic nervous system increases the cardiac output
without affecting peripheral vascular resistance.
The raised cardiac output increases flow to the
vascular bed, and as the cardiac output increases,
the autoregulatory response of the vascular bed is
activated. This autoregulatory response results in
constriction of the arterioles to prevent the pres-
sure from reaching the capillaries and affecting cell
hemostasis [21].
In addition, endothelial function plays a central
role in blood pressure maintenance. The endothe-
lium secretes nitric oxide, prostacylin, and endo-
thelin, which modulate vascular tone. Nitric oxide
is released by endothelial agonists such as acetyl-
choline and norepinephrine and in response to
shear stress [21]. Endothelin-1 has great vasocon-
strictive activities and may cause a salt-sensitive
rise in blood pressure and a rise in blood pressure
by triggering the renin-angiotensin-aldosterone
system [24]. Other vasoactive substances involved
in blood pressure maintenance include bradykinin
and natriuretic peptides. Bradykinin is a potent
vasodilator that is inactivated by angiotensin-
converting enzyme. Natriuretic peptides are se-
creted from the heart in response to increase in
blood volume and cause an increase in sodium
and water excretion.
Altered mechanisms in hypertension and
hypertensive crises
The pathophysiology of hypertensive crisis is
not well understood. It is thought that an abrupt
rise in blood pressure, possibly secondary to
a known or unknown stimulus, may trigger the
event. During this abrupt initial rise in blood
pressure, the endothelium tries to compensate for
the change in vasoreactivity by releasing nitric
oxide. When the larger arteries and arterioles
sense elevated blood pressures, they respond
with vasoconstriction and subsequently with hy-
pertrophy to limit pressure reaching the cellular
level and affecting cellular activity. Prolonged
smooth muscle contraction leads to endothelial
dysfunction, loss of nitric oxide production, and
irreversible rise in peripheral arterial resistance.
Without the continuous release of nitric oxide, the
hypertensive response becomes more severe, pro-
moting further endothelial damage, and a vicious
cycle continues. The endothelial dysfunction is
further triggered by inflammation induced by
mechanical stretch. The expression of inflamma-
tory markers such as cytokines, endothelial adhe-
sion molecules, and endothelin-1 is increased
[25,26]. These molecular events probably increase
the endothelial permeability, inhibit fibrinolysis,
and, as a result, activate coagulation. Coagulation
along with platelet adhesion and aggregation re-
sults in deposition of fibrinoid material, increased
inflammation, and the vasoconstriction of the ar-
teries, resulting in further endothelial dysfunction.
The role of the renin-angiotensin-aldosterone sys-
tem also seems to be important in hypertensive
emergency. There seems to be an amplification
of this system that contributes to vascular injury
and tissue ischemia [27].
The blood pressure at which the acute end-
organ damage starts occurring is different in each
individual. Patients who are more chronically
hypertensive have had more smooth muscle con-
traction and subsequent arterial hypertrophy,
which lessens the effect of acute rise in blood
pressure on the capillary circulation. Although
malignant hypertension is defined as a diastolic
blood pressure greater than 130 mm Hg, normo-
tensive patients who have not had time to establish
compensatory autoregulatory mechanisms are
more sensitive to elevations in blood pressure and
may suffer end-organ damage when diastolic blood
pressure becomes greater than 100 mm Hg.
Clinical manifestations
Hypertensive crisis shares all of the pathologic
mechanisms and end-organ complications of the
milder forms of hypertension [27]. In one study of
the prevalence of end-organ complications in
 HYPERTENSIVE CRISIS
hypertensive crisis, central nervous system abnor-
malities were the most frequent. Cerebral infarc-
tions were noted in 24%, encephalopathy in
16%, and intracerebral or subarachnoid hemor-
rhage in 4% of patients. Central nervous system
abnormalities were followed in incidence by car-
diovascular complications such as acute heart fail-
ure or pulmonary edema, which were seen in 36%
of patients, and acute myocardial infarction or
unstable angina in 12% of patients. Aortic dissec-
tion was noted in 2%, and eclampsia was noted in
4.5% of patients [16]. The end-organ damage is
outlined in Box 2.
Acute neurologic syndromes
The cerebral vasculature must maintain a con-
stant cerebral perfusion despite changes in blood
pressure. Cerebral autoregulation is the inherent
ability of the cerebral vasculature to maintain this
constant cerebral blood flow [28,29]. Normoten-
sive people maintain a constant cerebral blood
flow between mean arterial pressures of 60 mm
Hg and 120 mm Hg. As the mean arterial pressure
increases, there is disruption of the cerebral endo-
thelium and interruption of the blood–brain bar-
rier. Fibrinoid material deposits in the cerebral
vasculature and causes narrowing of the vascular
lumen. The cerebral vasculature, in turn, attempts
to vasodilate around the narrowed lumen, which
leads to cerebral edema and microhemorrhages
[30]. The changes in cerebral vasculature and cere-
bral perfusion seem to affect primarily the white
matter in the parieto-occipital areas of the brain
[31]. The predilection toward the parieto-occipital
Box 2. End-organ damage in
hypertension
Acute neurologic syndromes
Hypertensive encephalopathy
Cerebral infarction
Subarachnoid hemorrhage
Intracranial hemorrhage
Myocardial ischemia and infarction
Acute left ventricular dysfunction
Acute pulmonary edema
Aortic dissection
Retinopathy
Renal insufficiency
Eclampsia
139
regions possibly results from decreased sympa-
thetic innervation of the vessels in this region
[32]. There are also reports of brainstem involve-
ment, however [33].
Normotensive patients may develop endothe-
lial dysfunction at lower mean arterial pressures,
whereas chronically hypertensive patients can
tolerate higher mean arterial pressures before
they develop such a dysfunction. Chronically
hypertensive patients have the capacity to autor-
egulate and have cerebral blood flow and oxygen
consumption similar to those in normotensive
persons [34]. Changes in the structure of the arte-
rial wall cause increased stiffness and higher cere-
brovascular resistance, however [35]. Although
a higher threshold must be reached before they
have disruption of their autoregulation system,
hypertensive patients, because of the increased ce-
rebrovascular resistance, are more prone to cere-
bral ischemia when flow decreases [30].
Hypertensive encephalopathy is one of the
clinical manifestations of cerebral edema and
microhemorrhages seen with dysfunction of cere-
bral autoregulation. It is defined as an acute
organic brain syndrome or delirium in the setting
of severe hypertension. Symptoms include severe
headache, nausea, vomiting, visual disturbances,
confusion, and focal or generalized weakness. Signs
include disorientation, focal neurologic defects,
focal or generalized seizures, and nystagmus. If
not adequately treated, hypertensive encephalopa-
thy can lead to cerebral hemorrhage, coma, and
death, but with proper treatment it is completely
reversible [36]. The diagnosis of hypertensive en-
cephalopathy is a clinical diagnosis. Stroke, sub-
arachnoid hemorrhage, mass lesions, seizure
disorder, and vasculitides need to be ruled out.
Cerebral infarction, caused by an imbalance
between supply and demand, is another neurologic
sequela of severe acute rises in blood pressure [37].
Intracranial and subarachnoid hemorrhages are
other possible neurologic complications of hyper-
tensive crisis. The risk is increased in patients
who have intracranial aneurysms and in those
taking anticoagulant medications.
Myocardial ischemia
Hypertension affects the structure and function
of the coronary vasculature and left ventricle.
Activation of the renin-angiotensin-aldosterone
system in hypertension constricts systemic vascu-
lature and, thereby, increases myocardial oxygen
demand by increasing left ventricular wall tension.
140 AGGARWAL & KHAN
Increasing wall tension leads to hypertrophy of
the left ventricular myocytes and to deposition of
protein and collagen in the extracellular matrix
of the ventricular wall. These actions increase
ventricular mass, which further increases oxygen
demand on the heart. A second effect of the
hypertrophy is that the newly thickened ventricle
can cause coronary compression and decreased
luminal blood flow. Thirdly, hypertension can
increase the epicardial coronary wall thickness,
which increases the wall-to-lumen ratio and de-
creases coronary blood flow reserve. Concomitant
atherosclerosis worsens the wall-to-lumen ratio,
further decreases coronary flow reserve, and leads
to coronary ischemia [38]. Acute rise in blood
pressure also results in endothelial injury at the
level of the coronary capillaries.
Left ventricular failure
Another effect of hypertensive crisis on the
heart is left ventricular failure and acute pulmo-
nary edema. In certain cases, despite increasing
wall tension, the left ventricle cannot hypertrophy
enough to overcome the acute rise in systemic
vascular resistance. This inability to compensate
leads to left ventricular failure and a backup of
flow causing pulmonary edema. Secondly, neuro-
hormonal activation of the renin-angiotensin-
aldosterone system leads to increased sodium
content and increased total body water. In addi-
tion, left ventricular hypertrophy leads to focal
ischemia and subsequent inadequate diastolic
filling, which can result in imbalance between
left ventricular contraction and relaxation, leading
to pulmonary edema [5]. Clinically, patients show
signs of volume overload or signs of reduced tis-
sue perfusion such as cool limbs.
Aortic dissection
Aortic dissection is the most rapidly fatal
complication of hypertensive crises. Risk factors
for dissection include untreated hypertension,
advanced age, and diseases of the aortic wall.
Dilation of the aorta caused by atherosclerosis
and high blood pressures tear the intima of the
vessel, allowing a surge of blood into the aortic
wall. The blood driven by pulsatile pressure
separates the arterial wall into two layers [39].
Clinically, patients complain of retrosternal or in-
terscapular chest pain that migrates to the back. If
dissection extends proximally, it can lead to aortic
insufficiency or a pericardial effusion. Dissection
can lead to compression or occlusion of a branch
of the aorta and subsequently lead to organ ische-
mia. Clinical signs that are notable with dissection
include discrepancies between pulses, murmur of
aortic insufficiency, and neurologic deficits [40].
Diagnosis of aortic dissection can be confirmed
with transesophageal echocardiography, CT, or
MRI [41].
Hypertensive retinopathy
Retinopathy was one of the first signs of
hypertension, noted in 1914. In the early years
fundoscopy was considered a definitive tool in
diagnosing hypertensive encephalopathy. Papille-
dema was noted in patients who had hypertensive
encephalopathy but was not necessary for di-
agnosis [42]. Retinal hemorrhages and exudates
were considered indicative of malignant hyperten-
sion. Since 1914, multiple studies have looked at
retinopathy in the setting of hypertension. In
mild to moderate hypertension, the degree of focal
narrowing of arterioles has been associated with
the level of blood pressure rise. No relationship
has been found between retinal changes and the
end-organ damages such as ventricular hypertro-
phy or microalbuminuria, however [43]. Ophthal-
moscopy may be useful in recognizing acute
hypertensive target-organ damage such as hyper-
tensive encephalopathy, but the absence of retinal
exudates, hemorrhages, or papilledema does not
exclude the diagnosis [11].
Acute renal insufficiency
Acute renal insufficiency may be a cause or
result of rapidly progressive hypertension. Impor-
tant causes of hypertension are parenchymal
disease, such as acute glomerulonephritis or renal
artery stenosis, or cyclosporine use in kidney
transplant patients. Renal insufficiency could also
be a result of hypertension and hypertensive crisis,
however. Normal renal autoregulation enables the
kidney to maintain a constant renal blood flow and
glomerular filtration rate for mean arterial pres-
sures between 80 and 160 mm Hg. Under normal
conditions, autoregulatory vasodilation is maxi-
mal at a mean arterial pressure of about 80 mm Hg.
In chronic hypertension, the small arteries of the
kidney, including the afferent arteriole, undergo
pathologic changes that alter renal autoregulation,
showing signs of endothelial dysfunction and
impaired vasodilation. Structural changes initially
are probably protective of the kidney, but over
time progressive narrowing of the preglomerular
vessels results in ischemic injury, tubular atrophy,
 HYPERTENSIVE CRISIS
and fibrosis. With the impairment of the renal
autoregulatory system, the intraglomerular pres-
sure begins to vary directly with systemic arterial
pressure [44]. As such, the afferent vasculature be-
comes a passive conduit and cannot prevent the
kidney from being affected by fluctuations in pres-
sure and flow, leading to acute renal ischemia in
cases of hypertensive crisis.
Pregnancy-induced hypertension
Pre-eclampsia is characterized as a syndrome of
pregnancy-induced hypertension, edema, and pro-
teinuria in a pregnant woman after the twentieth
week of gestation [27]. Eclampsia is the end result
of this spectrum and is associated with acute hyper-
tension, edema, proteinuria, and concomitant seiz-
ures. Although the pathophysiologic mechanisms
of pre-eclampsia and eclampsia are not well under-
stood, blood pressure elevation is characterized by
an increased responsiveness to vasoconstrictors,
especially angiotensin II. There is also decreased
sensitivity to endothelium-derived vasodilators
[27]. Pregnancy-induced hypertension usually re-
solves spontaneously after delivery of baby.
Postoperative hypertension
The postoperative hypertensive crisis is classi-
cally an acute rise in blood pressure within the
first 2 hours after surgery and is typically short in
duration, with most patients requiring treatment
for 6 hours or less [45]. Although postoperative
hypertensive crises can occur with any surgery,
they are more common with cardiothoracic, vas-
cular, head, neck, and neurosurgical procedures
[45]. In one study of patients undergoing radical
neck dissection, the frequency of hypertensive
crisis ranged from 9% to 25% [46]. A feared com-
plication of postoperative hypertensive crisis is
bleeding from operation site [47]. The pathophys-
iology of postoperative hypertensive crisis is prob-
ably related to stimulation of the sympathetic
nervous system and catecholamine surge [48].
Clinical evaluation
History and detailed physical examination are
important in all patients who present with severe
hypertension. A thorough history is important to
determine the time since diagnosis of hypertension,
the severity, and the baseline blood pressures at
home. Determining the presence of end-organ
damage and other comorbidities is important,
because both are crucial factors influencing the
141
choice of antihypertensive drugs. Knowledge of the
patient’s medications and compliance with these
medications, including over-the-counter medica-
tions and recreational drugs, is essential, because
both could contribute to an acute rise in blood
pressure. Blood pressure should be checked in both
arms and should be done in supine and standing
positions, if possible, to determine volume status.
Neurologic examination is important to determine
the focal signs of an ischemic or hemorrhagic
stroke. The presence of delirium, nausea, vomiting,
and seizures suggests hypertensive encephalopa-
thy. Fundoscopic examination could be of help,
because the presence of exudates, hemorrhage, or
papilledema supports the diagnosis of hypertensive
encephalopathy. Cardiovascular examination in-
cludes listening for new murmurs of aortic in-
sufficiency associated with dissection or of ischemic
mitral regurgitation. A gallop or left ventricular
heave could suggest heart failure. Crackles in the
lung fields suggest pulmonary edema. Laboratory
studies should include serum electrolytes, blood
urea nitrogen, serum creatinine level, blood cell
count, and peripheral smear. An ECG should be
taken for myocardial ischemia and left ventricular
hypertrophy, and a chest radiograph should be
obtained for cardiac enlargement and widened
mediastinum. Urine analysis is indicated for as-
sessment of proteinuria and tubular casts. Plasma
renin and aldosterone levels could be obtained if
patient is not taking diuretics or other medications
that could have affected these levels [22].
Treatment
Hypertensive urgency can be treated in a non-
ICU setting with oral medications over 24 to 48
hours. Medications such as beta-blockers, diu-
retics, angiotensin-converting enzyme inhibitors,
and calcium-channel blockers can be titrated
initially as an inpatient; then the patient can be
discharged with close follow-up. If there is acute
end-organ damage, however, the patient should
be admitted to the ICU and treated with in-
travenous medications. The goal of therapy is
prompt but gradual reduction in blood pressure.
The most reasonable goal is to lower the mean
arterial pressure by about 25% or to reduce the
diastolic blood pressure to 100 to 110 mm Hg.
Medication options for treatment
There have been no large clinical trials to
determine the optimal pharmacologic therapy in
hypertensive emergency patients, primarily
142 AGGARWAL & KHAN

because of the heterogeneity among the patients
and their end-organ damage. As such, manage-
ment of hypertensive crisis should be dictated by
individual presentation and should be specific to
the end organ at risk (Tables 2, 3).
Sodium nitroprusside is the drug of choice for
most hypertensive emergencies because it has an
immediate onset of action and can be titrated
quickly and accurately. The duration of effect is
1 to 2 minutes. The mechanism of action of this
agent is probably similar to that of endogenous
nitric oxide. Sodium nitroprusside is an endoge-
nous arteriolar and venous dilator and has no
effects on the autonomic or central nervous system
[49]. The venous dilation decreases preload to the
heart and subsequently decreases cardiac output,
whereas the arterial dilation inhibits the reflex rise
in blood pressure from the drop in cardiac output.
Because sodium nitroprusside is a direct vasodila-
tor, one might think that it increases cerebral blood
flow and intracranial pressure. The fall in systemic
pressure, however, seems to inhibit the rise in cere-
bral blood flow, and patients who have neurologic
damage respond well to this agent [18]. Sodium ni-
troprusside must be administered as an infusion
and thus requires continuous surveillance with in-
tra-arterial monitoring. An end product of nitro-
prusside is thiocyanate, a precursor to cyanide
that can causes nausea, vomiting, lactic acidosis,
and altered mental status. Cyanide toxicity can be
rapidly fatal. Sodium nitroprusside is broken
down by the liver and cleared through the kidney,
and thus thiocyanate levels must be followed in
patients who have hepatic or renal insufficiency
to ensure prevention of a toxic buildup [21].
Labetalol is another first-line agent for hyper-
tensive emergency. It is a combined alpha- and
beta-blocking agent; the beta-blocking activity of
labetalol is five- to tenfold that of the alpha
component [50]. The beta effects of labetalol are
only about one fifth the activity of propranolol,
however [18]. Its onset of action is within 5 to
10 minutes, and the duration of action is about
3 to 6 hours. Labetalol can be used safely in
most patients, but caution should be exercised in
patients who have severe bradycardia, congestive
heart failure, or bronchospasm.
Fenoldopam is the first selective dopamine-1
receptor agonist approved for in-hospital shorter-
term management of severe hypertension up to
the first 48 hours of treatment [51–54]. The mech-
anism of action involves activating dopamine at
the level of the kidney. Dopamine is a well-known
vasoconstrictor and is sympathomimetic at inter-
mediate to high doses. At low doses, however, do-
pamine lowers diastolic blood pressure and,
importantly, increases renal perfusion and pro-
motes diuresis. Fenoldopam is administered by
parenteral continuous infusion and has 50% of
its maximal effect within 15 minutes. Its duration
of action is about 10 to 15 minutes; thus, it can be
discontinued swiftly if the decrease in blood pres-
sure is too rapid. The efficacy of fenoldopam in re-
nal perfusion is equal to or possibly better than
that of sodium nitroprusside. There are no toxic
metabolites of fenoldopam; however, the onset
of action is slower, and the duration of effect is
longer than that of sodium nitroprusside. Patients
develop tachyphylaxis to fenoldopam after 48
hours, and headache can be a side effect.

Table 2
Medications used in hypertensive emergencies
Onset of Duration Cardiac Renal
Name Dosing Action of Action Preload Afterload Output perfusion
Sodium IV 0.25–10 mg/kg/min within 1–2 min decreased decreased no effect decreased
nitroprusside seconds
Labetolol IV (20- to 80-mg 5–10 min 2–6 hr no effect decreased decreased no effect
bolus/10 min)
Fenoldopam IV 0.1–0.6 mg/kg/min 10–15 min 10–15 min no effect decreased increased increased
Nicardipine IV 2–10 mg/hr 5–10 min 2–4 hr no effect decreased increased no effect
Esmolol IV 80-mg bolus over 6–10 min 20 min no effect no effect decreased no effect
30 seconds, followed by
150 mg/kg/min infusion
Methyldopa IV (250-to 1000-mg bolus 3–6 hr up to 24 hr no effect decreased decreased no effect
every 6 hr)
Hydralazine IV bolus (10–20 mg) 10 min 2–6 hr no effect decreased increased no effect
Abbreviation: IV, intravenous.
 HYPERTENSIVE CRISIS 143

Table 3 the goal of therapy in hypertensive encephalopa-

Major side effects of medications thy is to reduce the mean arterial pressure gradu-
Name Comments Major Side Effects ally by no more than 25% or to a diastolic blood
Sodium Need to measure Cyanide toxicity:
pressure of 100 mm Hg, whichever is higher, dur-
nitroprusside thiocyanate nausea, vomiting, ing the first hour. If neurologic function worsens,
levels, caution altered mental the therapy should be suspended, and blood
in renal status, lactic pressure should be allowed to increase [20]. In in-
insufficiency acidosis,death tracerebral or subarachnoid hemorrhage, blood
Labetolol Alpha and Bradycardia, pressure reduction is necessary to stop the bleed-
beta blocker, bronchospasm, ing and can be facilitated by decreasing pressures
contraindicated nausea by 25%. It is important to reduce blood pressure
in acute heart slowly to prevent cerebral hypoperfusion to the al-
failure
ready ischemic areas [36]. Often after a stroke
Nicardipine Safe in coronary Reflex tachycardia,
bypass patients flushing
there is a loss of cerebral autoregulation in the
Esmolol Short-acting Bradycardia, infarct/ischemic region. This loss of cerebral au-
beta blocker, bronchospasm toregulation makes blood pressure reduction cau-
contraindicated tionary, because the ischemic region is more prone
in acute heart to hypoperfusion during blood pressure reduc-
failure tion. As such, blood pressure reduction is not rec-
Methyldopa Safe in pregnancy, Drowsiness, fever, ommended after stroke, except in cases of extreme
needs renal jaundice blood pressure elevation (diastolic blood pressure
dosing O 130 mm Hg). If neurologic function deterio-
Hydralazine Safe in pregnancy Reflex tachycardia,
rates with reduction of blood pressure, therapy
lupus-like
syndrome
should be suspended, and blood pressure should
be allowed to rise. Blood pressure usually declines
spontaneously to prestroke levels within 4 days of
Nicardipine is a calcium-channel blocker ad- an acute ischemic stroke without any antihyper-
ministered parenterally by continuous infusion for tensive treatments [37]. Sodium nitroprusside is
hypertensive crises. The onset of action of this the drug of choice for treatment of acute neuro-
drug is 5 to 10 minutes, and the duration of action logic syndromes in hypertensive crisis. Labetolol
is 1 to 4 hours. Adverse reactions are reflex is a good alternative unless there is evidence of
tachycardia and headache [55]. Nicardipine is severe bradycardia associated with the cerebral
contraindicated in patients who have heart failure. edema. Clonidine and methyldopa should not
Esmolol is a cardioselective beta-blocker with be used, because they can cause central nervous
a short duration of action. It reduces the systolic system depression and complicate the clinical
blood pressure and mean arterial pressure as well picture.
as heart rate, cardiac output, and stroke volume. Severe acute hypertension often results in
There is a notable decrease in myocardial oxygen myocardial ischemia even with patent coronary
consumption. Peak effects are generally seen arteries. In this situation, intravenous nitroglyc-
within 6 to 10 minutes after a bolus dose. The erin is effective in reducing systemic vascular
effects resolve 20 minutes after discontinuation of resistance and improving coronary perfusion.
infusion. The elimination half-life of esmolol is Nitrates should be given until symptoms subside
about 8 minutes. or until diastolic blood pressure is 100 mm Hg.
Beta-blockers and calcium-channel blockers are
also potential options; both can decrease blood
Choice of agent
pressure while improving myocardial oxygena-
The choice of pharmacologic agent to treat tion. Calcium-channel blockers should be used
hypertensive crisis should be tailored to each with caution in patients who have possible heart
individual based on risks, comorbidities, and the failure. Acute pulmonary edema that is precipi-
end-organ damage. The lower limit of cerebral tated by hypertension is best treated with sodium
blood flow autoregulation is reached when blood nitroprusside. The concomitant venous and arte-
pressure is reduced by 25%, and the cerebral rial dilation improve forward flow and cardiac
ischemia can be precipitated with rapid reductions output. This agent should be used in conjunction
of blood pressure of greater than 50% [35]. Hence, with morphine, oxygen, and a loop diuretic [49].
144 AGGARWAL & KHAN
Treatment of an aortic dissection depends on
location of the injury [41]. Type A or proximal
aortic dissections need immediate institution of
antihypertensive medications and immediate sur-
gery, but type B or distal aorta dissections can
be controlled medically. The medical therapy of
aortic dissection is aimed at reducing the shear
stress on aortic wall. This reduction is achieved
by lowering diastolic blood pressure to less than
100 to 110 mm Hg and by decreasing heart rate.
This reduction is best achieved with a combination
of an intravenous beta-blocker and sodium nitro-
prusside to decrease both the blood pressure and
the heart rate. Another option is the use of labeto-
lol, which has both alpha- and beta-blocking
effects.
Renal insufficiency is a cause and a conse-
quence of severe hypertension. These patients
need reduction of systemic vascular resistance
without compromising renal blood flow or glo-
merular filtration rate. Fenoldopam is a good
choice in these patients because of the improve-
ment in renal perfusion, diuresis, and lack of
production of toxic metabolites. Tolerance does
develop to fenoldopam after 48 hours [53]. Sodi-
um nitroprusside can also be used, but there is
a risk of thiocyanate toxicity, and the thiocyanate
level needs to be closely monitored. Calcium-
channel blockers are effective and well tolerated
in renal transplant patients. Beta-blockers are
also useful agents in hypertension in kidney dis-
ease. Calcium-channel blockers and beta-blockers
have no clinically important effects on glomerular
filtration or renal hemodynamics [44]. Patients
who have renal insufficiency should not be treated
with angiotensin-converting enzyme inhibitors or
angiotensin receptor blockers.
In pre-eclampsia and eclampsia, blood pressure
control is essential. Many of the traditional anti-
hypertensive medications are contraindicated in
pregnancy because of their detrimental effects on
the fetus. Angiotensin-converting enzyme inhibi-
tors and angiotensin receptor blockers can ad-
versely affect fetal development and increase fetal
morbidity. Calcium-channel blockers reduce blood
pressure but decrease uterine blood flow and can
inhibit labor. Methyldopa is the mainstay of
treatment of blood pressure in pregnant patients.
It works centrally in reducing blood pressure and
heart rate. Methyldopa can be administered orally
or intravenously. With renal dysfunction the doses
need to be adjusted. The side effects are drowsi-
ness, fever, and jaundice. The medication does
cross into the placenta but is considered class B in
pregnancy [2]. Hydralazine is another agent that is
safe in pregnancy and can be administered safely
parenterally. Hydralazine may cause reflex tachy-
cardia and fluid retention because it activates the
renin-angiotensin-aldosterone system [27]. Data
show that labetolol is probably effective in reduc-
ing blood pressure in treatment of eclampsia with-
out inducing fetal distress [56].
Pheochromocytoma is a rare cause of parox-
ysmal or sustained blood pressure and can induce
hypertensive crisis. The treatment of choice in
these patients is labetolol or phentolamine (an
alpha-blocking agent). It is important not to use
beta-blockers alone, because then there is an
unopposed alpha activity that will worsen the
vasoconstriction, resulting in a further increase in
blood pressure. A reflex hypertensive crisis can
develop in patients who have abruptly stopped
taking antihypertensives, particularly clonidine or
beta-blockers. The treatment in these cases is to
restart previous medications after the initial re-
duction of blood pressure with labetolol or
sodium nitroprusside. Postoperative hypertension
is typically related to catecholamine surge from
activation of the sympathetic nervous system.
Therefore, the treatment of choice for postopera-
tive hypertensive crisis is with a beta-blocker or
labetolol.
Summary
Hypertensive crisis is a serious condition that is
associated with end-organ damage or may result
in end-organ damage if left untreated. Causes of
acute rises in blood pressure include medications,
noncompliance, and poorly controlled chronic
hypertension. Treatment of a hypertensive crisis
should be tailored to each individual based on the
extent of end-organ injury and comorbid condi-
tions. Prompt and rapid reduction of blood
pressure under continuous surveillance is essential
in patients who have acute end-organ damage.
References
[1] Heart disease and stroke statisticsd2004 update.
Dallas (TX): American Heart Association; 2003.
[2] Chobanian AV, Bakris GL, Black HR, et al. Sev-
enth report of the Joint National Committee on
Prevention, Detection, Evaluation, and Treatment
of High Blood Pressure. Hypertension 2003;42:
1206–52.
[3] Stamler J. Blood pressure and high blood pressure:
aspects of risk. Hypertension 1991;18:I95–I107.
 HYPERTENSIVE CRISIS
[4] Bennett NM, Shea S. Hypertensive emergency: case
criteria, sociodemographic profile, and previous care
of 100 cases. Am J Public Health 1988;78:636–40.
[5] Varon J, Marik PE. The diagnosis and management
of hypertensive crises. Chest 2000;118:214–27.
[6] Kincaid-Smith P. Malignant hypertension. J Hyper-
tens 1991;9:893–9.
[7] Calhoun DA, Oparil S. Hypertensive crises since
FDR: a partial victory. N Engl J Med 1995;332:
1029–30.
[8] Volhard F, Fahr TH. Die Brightsche Neirenkrank-
heit: KlinikPathlogie und Atlas, vol. 2. Berlin:
Springer Verlag; 1914. p. 247–65.
[9] Keith NM, Wagener HP, Keronohan JW. The syn-
drome of malignancy hypertension. Arch Intern
Med 1928;4:264–78.
[10] Ahmed ME, Walker JM, Beevers DG, et al. Lack of
difference between malignant and accelerated hyper-
tension. BMJ 1986;292:235–7.
[11] Bakker RC, Verburgh CA, van Buchem MA, et al.
Hypertension, cerebral edema and fundoscopy.
Nephrol Dial Transplant 2003;18:2424–7.
[12] Oppenheimer B, Fishberg AM. Hypertensive en-
cephalopathy. Arch Intern Med 1928;41:264–78.
[13] He J, Whelton PK. Epidemiology and prevention of
hypertension. Med Clin North Am 1997;81:1077–97.
[14] Wang Y, Wang QJ. The prevalence of prehyperten-
sion and hypertension among US adults according
to the new Joint National Committee guidelines.
Arch Intern Med 2004;164:2126–34.
[15] Gudbrandsson T. Malignant hypertension: a clinical
follow-up study with special reference to renal and
cardiovascular function and immunogenic factors.
Acta Med Scand Suppl 1981;650:1–62.
[16] Zampaglione B, Pascale C, Marchisio M, et al. Hy-
pertensive urgencies and emergencies: prevalence
and clinical presentation. Hypertension 1996;27:
144–7.
[17] Keith NM, Wagener HP, Barker NW. Some differ-
ent types of essential hypertension: their course
and prognosis. Am J Sci Med 1939;197:332–43.
[18] Kaplan N. Management of hypertensive emergen-
cies. Lancet 1994;344:1335–8.
[19] Webster J, Petrie JC, Jeffers TA, et al. Accelerated
hypertension patterns of mortality and clinical fac-
tors affecting outcome in treated patients. Q J Med
1993;96:485–93.
[20] Calhoun DA, Oparil S. Treatment of hypertensive
crisis. N Engl J Med 1990;323:1177–83.
[21] Vaughn CJ, Delanty N. Hypertensive emergencies.
Lancet 2000;356:411–7.
[22] Mullins JJ, Peters J, Ganten D. Fulminant hyperten-
sion in transgenic rats harbouring the mouse Ren-2
gene. Nature 1990;344:541–4.
[23] Laragh JH. Vasoconstriction-volume analysis for
understanding and treating hypertension: the use
of renin and aldosterone profiles. Am J Med 1973;
55:261–74.
145
[24] Beevers G, Lip GY, O’Brien E. The pathophysio-
logy of hypertension. BMJ 2001;322:912–6.
[25] Okada M, Matsumori A, Ono K, et al. Cyclic stretch
upregulates production of interleukin-8 and mono-
cyte chemotactic and activating factor/monocyte
chemoattractant protein-1 in human endothelial
cells. Arterioscler Thromb Vasc Biol 1998;18:
894–901.
[26] Verhaar MC, Beutler JJ, Gaillard CA, et al. Progres-
sive vascular damage in hypertension is associated
with increased levels of circulating P-selectin.
J Hypertens 1998;16:45–50.
[27] Blumenfeld JD, Laragh JH. Management of hyper-
tensive crises: the scientific basis for treatment deci-
sions. Am Heart J 2001;14:1154–67.
[28] Lassen NA. Cerebral blood flow and oxygen con-
sumption in man. Physiol Rev 1959;39:183–238.
[29] Van Lieshout JJ, Wieling W, Karemaker JM, et al.
Syncope, cerebral perfusion, and oxygenation.
J Appl Physiol 2003;94:833–48.
[30] Traon AP, Costes-Salon MC, Galinier M, et al.
Dynamics of cerebral blood flow autoregulation in
hypertensive patients. J Neurol Sci 2002;195:139–44.
[31] Garg RK. Posterior leukoencephalopathy. Postgrad
Med 2001;77:24–8.
[32] Beausang-Linder M, Bill A. Cerebral circulation in
acute arterial hypertension: protective effects of sym-
pathetic nervous activity. Acta Physiol Scand 1981;
111:193–9.
[33] Grond M, Reul J. Brainstem edema during a hyper-
tensive crisis with vasogenic and cytotoxic concerns.
Deutsch Med Wochenschr 2003;128:2487–9.
[34] Immink RV, van den Born BJ, van Montfrans GA,
et al. Impaired cerebral autoregulation in patients
with malignant hypertension. Circulation 2004;110:
2241–5.
[35] Strandgaard S, Paulson OB. Cerebral autoregula-
tion. Stroke 1984;15:413–6.
[36] Lavin P. Management of hypertension in patients
with acute stroke. Arch Intern Med 1986;146:66–8.
[37] Wallace J, Levy LL. Blood pressure after stroke.
JAMA 1981;246:2177–80.
[38] Frohlich ED. Target organ involvement in hyperten-
sion: a realistic promise of prevention and reversal.
Med Clin North Am 2004;88:1–9.
[39] Robicsek F, Thubrikar MJ. Hemodynamic consider-
ations regarding mechanism and prevention of aortic
dissection. Ann Thorac Surg 1994;58:1247–53.
[40] Khan IA. Clinical manifestations of aortic dissec-
tion. J Clin Basic Cardiol 2001;4:265–7.
[41] Khan IA, Nair CK. Clinical, diagnostic and man-
agement perspectives of aortic dissection. Chest
2002;122:311–28.
[42] McGregor E, Isles CG, Jay JL, et al. Retinal changes
in malignant hypertension. BMJ 1986;292:233–4.
[43] Dimmitt SB, West JN, Eames SM, et al. Usefulness
of ophthalmoscopy in mild to moderate hyperten-
sion. Lancet 1989;20:1103–6.
146 AGGARWAL & KHAN
[44] Palmer BF. Renal dysfunction complicating the
treatment of hypertension. N Engl J Med 2002;
347:1256–61.
[45] Haas CE, LeBlanc JM. Acute postoperative hyper-
tension: a review of therapeutic options. Am J
Health Syst Pharm 2004;61:1661–80.
[46] McGuirt WF, May JS. Postoperative hypertension
associated with radical neck dissection. Arch Otolar-
yngol Head Neck Surg 1987;113:1098–100.
[47] Gal TJ, Cooperman LH. Hypertension in the imme-
diate postoperative period. Br J Anaesth 1975;47:
70–4.
[48] Roberts AJ, Niarchos AP, Subramanian VA, et al.
Systemic hypertension associated with coronary
artery bypass surgery. J Thorac Cardiovasc Surg
1977;74:856–9.
[49] Shepherd AM, Irvine NA. Differential hemody-
namic and sympathoadrenal affects of sodium nitro-
prusside and hydralazine in hypertensive subjects.
J Cardiovasc Pharmacol 1986;8:527–33.
[50] Kirsten R, Nelson K, Kirsten D, et al. Clinical phar-
macokinetics of vasodilators. Clin Pharmacokinet
1998;35:9–36.
[51] Murphy MB, Murray C, Shorten GD. Fenoldopam:
a selective peripheral dopamine-receptor agonist for
the treatment of hypertension. N Engl J Med 2001;
345:1548–57.
[52] Panacek EA, Bednarczyk EM, Dunbar LM, et al.
Randomized, prospective trial of fenoldopam vs so-
dium nitroprusside in the treatment of acute severe
hypertension. Acad Emerg Med 1995;2:959–65.
[53] Tumlin JA, Dunbar LM, Oparil S, et al. Fenoldo-
pam, a dopamine agonist, for hypertensive emer-
gency: a multicenter randomized trial. Acad Emerg
Med 2000;7:653–62.
[54] Goldberg ME, Cantillo J, Nemiroff MS, et al. Fenol-
dopam infusion for the treatment of postoperative
hypertension. J Clin Anesth 1993;5:386–91.
[55] Squara P, Denjean D, Godard P, et al. Enoximome
vs nicardipine during the early postoperative course
of patients undergoing cardiac surgery: a prospective
study of two therapeutic strategies. Chest 1994;106:
52–8.
[56] Lunell NO, Nylund L, Lewander R, et al. Acute effect
of an antihypertensive drug labetolol, on uteroplacen-
tal blood flow. Br J Obstet Gynaecol 1982;89:640–4.
 Cardiol Clin 24 (2006) ix

Foreword
Interventional Cardiology

Michael H. Crawford, MD
Consulting Editor

Interventional cardiology has moved from
coronary angioplasty to percutaneous interven-
tions on the heart and vasculature. This issue of
Cardiology Clinics focuses on coronary artery in-
terventions. We are fortunate to have Dr. Samin
Sharma, who directs one of the premier cardiac
interventional laboratories in the United States,
as guest editor for this issue. He and his colleagues
have contributed a great deal to the advancement
of the technical aspects of coronary interventions.
They and other experts from the United States
have put together an outstanding comprehensive
discussion of current issues in coronary interven-
tions, including a discussion of who should have
a percutaneous coronary intervention and safety
concerns.
The appropriate application of any technology
is often its Achilles’ heel. Being accomplished at
doing something does not necessarily translate to
having good judgment about when to use this
skill. Ideally good judgment is developed over
time; hence the development of interventional
cardiology training programs, boards, and the spe-
cial certificate in this subdiscipline of cardiology.
With coronary artery bypass surgery providing
very similar outcomes, patient selection in more
challenging patients is crucial. Also, advances in
the management of cardiovascular disease and
heightened expectations of patients are bringing
many 80- and 90-year-olds who have advanced
vascular disease to the catheterization laboratory.
Although bypass surgery seems overly aggressive
and is often not welcomed in such patients, one
must be able to comprehend the limits of percu-
taneous interventions also. Many difficult situa-
tions confront the interventional cardiologist
today, and this issue of Cardiology Clinics pro-
vides an excellent database for approaching man-
agement decisions in patients who have coronary
artery disease.
Michael H. Crawford, MD
Division of Cardiology
Department of Medicine
University of California, San Francisco Medical
Center
505 Parnassus Avenue, Box 0124
San Francisco, CA 94143, USA
E-mail address: crawafordm@medicine.ucsf.edu

0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.009 cardiology.theclinics.com
 Cardiol Clin 24 (2006) xi

Preface
Interventional Cardiology

Samin K. Sharma, MD
Guest Editor

Since the introduction of balloon angioplasty
by Andreas Gruentzig in 1977, interventional
cardiology has immensely proliferated in the last
three decades. In 2005, approximately 1.2 million
percutaneous coronary interventional (PCI) pro-
cedures were performed in the United States, and
about 2 million were performed worldwide. This
unprecedented exponential growth in interven-
tional cardiology has been possible because of (1)
continued refinement in technique and the advent
of new devices to improve the success and safety of
PCI, (2) expanded indications by well-defined
randomized clinical trials, and (3) dramatically
reduced rates of restenosis by the advent of drug-
eluting stents (DES). Despite the increasing com-
plexity of cases, the outcome of PCI has continued
to improve, and selected cases are being done on
an ambulatory basis. The need for urgent bypass
surgery resulting from PCI complications has been
almost eliminated, pushing this field further by
allowing freestanding catherization laboratories to
perform PCI without on-site surgery. Interven-
tional cardiologists have now expanded their skills
outside the realm of coronary tree to include
noncoronary vascular interventions: carotids, sub-
clavian, renal, iliac, among others.
This issue of Cardiology Clinics has compiled
a group of world-class authors to provide the
most updated view in this ever-changing and grow-
ing field of interventional cardiology. Important
topics such as lesion classification, intravascular
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.010
ultrasound, current antiplatelet therapy, and vascu-
lar closure devices are presented in a practical
manner. Lesion-specific approaches using various
debulking and thrombectomy devices, bifurcation
lesions, and total occlusions are discussed in detail
with practical tips. The section on DES provides
the update of various DES program and trial results.
The controversies of PCI versus coronary artery
bypass grafting along with recent PCI guidelines
are outlined in a fairly balanced overview. Finally,
quality issues in the catheterization laboratory have
been emphasized to continue to do the best safely.
I am indebted to all the authors who have
made possible this world-class handbook for the
interventional cardiologists, fellows, and support
staff. Undoubtedly it will serve as an important
reference resource for those who are participating
in the field of interventional cardiology. I wish
to thank my colleagues and staff of the Cardiac
Catheterization Laboratory who have helped
refine the overall care of cardiac interventional
patients at the Mount Sinai Hospital, making it
one of the premier institutions in the state.
Samin K. Sharma, MD
Cardiac Catheterization Laboratory &
Interventional Cardiology
Mount Sinai Hospital
One Gustave L. Levy Place, Box 1030
New York, NY 10029, USA
E-mail address: samin.sharma@msnyuhealth.org
cardiology.theclinics.com
 Cardiol Clin 24 (2006) 153–162
Coronary Angiography, Lesion Classification
and Severity Assessment
Annapoorna S. Kini, MD, MRCP
Cardiac Catheterization Laboratory, Cardiovascular Institute, Mount Sinai Hospital,
Box 1030, One Gustave Place, New York, NY 10029, USA
Advances in the technique of coronary in-
tervention over the years have changed the
management of patients with coronary artery
disease, resulting in safer and more effective
percutaneous revascularization in patients pre-
viously deemed at high risk for nonsurgical
approaches. Angiographic factors contributing
to an untoward procedural outcome after coro-
nary revascularization have been characterized,
and a lesion complexity score has been developed.
Recognition of these angiographic risk factors has
proven invaluable for triaging patients to coro-
nary intervention, coronary bypass surgery, or
medical therapy. The post-treatment lumen di-
ameter has been the most important predictor of
late clinical and angiographic recurrence after
coronary revascularization in several single and
multi-center clinical trials. Coronary angiography
(visual or quantitative) is a simple, easy, and
mostly reliable tool for assessing lesion severity,
but it may be inconclusive in the borderline lesions
(40% to 60% diameter obstruction). Anatomical
(using intravascular ultrasound) and physiological
(using coronary flow reserve or fractional flow
reserve) lesion assessment may be required for
adequate lesion evaluation, before and after
percutaneous coronary intervention (PCI).
Lesion classification
Coronary lesion classification is based most
commonly on American College of Cardiology/
American Heart Association (ACC/AHA) task
E-mail address: annapoorna.kini@msnyuhealth.org
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.002
force classification, along with other recent clas-
sifications [1].
The American College of Cardiology/American
Heart Association task force on lesion morphology
A joint task force of the ACC and the AHA
established criteria in 1988 to estimate procedural
success and complication rates after balloon
angioplasty, based on the presence or absence of
specific lesion characteristics (Box 1) [2]. Although
these criteria were developed based solely upon
the task force’s clinical impressions in the era of
balloon angioplasty, the estimates of procedural
success and complications were correlated closely
with the procedural outcomes subsequently dem-
onstrated in patients undergoing multi-vessel cor-
onary angioplasty or stenting. Procedural success
and complication rates were 92% and 2%, respec-
tively, for type A lesions; 76% and 10%, respec-
tively, for type B lesions; and 61% and 21%,
respectively, for type C lesions. Lesions with two
or more type B characteristics (modified ACC/
AHA type B2) had an intermediate risk between
lesions with one type B characteristic (modified
ACC/AHA type B1) and type C lesions. Most of
the studies have categorized B2 and C lesion types
as complex or high-risk lesion characteristics. Spe-
cific lesion characteristics associated with an ad-
verse outcome included chronic total occlusion,
high-grade stenosis, stenosis on a bend of 60 de-
grees or more, and lesions located in vessels with
proximal tortuosity.
Despite the advantages of this composite
approach for estimating lesion complexity, the
ACC/AHA classification system has certain lim-
itations. The definitions used in the classification
cardiology.theclinics.com
154
Box 1. Characteristics of type A, B and C
coronary lesions as per American
College of Cardiology/American Heart
Association classification
Type A lesion (high success, greater
than 85%; low risk)
 Discrete (less than 10 mm)
 Little or no calcium
 Concentric
 Less than totally occlusive
 Readily accessible
 Not ostial in location
 Nonangulated segment, less than 45

 No major side branch involvement
 Smooth contour
 Absence of thrombus
Type B lesions (moderate success, 60%
to 85%; moderate risk)
 Tubular (10 to 20 mm length)
 Moderate to heavy calcification
 Eccentric
 Total occlusions less than 3 months old
 Moderate tortuosity of proximal
segment
 Ostial in location
 Moderately angulated segment,
at least 45
, less than 90

 Bifurcation lesion requiring double
guidewire
 Irregular contour
 Some thrombus present
Type C lesions (low success, less than
60%; high risk)
 Diffuse (at least 2 cm length)
 Total occlusion more than 3 months
old
 Excessive tortuosity of proximal
segment
 Inability to protect major side branches
 Extremely angulated segments at least
90

 Degenerated vein grafts with friable
lesions
system (eg, lesion eccentricity, irregularity, angu-
lation, and tortuosity) are subject to individual
interpretations. As a result, considerable observer
variability has been reported. Some ACC/AHA
lesion features are associated with a complicated
procedure (eg, thrombus and angulated seg-
ments), whereas others are associated with an
KINI
unsuccessful but uncomplicated procedure (eg,
old total occlusions or longer lesions). Owing to
the heterogeneity of the morphologic features
within this classification system, its generalized
use in estimating procedural outcome for all
patients may be problematic. Various preproce-
dural lesion morphologies have been outlined in
Box 2.
Angulated lesions
Vessel curvature at the site of maximum
stenosis should be measured in the most unfore-
shortened projection using a length of curvature
that approximates the balloon length used for
coronary dilation. Percutaneous interventions of
highly angulated (at least 45
) lesions have been
associated with an increased risk of procedural
complications (13% versus 3.5% in nonangulated
stenoses, P !.001), most commonly owing to cor-
onary dissection.
Bifurcation lesions
The risk of side branch (SB) occlusion in
bifurcation lesions relates to the extent of athero-
sclerotic involvement of the SB within its origin
from the parent vessel (14% to 27% in SB with
ostial involvement). To accurately assess the risk
of SB occlusion and avoid conflicting definitions
of SB and ostial stenoses, few classifications were
accepted, most used being Duke classification
(Fig. 1).
Lesion calcification
Angiographic and intravascular studies have
shown that the presence of coronary artery
calcium is an important marker for coronary
atherosclerosis. The presence of coronary artery
calcium also has been related to reduced pro-
cedural success rates after coronary interventions,
in part because of lesion rigidity, development of
dissections at calcified plaque-normal wall inter-
face, and the inability of the atherectomy cutting
chamber to excise the fibrocalcific plaque. In
contrast, higher (greater than 90%) procedural
success rates have been reported after rotational
atherectomy (an atheroablative device creates
microdissection planes within the fibrocalcific
plaque and allows more effective arterial expan-
sion after balloon angioplasty or stent placement).
Despite the prognostic importance of lesion
calcium on procedural outcome after coronary
intervention, conventional angiography has been
 CORONARY ANGIOGRAPHIC LESION ASSESSMENT 155

Box 2. Definitions of preprocedural lesion morphology

Lesion angulationdvessel angle formed by the center line through the lumen proximal
to the stenosis and extending beyond it and a second center line in the straight portion
of the artery distal to the stenosis
 Moderatedlesion angulation at least 45

 Severedlesion angulation at least 90

Calcificationdreadily apparent densities noted within the apparent vascular wall at the site
of the stenosis
 Moderateddensities noted only with cardiac motion prior to contrast injection
 Severedradiopacities noted without cardiac motion prior to contrast injection
Eccentricitydstenosis that is noted to have one of its luminal edges in the outer one quarter
of the apparently normal lumen
Filling defectdan angiographic lucency, usually globular, with contrast surrounding at least
3 sides (or equivalent), divided into three grades: 1 = haziness alone, 2 = defect 1 to 2 mm,
3 = defect greater than 2 mm in diameter
Irregularitydcharacterized by lumen ulceration, intimal flap, aneurysm, or sawtooth pattern
 Ulcerationdlesions with a small crater consisting of a discrete luminal widening in the area
of the stenosis is noted, provided it does not extend beyond the normal arterial lumen
 Intimal flapda mobile, radiolucent extension of the vessel wall into the arterial lumen
 Aneurismal dilationdsegment of arterial dilation larger than the dimensions of the normal
arterial segment
 Sawtooth patterndmultiple, sequential stenosis irregularities
Lesion angleddegrees at end diastole in nonforeshortened view subtended by a 15 mm
treatment device
Lesion ectasiadat least 150% of reference diameter
Lesion lengthdmeasured shoulder-to-shoulder in an unforeshortened view
 Discretedlesion length less than 10 mm
 Tubulardlesion length 10 to 20 mm
 Diffusedlesion length at least 20 mm
Ostial locationdorigin of the lesion within 3 mm of the vessel origin
Total occlusiondTIMI 0 or 1 flow
Nonchronic total occlusiondtotal occlusion not meeting the criteria for chronicity
Thrombusddiscrete, intraluminal filling defect is noted with defined borders and is
separated largely from the adjacent wall. Contrast staining may or may not be present.
Bifurcation stenosisdpresent if a medium or large branch (greater than 1.5 mm) originates
within the stenosis and if the side branch is surrounded completely by stenotic portions of
the lesion to be dilated
Lesion accessibility (proximal tortuosity)
 Moderate tortuositydlesion is distal to two bends at least 75

 Severe tortuositydlesion is distal to three bends at least 75

Degenerated saphenous vein graftdgraft characterized by luminal irregularities or ectasia
comprising more than 50% of the graft length

shown to have limited sensitivity for detecting
smaller amounts of calcium within the coronary
artery.
Degenerated saphenous vein grafts
Few criteria have been proposed for classifying
the degree of graft degeneration, although such
a definition should include an estimate of the
percentage of graft irregularity and ectasia, fria-
bility, presence of thrombus, and number of
discrete or diffuse lesions (greater than 50%
stenosis) located within the graft. These patho-
logic features have been correlated clinically
with graft atherosclerosis and may predispose to
distal microembolization, thrombosis, and other
156
Fig. 1. Bifurcation lesion classification.
complications during saphenous vein graft
intervention.
Eccentric lesions
Pathologic studies have demonstrated that
balloon angioplasty of eccentric lesions may result
in the asymmetric expansion of normal vessel
wall, with little change to the underlying athero-
sclerotic segment. Historically, reduced proce-
dural success rates have been attributed to
eccentric lesions, presumably owing to greater
degrees of elastic recoil and larger residual percent
diameter stenoses in these lesions.
Inaccessible lesions
The vessel tortuosity is assessed before the
lesion that is going to be treated percutaneously
(Fig. 2).
Irregular lesions
Lesion irregularity includes those narrowings
with ulceration, aneurysms proximal or distal to
stenoses, sawtoothed contour suggesting a friable
lesion, and intimal flaps. The presence of lesion
irregularity correlates pathologically with plaque
fissuring, rupture, and platelet and fibrin aggre-
gation. Accordingly, complex, irregular plaques
have been associated with unstable coronary
syndromes and progression to total occlusion,
whereas smooth lumen contours are more sugges-
tive of stable angina. Other surface morphology
KINI
features associated with unstable angina and
infarction include lesion ulceration, sharply angu-
lated leading or trailing borders, multiple serpig-
inous channels, and discrete intraluminal filling
defects. A qualitative scoring index for lesion
irregularity was proposed in 1985, classifying
lesions as:
 Concentric (symmetric narrowing)
 Type I eccentric (asymmetric narrowing with
a broad neck)
 Type II eccentric (asymmetric narrowing with
a narrow neck related to one or more over-
hanging edges or scalloped borders)
 Multiple irregular coronary narrowings in
series
Type II eccentric narrowings were more com-
mon in patients with unstable angina, whereas
concentric and type I eccentric narrowings were
demonstrated more often in those with stable
angina.
Long lesions
Several criteria have been used to assess the
axial length of the atherosclerotic obstruction in
patients undergoing percutaneous coronary in-
tervention. Lesion length may be estimated as
the shoulder-to-shoulder extent of atherosclerotic
narrowing greater than 20% or by the lesion
length with greater than 50% visual diameter
stenosis. Sequential stenoses may be included in
the estimation of lesion length, provided that the
 CORONARY ANGIOGRAPHIC LESION ASSESSMENT 157

Fig. 2. Assessment of lesion tortuosity. (A) No tortuosity or mild tortuosity: no bend or one 75
bend before lesion. (B)
Moderate tortuosity: two bends of 75
before the lesion or one 90
bend before the lesion. (C) Excessive tortuosity: more
than two bends of 75
before the lesion or more than one 90
bend before the lesion.

distance between the sequential lesions does not
exceed 5 mm.
Ostial lesions
Balloon angioplasty of aorto–ostial lesions and
lesions involving the proximal 3 mm of left
anterior descending coronary artery (LAD) or
left circumflex coronary artery (LCX) has been
associated with an unfavorable procedural out-
come, potentially owing to smooth muscle and
eccentric intimal proliferation noted pathologi-
cally in ostial lesions. Technical factors account-
ing for the suboptimal success rates included
difficulties with guide catheter support, lesion
inelasticity precluding maximal balloon inflation,
and the need for multiple balloon exchanges.
Clinical restenosis developed in nearly 50% of
patients over the subsequent 6 months. Although
directional, rotational, and extractional atherec-
tomy; intracoronary stenting; and excimer laser
coronary angioplasty have been used in patients
with ostial lesions, late clinic recurrence still may
be problematic in this location.
quantifying the length of the totally occluded
segment. The risk of an unsuccessful procedure
relates to the duration of the occlusion and certain
lesion morphologic features, such as bridging
collaterals, occlusion length greater than 15 mm,
occlusion duration greater than 3 months, and the
absence of a nipple to guide wire advancement.
Thrombus
The presence of angiographic thrombus, gen-
erally identified by the appearance of discrete,
intraluminal filling defects within the arterial
lumen, also has been associated with a higher,
although widely variable (6% to 73%), incidence
of ischemic complications after coronary inter-
vention, primarily resulting from the occurrence
of distal embolization and thrombotic occlusions.
Although antithrombotic and thrombolytic
agents and mechanical devices have been recom-
mended for selected thrombus-containing lesions,
it often has been difficult to quantitate the in-
cremental benefit achieved with these techniques
over conventional methods.

Total occlusion
Thrombus grades
Total coronary occlusion generally is identified
There are six thrombus grades:
on the cineangiogram as an abrupt termination of
the epicardial vessel; anterograde and retrograde  Grade 0dno thrombus
collaterals may be present and are helpful in  Grade 1dpossible thrombus ¼ mural opacities
158
 Grade 2dmall thrombus ¼ size ! 0.5  nor-
mal lumen diameter
 Grade 3dmedium thrombus ¼ size 0.5–1.5 
normal lumen diameter
 Grade 4dlarge thrombus ¼ size O 1.5  nor-
mal lumen diameter
 Grade 5drecent thrombotic occlusion ¼
fresh thrombus with dye stasis and delayed
washout
 Grade 6dchronic total occlusion ¼ smooth,
abrupt, and with no dye stasis and brisk
flow
Other lesion classifications
Society for Cardiac Angiography and Interventions
The Society for Cardiac Angiography and
Interventions (SCAI) [3] lesion classification sys-
tem divided ACC/AHA lesions into type I to IV:
 Type I lesions (highest success, lowest risk)d
patent and do not meet criteria for C lesions
 Type II lesionsdpatent and meet any of the
criteria for ACC/AHA type C lesions
 Type III lesionsdoccluded and do not meet
criteria for C lesions
 Type IV lesionsdoccluded and meet any of
the criteria for ACC/AHA type C lesions
(Box 3).
Ellis classification
With newer device strategy and use of dual
antiplatelet therapy and intravenous GP IIb/IIIa
inhibitors [4], major complications of coronary
interventions have declined significantly. Hence
a newer scheme to update lesion classification based
on the lesion outcome of over 10,000 patients was
suggested by Ellis and colleagues [5]. Nonchronic
total occlusion (total occlusion not meeting criteria
for chronicity) had the highest odds ratio [4.75 (2.69
to 8.38, P ! .001)] on multivariate analysis to
predict periprocedural complications, along with
degenerated vein graft lesions.
The strongest correlation was with nonchronic
total occlusion with degenerated saphenous vein
graph. There was a moderately strong correlation
with:
 Lesion length of at least 10 mm
 Lumen irregularities or saw tooth
 Lumen edges in absence of thrombus
 Filling defect greater than 2 mm in diameter
 Calcification
KINI
Box 3. American College of Cardiology/
Society for Cardiac Angiography and
Interventions lesion classification
system
Type I lesions (highest success
expected, lowest risk)
Does not meet criteria for ACC/AHA type
C lesion
Nontotal occlusion
Type II lesions
Meets any of these criteria for ACC/AHA
type C lesion:
 Diffuse (greater than 20 mm length)
 Excessive tortuosity of proximal
segment
 Extremely angulated segments,
greater than 90

 Inability to protect major SB
 Degenerated vein grafts with friable
lesions
Nontotal occlusion
Type III lesions
Does not meet criteria for type C lesion
Total occlusion
Type IV lesions
Meets any of these criteria for ACC/AHA
C lesion:
 Diffuse (greater than 2 cm length)
 Excessive tortuosity of proximal
segment
 Extremely angulated segments,
greater than 90

 Inability to protect major SB
 Degenerated vein grafts with friable
lesions
 Occluded for more than 3 months
Total occlusion
 Lesion angle of at least 45

 Eccentricity
 SVG age of at least 10 years
Based on these correlates and PCI outcomes,
lesions were divided further as:
 Class I (low risk)dno risk factors
 Class II (moderate)done to two moderate
correlates and absence of strong correlations
 Class III (high risk)dat least three moderate
correlates and absence of strong correlates
 CORONARY ANGIOGRAPHIC LESION ASSESSMENT
 Class IV (highest risk)deither of strongest
correlates.
Evaluation of lesion severity
Quantitative coronary angiography
Generally, there is good agreement among
interventional cardiologists who visually estimate
stenosis severity regarding the severity of mild or
severe stenosis. In contrast, there is a great deal of
intraobserver and interobserver variability regard-
ing intermediate stenosis. In addition, there is
some variability in the visual estimate of vessel
dimensions. Computer-assisted methods have
been developed to provide a more accurate and
unbiased assessment of absolute and relative
coronary artery dimensions during angiography,
a technique called quantitative coronary angiog-
raphy (QCA). QCA entails digitization of the film,
image calibration, arterial contour editing, and
observer editing. In addition to the inherent
shortcomings of individual QCA systems, there
are errors common to all systems at each stage (ie,
image acquisition and analysis may be performed
during systole, thus skewing the results). More-
over, observer editing may render this objective
technique operator-dependent and susceptible to
bias. The major advantage of coronary angiogra-
phy over other techniques of lesion assessment is
the ability to assess the severity of the lesion
without the need to cross the lesion with a guide-
wire or other devices.
Pitfalls of coronary angiography
The angiographic assessment of coronary artery
lesions is based on the comparison of radio-
contrast dye opacification of the lesion relative to
a presumed normal reference segment. The true
lumen diameter of the reference segment, however,
may be measured inaccurately, because it is dis-
eased diffusely and concentrically. Likewise, if the
true dimensions of the vessel are underestimated,
undersized balloon and stents may be selected for
coronary intervention, resulting in suboptimal re-
sults and use of additional equipment [6].
Also, because of vascular remodeling, the
vessel dimensions at the site of stenosis may
change over time relative to the reference segment.
The vessel may shrink or grow in diameter focally
at the site of stenosis.
Another pitfall of coronary angiography may
stem from the eccentricity of the lesion. Even
using multiple planar views, it is often difficult to
159
fully delineate the lesion, especially in the presence
of ostial or bifurcation lesions or overlapping
branches. Therefore, the lesion may be much more
severe than appreciated angiographically.
Coronary angiography also may fail to char-
acterize the composition of the atheroma accu-
rately, which is an important factor in the choice
of the device used in a possible intervention. For
example, calcifications and intracoronary thrombi
often produce the same angiographic picture. A
heavily calcified lesion may be more suitable for
rotational atherectomy, whereas a lesion with
a large thrombus burden would contraindicate
rotational atherectomy and would be dealt with
better with thrombectomy devices.
Intravascular ultrasound
Intravascular ultrasound (IVUS) has become
the gold standard for delineating vessel wall
anatomy and plaque morphology [7]. Current
IVUS catheters with an outer diameter of between
2 to 3 Fr can be introduced by means of 6 Fr guid-
ing catheters. The catheter is placed distal to the
segment of interest and is pulled back gradually.
Motorized pull-back devices are available, en-
abling three-dimensional reconstruction of the
vessel wall. In many large prospective series, in
approximately 20% of examinations before coro-
nary interventions, IVUS changed the treatment
strategy by demonstrating more severe or milder
coronary artery disease than appreciated by angi-
ography. Also, IVUS remains an invaluable tool
to evaluate adequate stent expansion or other sub-
optimal angiographic results.
Physiologic evaluation
Intracoronary Doppler
The coronary angiogram or IVUS offers no
information regarding the coronary microcircula-
tion, or of the physiologic significance of lesions.
A physiologically significant lesion impairs coro-
nary blood flow (CBF) at rest, or more commonly
during stress [8,9].
At rest, myocardial demand is low, and accord-
ingly CBF is at its lowest level. Under conditions of
increased stimulation, the normal physiologic re-
sponse to an increase in myocardial demand is
enhanced CBF by vasodilation of epicardial and
resistance vessels. The ability to increase CBF from
resting CBF by reducing vasomotor tone to meet
myocardial demand (hyperemia) is called coronary
flow reserve (CFR). Normal individuals can
160
increase CBF three- to fivefold to meet increased
myocardial demand.
In the presence of a physiologically significant
lesion, the resistance vessels compensate for the
impaired CBF by vasodilating. In the case of
a severe lesion, the resistance vessels are dilated
fully. Thus, in response to a physiological or
pharmacological stimulus that increases myocar-
dial demand, the resistance vessels are not capable
of further vasodilating, constituting a state of
impaired CFR. Gould and Lipscomb demon-
strated that the CFR is attenuated beginning
with coronary artery stenosis of more than 50%
of the diameter. Compensatory vasodilation of
the distal coronary vascular bed maintains near
normal resting flow for lesions between 70% and
85% of diameter stenosis, but adaptive vasodila-
tion fails to compensate for lesions greater than
85% of diameter stenosis. These findings have
served as reference for the current definition of
obstructive coronary artery disease. Thus, it is
accepted that at least 70% stenosis of an epicar-
dial artery constitutes significantly obstructive
coronary artery disease. It is clear, however, that
lesions estimated to be 50% to 70% of diameter
stenosis also may be physiologically significant,
and hence may merit further evaluation. Using
physiological assessments of intermediate lesions,
it has been demonstrated that it is possible to
safely defer an intervention in patients who have
normal physiological parameters [10–12].
The CFR, defined as the ratio of hyperemic
blood flow to resting blood flow, can be measured
using a 0.014 inch intracoronary Doppler guide
wire. CFR blood flow is calculated using the
formula pD2  APV O 8, where D represents
the coronary diameter measured 5 mm distal to
the tip of the Doppler wire (by quantitative angi-
ography or IVUS), and APV equals the average
peak velocity from the Doppler tracing. The
CFR is calculated by the ration of peak-to-base-
line CBF in response to drug infusion or injection.
When coronary artery diameter is presumed to
remain unchanged in response to drug manipula-
tion, the CFR is calculated by the ratio of peak-
to-baseline flow velocity (APV). Normal value is
greater than 2.5.
Common pharmacological drugs are used to
detect abnormalities in CFR. Adenosine is
thought to act on the coronary vasculature by
stimulating the adenosine A2 receptor on smooth
muscle cells. At high doses, adenosine can cross
the endothelial barrier and stimulate the receptor
on the smooth muscle directly in an endothelium-
KINI
independent mechanism. Adenosine acts predom-
inantly on vessels less than 150 mm in diameter,
and, therefore, mainly assesses changes in the
coronary resistance vessels as reflected by changes
in coronary flow. The administration of adenosine
provides mainly an endothelium-independent
evaluation of the coronary microvasculature.
Adenosine may cause bradyarrhythmias including
sinus bradycardia and atrioventricular block,
facial flushing, and bronchoconstriction. Because
of the short half-life of adenosine, the duration of
these effects is very brief.
It is important to recognize several possible
pitfalls in the measurement of CFR using Doppler
wire. Systemic conditions that may affect systemic
hemodynamics such as thyrotoxicosis and anemia
also may affect basal CFR. The microvasculature
in infarcted areas of the myocardium may be
impaired functionally. Caution should be exer-
cised that the guidewire tip should not abut the
vessel wall, should not be in a small branch,
should not be distorted grossly in shape, and
should not be placed in the proximity to a major
bifurcation. In the left coronary artery system, the
blood flow velocity in diastole is greater than in
systole, perhaps because of the greater compres-
sion of the left ventricle during systole, whereas
for the right coronary artery, the flow velocities
are fairly similar during both phases. Although
the basal CBF may be highest in the left anterior
descending coronary artery and lowest in the right
coronary artery, the CFR is fairly similar for all
three major epicardial arteries. Because of the
increased sensitivity of the right coronary artery
to the pharmacological agents used, however,
especially the chronotropic effects of adenosine,
lower doses (18 to 24 mg) are recommended ini-
tially for injections to the right coronary artery,
and doses of 30 to 40 mg are recommended for
the left coronary system.
Coronary microvessels may have reduced vas-
odilating abilities because of structural or func-
tional abnormalities (such as hypertension,
diabetes mellitus), resulting in the ability to de-
crease vasomotor tone during stress. Therefore,
in case of a low CFR ratio, CFR should be
measured in another angiographically normal
coronary artery. If the CFR is abnormally low
in the control artery, diffuse coronary microvas-
cular disease or another pathology should be
sought (eg, diabetes mellitus or left ventricular
hypertrophy). If the CFR in the control artery
is normal, it is reasonable to assume that the le-
sion in question is physiologically significant.
 CORONARY ANGIOGRAPHIC LESION ASSESSMENT
Fig. 3. Schematic flow chart to use anatomical and physiological lesion testing in decision-making process in the catheter
laboratory.
Calculation of the relative CFR as the ratio of the
CFR of the target lesion divided by the CFR of
the normal vessel may be a more accurate
assessment of the culprit lesion and generalized
coronary microvasculature.
The use of CFR in cases with microvascular
disease remains controversial. The inability of the
microcirculation to respond to adenosine may
result in inaccurate results. The ability of the
Doppler wire to assess the microcirculation,
however, may be considered an advantage in
certain circumstances (ie, an impaired CFR mea-
sured in more than one epicardial artery may
indicate coronary microvessel disease that may
account for the patient’s symptoms).
Fractional flow reserve
Fractional flow reserve (FFR) is a method for
assessing indeterminate coronary artery stenoses
based on pressure wire analysis during maximal
flow. The concept of myocardial FFR, defined as
the maximal blood flow to the myocardium in the
presence of a stenosis in the supplying coronary
artery, divided by the theoretical normal maximal
flow in the same distribution, has been developed as
an index of physiologic severity of the lesion. FFR
represents the fraction of the normal maximal
myocardial flow that can be achieved despite the
coronary stenosis. This index can be calculated
from the ratio of the mean distal coronary artery
pressure to the aortic pressure during hyperemic
maximal vasodilation. It is independent of changes
161
in systemic blood pressure, heart rate, and status of
microvascular circulation [13].
Moreover, the FFR takes into account the
collateral blood supply. Per definition, the normal
value of the FFR is 1.0 for any vessel investigated.
Based on prior studies, it is accepted that an index
value of less than 0.75 is abnormal and correlates
well with pathological findings using noninvasive
techniques. Appropriate interventions may reduce
the future cardiac events. Values between 0.75 and
0.90 are of intermediate significance, and clinical
correlation is suggested.
After calibration at zero, the wire is positioned
distal to the lesion at question, and adenosine is
administered (18 to 40 mg bolus) into the ostium
of the coronary artery through the guiding cathe-
ter. The distal coronary pressure is monitored;
maximal CBF is achieved with minimal coronary
distal pressure. When the coronary distal pressure
reaches a new steady state, the FFR is calculated
by dividing the mean distal intracoronary pressure
(measured by the pressure wire) by the mean arte-
rial pressure (measured by the guiding catheter
positioned in the ostium of the coronary artery).
Summary
Coronary angiography remains the gold stan-
dard of day-to-day lesion assessment. Other
modalities may help to further improve the un-
derstanding of the lesion morphology for better
management (Fig. 3).
162
References
[1] Ellis SG, Vandormael MG, Cowley MJ, et al. Coro-
nary morphologic and clinical determinants of pro-
cedural outcome with angioplasty for multi-vessel
coronary disease: implications for patient selection.
Circulation 1990;82:1193–202.
[2] Ryan TJ, Faxon DP, Gunnar RM, et al. Guidelines
for percutaneous transluminal coronary angio-
plasty: a report of the American College of Cardiol-
ogy/American Heart Association Task Force on
assessment of diagnostic and therapeutic cardiovas-
cular procedures (subcommittee on percutaneous
transluminal coronary angioplasty). J Am Coll Car-
diol 1988;12:529–45.
[3] Krone RJ, Shaw RE, Klein LW, et al. Evaluation of
the American College of Cardiology/American
Heart Association and the Society for Coronary An-
giography and Interventions lesion classification sys-
tem in the current stent era of coronary interventions
(from the ACC-National Cardiovascular Data Reg-
istry). Am J Cardiol 2003;92:389–94.
[4] Investigators EPISTENT. Randomised placebo-
controlled and balloon angioplasty-controlled trial
to assess safety of coronary stenting with use of
platelet glycoprotein-IIb/IIIa blockade. Lancet
1998;352:87–92.
[5] Ellis SG, Guetta S, Miller D, et al. Relation between
lesion characteristics and risk with percutaneous in-
tervention in the stent and glycoprotein IIb/IIIa
eradan analysis of results from 10 907 lesions and
proposal for new classification scheme. Circulation
1999;100:1971–6.
KINI
[6] Saucedo JF, Lansjy AJ, Ito S, et al. A practical ap-
proach to quantitative coronary angiography. In:
Beyar R, Keren G, Leon MB, et al, editors. Frontiers
in interventional cardiology. London: Martin
Dunitz Publishers; 1997. p. 281–96.
[7] Di Mario C, Gorge G, Peters R, et al. Clinical appli-
cation and image interpretation in intracoronary
ultrasound. Eur Heart J 1998;19:207–29.
[8] Pijls NH, de Bruyne B, Peels K, et al. Measurement
of fractional flow reserve to assess the functional se-
verity of coronary artery stenoses. N Engl J Med
1996;334:1703–8.
[9] Gould KL, Lipscomb K. Effects of coronary steno-
ses on coronary flow reserve and resistance. Am
J Cardiol 1974;34:48–55.
[10] Kern M, Donohue T, Aguirre F, et al. Clinical
outcome of deferring angioplasty in patients with
normal translesional pressure flow velocity measure-
ments. J Am Coll Cardiol 1995;25:178–87.
[11] Serruys P, di Mario C, Piek J, et al. Prognostic value
of intracoronary flow velocity and diameter stenosis
in assessing the short- and long-term outcomes of
coronary balloon angioplasty: the DEBATE study
(Doppler Endpoints Balloon Angioplasty Trial
Europe). Circulation 1997;96:3369–77.
[12] Kern MJ, Meier B. Evaluation of the culprit plaque
and the physiological significance of coronary ath-
erosclerotic narrowings. Circulation 2001;103:
3142–9.
[13] Pijls NH, de Bruyne B, Peels K, et al. Measurement
of fractional flow reserve to assess the functional
severity of coronary-artery stenoses. N Engl J Med
1996;334:1703–8.
 Cardiol Clin 24 (2006) 163–173

Intravascular Ultrasound in the Current

Percutaneous Coronary Intervention Era
Ilke Sipahi, MDa, Stephen J. Nicholls, MBBS, PhDb,
E. Murat Tuzcu, MDa,c,*
a
Intravascular Ultrasound Core Laboratory, Department of Cardiovascular Medicine,
The Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk JJ65, Cleveland, OH 44195, USA
b
Interventional Cardiology, Department of Cardiovascular Medicine, The Cleveland Clinic Foundation,
9500 Euclid Avenue, Desk JJ65, Cleveland, OH 44195, USA
c
Cleveland Clinic Lerner College of Medicine, Case Western Reserve University, 9500 Euclid Avenue,
Cleveland, OH 44195, USA

Intravascular ultrasound (IVUS) is an imaging
modality that can bring a tomographic perspective
to percutaneous coronary interventions (PCI).
Although contrast angiography allows evaluation
of lumen of coronary arteries in a planar fashion, it
does not permit visualization of the arterial wall,
which harbors the atherosclerotic tissue. IVUS is
capable of showing the arterial wall and the lumen
of the coronary arteries with high spatial resolution
and across the full 360
circumference of the vessel.
Thus, it provides additional information beyond
what is obtained from angiography. The use of
IVUS in the cardiac catheterization laboratory has
continued to evolve since its introduction almost
15 years ago. In this review, the authors examine
the role of IVUS in the current PCI era, which is
dominated by the use of drug-eluting stents (DES).
Normal arterial anatomy and basic measurements
by intravascular ultrasound
Normal coronary arteries usually have a tri-
layered appearance on IVUS imaging, which
corresponds to the three histologic layers of the
arterial wall. The innermost layer is the echogenic
intima, the middle layer is the echolucent media,
and outermost layer is the echogenic adventitia
(Fig. 1). The upper limit of normal intimal thick-
ness is considered to be 0.25 to 0.50 mm [1,2]. Lu-
men cross-sectional area (CSA) is determined by
tracing the lumen–intima interface (Fig. 2). Mini-
mum and maximum lumen diameters are the
shortest and longest diameters through the center
of the lumen. Because the outer border of adven-
titia is usually indistinct on IVUS imaging, total
arterial CSA is measured by tracing the trailing
edge of media and is referred to as external elastic
membrane (EEM) CSA. Atheroma CSA is calcu-
lated as EEM CSA minus lumen CSA. Because
atheroma CSA also includes the area occupied
by the media, it is sometimes referred to as plaque
plus media CSA. Atheroma CSA divided by
EEM CSA yields percent CSA stenosis (or plaque
burden). Percent CSA stenosis is distinct from
percent diameter stenosis obtained by angio-
graphy because it is the fraction of arterial CSA
occupied by atheroma at a single cross-section
and does not use reference segments, unlike angi-
ography. Detailed descriptions of standards of
image acquisition and interpretation are available
in the American College of Cardiology and the
European Society of Cardiology expert consensus
documents on IVUS [3,4].

* Corresponding author. Department of Cardiovas- Evaluation of intermediate coronary lesions

cular Medicine, The Cleveland Clinic Foundation, 9500
Euclid Avenue, Desk F25, Cleveland, OH 44195. Making decisions for revascularization is a
E-mail address: tuzcue@ccf.org (E.M. Tuzcu). challenging task in patients who have intermediate
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.01.003 cardiology.theclinics.com
164 SIPAHI
Fig. 1. IVUS image showing the trilaminar appearance
of normal coronary arterial wall. White arrow points
to the intima and black arrow points to the adventitia.
The echolucent space between these two echogenic layers
is the media. Note that all three layers may not be appre-
ciable in some cases.
coronary lesions (ie, 40%–70% angiographic
diameter stenosis). Intermediate lesions can be
particularly troublesome in patients whose
symptomatic status is difficult to evaluate. Vessel
overlapping and tortuosity, lesion eccentricity,
ostial localization, and severely calcified lesions
can further compound the problem. In this context,
IVUS is the alternative imaging modality and
provides a tomographic perspective that often
permits precise quantification of lumen and plaque
Fig. 2. IVUS image (right panel ). Basic IVUS measurements (left panel ). The area bound by the outer border of the
echolucent media is the external elastic membrane (EEM) area. The area bound by the lumen-intima interface is the
lumen area. EEM area minus lumen area yields the atheroma area.
et al
sizes. In addition, IVUS can help to assess the
plaque morphology.
When an intermediate lesion is encountered,
functional assessment using pressure or flow
(Doppler) measurements is an important alterna-
tive to IVUS. Measurements of fractional flow
reserve (FFR) and coronary flow reserve using
miniaturized sensors have proved useful in iden-
tifying lesions of hemodynamic significance [5,6].
FFR data have been prospectively validated,
and these studies have shown that deferring le-
sions with intermediate severity that are hemo-
dynamically insignificant (ie, FFRO0.75) have
favorable clinical follow-up [7]. The use of IVUS
is preferred over functional studies when the ana-
tomic distribution (eg, involvement of ostium or
bifurcation) or the morphology of the lesion
(eg, ulceration versus calcification as the cause of
haziness on angiogram) is also important for
decision-making. Unlike the case with functional
studies, the accuracy of the IVUS cutoff values
has not been validated prospectively by noninva-
sive tests of myocardial ischemia such as myocar-
dial perfusion scintigraphy. Similarly, clinical
outcome data for these cutoff values have been
evaluated only in relatively small retrospective
studies. The cutoff values for IVUS measurements
that identify hemodynamically significant lesions
have mostly been determined by correlating the
lumen dimensions obtained during IVUS imaging
with FFR and coronary flow reserve. A study of
51 lesions (not involving the left main coronary
artery [LMCA]) found that a minimal lumen
CSA of less than 3 mm2 was the best cutoff value
 INTRAVASCULAR ULTRASOUND IN CURRENT PCI ERA
for predicting hemodynamically significant lesions
by FFR [8]. In a later study of 53 lesions, a mini-
mal lumen CSA of less than 4 mm2 and a minimal
lumen diameter of less than 1.8 mm were found to
be the best predictors of hemodynamic signifi-
cance [9]. The apparent discrepancy between the
cutoff values for minimal lumen CSA obtained
in these two studies (3 versus 4 mm2) may be
due to the different agents used for inducing hy-
peremia (papaverine versus adenosine) or due to
the very few number of lesions that had a minimal
lumen CSA between 3 and 4 mm2 in the former
study. Because favorable clinical follow-up data
are available for the minimum lumen area of 4.0
mm2 [10], it is the authors’ preference to use this
parameter for decision making regarding revascu-
larization for lesions that have intermediate sever-
ity. In the gray zone of 3 to 4 mm2, the additional
use of the criterion ‘‘percent CSA stenosis greater
than 70%’’ [9] can be decisive because combining
different criteria has been shown to increase spec-
ificity for hemodynamic significance [8]. It has
been shown that a more precise evaluation of le-
sion severity using IVUS can change the decision
about whether a lesion should be intervened in
about 13% of the cases [11].
Evaluation of the left main coronary artery
It is often difficult to quantify the severity of
LMCA stenoses by angiography [12,13]. Contrast
in the aortic cusp can obscure the proximal part of
the vessel, requiring the angiographer to engage
the LMCA with the catheter and depend on the
reflux of contrast to visualize the ostium. On the
distal end, bifurcation or trifurcation into daugh-
ter branches may preclude accurate assessment. In
some cases, the left main trunk is diffusely dis-
eased, leaving no normal segment that can be
used as a reference site. Because of these limita-
tions, the reproducibility of quantitative coronary
angiography (QCA) of the LMCA is worse than it
is for the other coronary arterial segments [14,15].
When angiographic interpretation of the LMCA
is uncertain, IVUS is usually a helpful modality
because it does not suffer from the previously
mentioned limitations of angiography (Fig. 3).
Certain technical issues should be considered
when imaging the LMCA with IVUS. The IVUS
probe should be placed in the straighter distal
vessel (commonly the left anterior descending)
and the guiding catheter should be disengaged to
miss the ostium during the pullback. Selection of
the guiding catheter is important to permit coaxial
165
imaging. Short-tip JL-4 guiding catheters gener-
ally provide the ability to withdraw the tip while
coaxiality is maintained.
Due to the larger caliber of the LMCA, criteria
for the significance of LMCA lesions as assessed
by IVUS are different from the rest of the
coronary tree. In a study of 55 patients who had
angiographically intermediate LMCA stenoses
(49 G 15% by QCA), a minimal luminal diameter
of less than 2.8 mm and a minimal luminal CSA of
less than 5.9 mm2 were found to predict hemody-
namically significant lesions with a sensitivity and
specificity greater than 90% [16]. The cutoff value
of 2.8 mm for minimal lumen diameter is also sup-
ported by a study with clinical end points [17].
This study, which included 122 patients who had
intermediate stenoses in the LMCA (42 G 16%
by QCA), showed that minimal luminal diameter
of the LMCA as measured by IVUS was the
strongest predictor of cardiac events during the
1-year follow-up. In this study, a minimal luminal
diameter of 3 mm performed best as a threshold
value and, accordingly, patients who had a mini-
mal lumen diameter greater than 3 mm had an
event rate of only 3%. A recent study that also
used clinical end points has suggested that a mini-
mal lumen CSA of 7.5 mm2, a value higher than
the 5.9 mm2 dictated by the FFR study, should
be used as the cutoff value for performing revascu-
larization [18]. This value, however, was obtained
not according to event rates but by calculating the
mean minus 2 SD of minimal lumen areas in
a group of patients who had completely normal
LMCAs. As a result, it is the authors’ preference
to use the criterion ‘‘minimal lumen diameter less
than 2.8 mm’’ for revascularization in LMCA
lesions, which has been validated by clinical end
point and functional studies.
Intravascular ultrasound for guidance of stenting
IVUS imaging has been instrumental in un-
derstanding the arterial responses to almost every
single interventional modality and has helped
to improve the technical details about the
manufacturing and the use of most of the coro-
nary devices. Despite its profound impact on the
refinement of various percutaneous coronary pro-
cedures, it should be noted that routine IVUS
guidance for coronary interventions, particularly
while using DES, is not a requirement.
IVUS imaging has played a pivotal role,
especially in optimizing the technique of stent
deployment as it is currently practiced. IVUS
166
Fig. 3. Coronary angiogram (left panel ) shows severe stenosis of the LMCA ostium; however, IVUS reveals that the left
main stem has no atherosclerotic disease (middle and right panels ). On the other hand, the lumen of the proximal vessel is
smaller than that of the distal vessel, which is the reason of the angiographic appearance. This phenomenon is referred to as
‘‘reverse tapering.’’
observations revealed that the earlier technique of
coronary stenting used in the initial randomized
stent trials frequently resulted in incomplete
expansion and imperfect apposition of the stent
struts, despite the satisfactory angiographic re-
sults [19,20]. These observations led to critical re-
finements in the technique of stenting, which
included higher-pressure postdilations and the
use of larger-sized balloons. These developments,
together with the use of dual antiplatelet therapy,
substantially reduced the incidence of subacute
stent thrombosis [21], and the use of oral anticoa-
gulation was no longer needed.
After the observation of the adverse conse-
quences of underexpanded and unapposed stents,
which included stent thrombosis and restenosis,
the possible benefits of IVUS-guided stenting over
angiography-guided stenting were evaluated vig-
orously in several nonrandomized registries
[22–26], randomized trials [27–29], and in a meta-
analysis of these data sources [30]. The incidence
of death or myocardial infarction was not altered
by IVUS guidance in any of these studies. In the
registries, IVUS guidance generally reduced the
angiographic binary restenosis and target vessel
revascularization rates, probably by promoting
larger postprocedural minimal lumen CSAs. How-
ever, randomized trials produced mixed results.
In the relatively small Restenosis After IVUS-
Guided Stenting (RESIST) trial, there was a non-
significant reduction in the angiographic binary
restenosis rate in the IVUS-guided arm (28.8%
to 22.5%, P ¼ 0.25) [27]. Optimization with
ICUS to reduce stent restinosis (OPTICUS),
the largest reported randomized trial on the
comparison of IVUS-guided versus angiogra-
phy-guided stenting, specifically included lesions
25 mm or less in length that had a 2.5 mm or
SIPAHI et al
greater reference segment diameter. Results of
OPTICUS showed similar restenosis rates
(24.5% versus 22.8%, P ¼ 0.68) and similar tar-
get vessel revascularization rates in the two arms
[28]. The thrombocyte activity evaluation and ef-
fects of ultrasound guidance in long coronary
stent replacement (TULIP) Study [29] investi-
gated the role of IVUS in stenting diffuse
(R20 mm) coronary lesions. In this study, the
restenosis rate was lower with IVUS guidance
(46% versus 23%, P ¼ 0.008), as was the target
lesion revascularization rate (14% versus 4%,
P ¼ 0.037). Postintervention minimal lumen dia-
meter, the number of stents used in each patient,
average stent length, and final balloon size were
all greater in the IVUS-guided group. The posi-
tive impact of IVUS guidance in TULIP, as op-
posed to the neutral results of OPTICUS,
implies that IVUS guidance can reduce resteno-
sis in patients who are at a particularly high risk
for restenosis. Various criteria have been sug-
gested for optimal stent expansion using IVUS-
guidance [31–33]. For implanting a bare metal
stent, the criteria ‘‘minimal stent CSA greater
than 7 mm2’’ and ‘‘minimal lumen CSA divided
by reference EEM CSA greater than 0.55’’ are
probably the most useful (Fig. 4) [33].
With DES, the frequency of restenosis has
been reduced to less than 10% [34,35]. Therefore,
with the more widespread use of DES, the poten-
tial advantage of IVUS guidance to reduce reste-
nosis is less significant. Stent underexpansion,
however, may still be a major cause of restenosis
with DES. Two studies have shown that with siro-
limus-eluting stents, a postprocedural minimum
stent area of less than 5 mm2 was a strong predic-
tor of angiographic binary restenosis or a follow-
up minimal lumen area of less than 4 mm2 [36,37],
 INTRAVASCULAR ULTRASOUND IN CURRENT PCI ERA 167

Fig. 4. Example of a well-expanded stent (upper panels). The stent area is 8.0 mm2. The reference lumen area is 9.3 mm2
and reference EEM area is 13.5 mm2 (lower panels). These measurements show that this is a well-expanded stent by most
criteria. Note the complete apposition of the stent struts to the arterial wall.

which is an indicator of a hemodynamically signif-
icant lesion [9]. In addition to its association with
restenosis, stent underexpansion has been associ-
ated with stent thrombosis in DES [38]; thus, it
is thought that in selected patients, higher infla-
tion pressures (O18 atm) may be needed to
achieve a minimum acceptable stent CSA and,
therefore, to prevent restenosis and thrombosis
in DES [37]. Although routine IVUS guidance
for better expansion of DES is probably not feasi-
ble, IVUS examination may be warranted when it
is difficult to choose the diameter of the stent to be
placed (eg, when there is a large difference in the
diameters of the proximal and the distal reference
segments or when the artery is diffusely diseased).
Although there is no firm guideline to determine
the stent size to be placed in a diffusely diseased
artery using IVUS imaging, one can use the
method followed in the Clinical Outcomes with
Ultrasound Trial (CLOUT) in which IVUS was
used for balloon sizing [39]. Accordingly, if a ‘‘nor-
mal’’ reference diameter is not available, one can
estimate the stent diameter to be the average of
the lumen and EEM diameters at the least dis-
eased site. In addition, stent length is particularly
important when a DES is used because in addition
to stent underexpansion, residual proximal or dis-
tal segment stenosis on IVUS imaging has been
associated with stent thrombosis [37]. Therefore,
it is important to cover the entire diseased area.
IVUS is also useful for stent positioning when
the degree of osteal involvement cannot be dis-
cerned by angiography. In such situations, IVUS
can reliably determine whether the ‘‘true ostium’’
is diseased, in which case the stent should cover
the ostium and can actually protrude into the par-
ent branch. In other cases in which a few millime-
ters of the most proximal portion of the target
168
vessel harbors no disease, the stent can be an-
chored to the most proximal disease-free segment.
In the past, IVUS has been used to evaluate
morphology of the lesions in an effort to optimize
selection of interventional devices (ie, rotational
atherectomy for heavily calcified lesions versus
directional atherectomy for others). With the pre-
dominance of stenting, and particularly of DES,
evidence supporting the use of IVUS imaging to aid
in PCI device selection has become obsolete. IVUS,
however, can still be helpful in cases of directional
coronary atherectomy because angiography is
sometimes incapable of localizing the tissue for
retrieval, especially in eccentric lesions.
Assessing complications of intervention using
intravascular ultrasound
Coronary dissection is the most common
reason for acute arterial closure during PCI and
can result in serious complications including
death, myocardial infarction, and emergent by-
pass surgery. In addition, residual dissection after
stenting remains a risk factor for subacute stent
thrombosis in the DES era [40]. For detection of
dissections, IVUS is a much more sensitive imag-
ing modality than angiography. The circumferen-
tial and the longitudinal extent of coronary
dissections can be better appreciated by IVUS
(Fig. 5). Because IVUS has not been shown to
improve outcomes when evaluating coronary dis-
sections, interventionalists prefer to assess this
condition by angiography alone. Most nonflow
limiting dissections that have no high-risk features
Fig. 5. IVUS image (left panel ). Example of a coronary dissection occurring after balloon dilatation (right panel ). The
intimal tear at the 12-o’clock position connects the true lumen with the false lumen.
SIPAHI et al
on angiography (ie, extraluminal dye staining, fill-
ing defects, or spiral dissections) can be treated
conservatively without additional mechanical in-
terventions. Whenever there is an indication for
stenting, IVUS imaging usually reveals involve-
ment of a longer arterial segment than can be
appreciated angiographically. This additional in-
volvement may be particularly important in cases
of bailout stenting for threatened acute closure, in
which it is critical to cover the entirety of the dis-
sected segment. In such cases, presence of a resid-
ual dissection in the vicinity of the bailout stent or
stents adds to the already higher risk of stent
thrombosis [41]. Residual dissection also remains
a risk factor for subacute stent thrombosis in the
DES era [40], although the absolute risk of throm-
bosis seems to be low.
‘‘Peristent haziness’’ is used to refer to the
nonhomogenous density or ground-glass appear-
ance on the angiogram that generally occurs
proximal or distal to the stented segment. In some
cases of peristent haziness, an obvious cause (eg,
a dissection or a thrombus) can readily be seen on
angiograms. In others, there is no obvious cause,
and such persistent haziness has been reported to
be 15% with high-pressure stenting. In these types
of cases in which the etiology cannot be readily
appreciated by angiogram, IVUS imaging can be
helpful. With IVUS, major causes of this phenom-
enon have been found to be dissections, significant
step-down of luminal area from the edge of the
stent to a segment of moderate disease [42], and cal-
cifications without any dissections [43]. In the cases
of calcification or luminal step-down as the cause of
persistent haziness, IVUS imaging can prevent
 INTRAVASCULAR ULTRASOUND IN CURRENT PCI ERA 169

Box 1. American College of Cardiology/American Heart Association recommendations

for coronary intravascular ultrasound

Class I
None
Class IIa
1. Assessment of the adequacy of deployment of coronary stents, including the extent of
stent apposition and determination of the minimum luminal diameter within the stent.
(Level of evidence: B)
2. Determination of the mechanism of stent restenosis (inadequate expansion versus
neointimal proliferation) and to enable selection of appropriate therapy (plaque ablation
versus repeat balloon expansion). (Level of evidence: B)
3. Evaluation of coronary obstruction at a location difficult to image by angiography in
a patient with a suspected flow-limiting stenosis. (Level of evidence: C)
4. Assessment of a suboptimal angiographic result following PCI. (Level of evidence: C)
5. Diagnosis and management of coronary disease following cardiac transplantation. (Level
of evidence: C)
6. Establish presence and distribution of coronary calcium in patients for whom
adjunctive rotational atherectomy is contemplated. (Level of evidence: C)
7. Determination of plaque location and circumferential distribution for guidance of
directional coronary atherectomy. (Level of evidence: B)
Class IIb
1. Determine extent of atherosclerosis in patients with characteristic anginal
symptoms and a positive functional study with no focal stenoses or mild coronary
artery disease on angiography. (Level of evidence: C)
2. Preinterventional assessment of lesional characteristics and vessel dimensions as
a means to select an optimal revascularization device. (Level of evidence: C)
Class III
1. When angiographic diagnosis is clear and no interventional treatment is planned. (Level
of evidence: C)
This document employs the American College of Cardiology/American Heart Association
style classification of class I, II, or III. These classes summarize the indications for PCI
as follows:
Class Idconditions for which there is evidence for and/or general agreement that the
procedure or treatment is useful and effective
Class IIdconditions for which there is conflicting evidence and/or a divergence of opinion
about the usefulness/efficacy of a procedure or treatment
Class IIadweight of evidence/opinion is in favor of usefulness/efficacy
Class IIbdusefulness/efficacy is less well established by evidence/opinion
Class IIIdconditions for which there is evidence and/or general agreement that the
procedure/treatment is not useful/effective and, in some cases, may be harmful.
The weight of evidence in support of the recommendation for each listed indication
is presented as follows:
Level of evidence Addata derived from multiple randomized clinical trials
Level of evidence Bddata derived from a single randomized trial or nonrandomized studies
Level of evidence Cdconsensus opinion of experts

From Smith SC Jr, Dove JT, Jacobs AK, et al. ACC/AHA guidelines for percutaneous coronary inter-
vention (revision of the 1993 PTCA guidelines)dexecutive summary: a report of the American College of
Cardiology/American Heart Association task force on practice guidelines (Committee to Revise the 1993
Guidelines for Percutaneous Transluminal Coronary Angioplasty) endorsed by the Society for Cardiac
Angiography and Interventions. Circulation 2001;103(24):3019–41.
170
unnecessary deployment of additional stents and,
therefore, may reduce the risk of restenosis and
the cost of the procedure.
Intravascular ultrasound for the management
of stent restenosis
Despite advances in the technique of stent
deployment, adjuvant pharmacologic therapies,
and the use of DES that inhibit neointimal pro-
liferation, stent restenosis remains a problem of
interventional cardiology.
IVUS has provided substantial insight into the
mechanisms of stent restenosis [44–46]. Using
IVUS, it was found that about 20% of bare metal
stent restenosis cases harbor unexpanded stents
[47,48]. The relative contribution of stent underex-
pansion is even greater in cases of DES restenosis
[37,49,50].
The authors believe that IVUS should be used
in decision making for the treatment of stent
restenosis. If IVUS identifies stent underexpan-
sion as the cause of restenosis site, balloon
dilatations based on IVUS stent area measure-
ments should be the mode of treatment. If neo-
intimal proliferation is found to be the cause, then
restenting should be considered. The exact local-
ization of the restenotic site is also important
when deciding among treatment options. If reste-
nosis is within the stent, then higher-pressure
inflations can be effective. Conversely, this ap-
proach may not be suitable for treating ‘‘edge
restenosis’’ that occurs in the unstented nearby
reference sites.
Fig. 6. Left panel shows a regular gray-scale IVUS image of an atherosclerotic plaque. Right panel shows tissue char-
acterization map obtained through spectral analysis of the same cross-section.
SIPAHI et al
In the DES era, the binary restenosis rate is
low and, therefore, in clinical studies large num-
bers of patients are needed to have enough power
to detect a difference in the efficacy of different
DES. In-stent late lumen loss, as determined by
IVUS, has been used by some investigators to
compare the efficacy of various strategies includ-
ing different DES or the long-term results in
different patient populations [51]. It is currently
unknown whether a larger late loss as determined
by IVUS is associated with significantly higher
clinical event rates.
Guidelines
The latest recommendations of the American
College of Cardiology and the American Heart
Association on indications for coronary IVUS
are published in the 2001 guidelines for PCI [52].
Box 1 summarizes these recommendations. The
results of recent research presented in this review
and the declining use of most of the interventional
devices other than DES are likely to change some
of these recommendations.
Identification of vulnerable plaques: the future
of intravascular ultrasound imaging?
Currently, decisions regarding the percutane-
ous treatment of coronary lesions are mostly
driven by the degree of luminal compromise as
assessed by the angiogram or by complementary
studies such as pressure measurements or IVUS.
 INTRAVASCULAR ULTRASOUND IN CURRENT PCI ERA
Accurate identification of plaques that do not
cause significant luminal compromise but are
vulnerable to future rupture can revolutionize
not only the percutaneous but also the pharma-
cologic treatment of coronary artery disease.
Although regular gray-scale IVUS can help to
estimate various histologic components of pla-
ques, particularly calcifications, it cannot reliably
differentiate vulnerable plaques from stable pla-
ques. Accordingly, various techniques based on
IVUS imaging (elastography, palpography, spec-
tral analysis) or other imaging modalities (optical
coherence tomography, thermography, intracoro-
nary magnetic resonance) have been developed for
detection of vulnerable plaques [53]. Of the IVUS-
based modalities, IVUS elastography evaluates
the local elastic properties (strain) of plaques
and has been shown to detect histologically vul-
nerable plaques with high sensitivity and specific-
ity [54]. IVUS palpography is an allied technology
that also evaluates tissue strain but is faster and
more robust than IVUS elastography [55]. Spec-
tral analysis of radiofrequency signals is another
promising new IVUS-based imaging modality.
This technique has been shown to classify various
components of plaques with high accuracy (Fig. 6)
[56,57] and has been used successfully to evaluate
changes in these components with lipid-lowering
therapy [58]. Ongoing research including the Pro-
viding Regional Observations to Study Predictors
of Events in the Coronary Tree (PROSPECT)
study will help define the value of radiofrequency
spectral analysis and other novel coronary imag-
ing modalities in determining or improving clini-
cal outcomes.
References
[1] St. Goar FG, Pinto FJ, Alderman EL, et al. Intra-
vascular ultrasound imaging of angiographically
normal coronary arteries: an in vivo comparison
with quantitative angiography. J Am Coll Cardiol
1991;18(4):952–8.
[2] Fitzgerald PJ, St. Goar FG, Connolly AJ, et al. In-
travascular ultrasound imaging of coronary arteries.
Is three layers the norm? Circulation 1992;86(1):
154–8.
[3] Mintz GS, Nissen SE, Anderson WD, et al. Ameri-
can College of Cardiology Clinical Expert Consen-
sus Document on Standards for Acquisition,
Measurement and Reporting of Intravascular Ultra-
sound Studies (IVUS). A report of the American
College of Cardiology Task Force on Clinical Expert
Consensus Documents. J Am Coll Cardiol 2001;
37(5):1478–92.
171
[4] Di Mario C, Gorge G, Peters R, et al. Clinical appli-
cation and image interpretation in intracoronary ul-
trasound. Study group on Intracoronary Imaging of
the Working Group of Coronary Circulation and of
the Subgroup on Intravascular Ultrasound of the
Working Group of Echocardiography of the Euro-
pean Society of Cardiology. Eur Heart J 1998;
19(2):207–29.
[5] Pijls NH, De Bruyne B, Peels K, et al. Measurement
of fractional flow reserve to assess the functional se-
verity of coronary-artery stenoses. N Engl J Med
1996;334(26):1703–8.
[6] Miller DD, Donohue TJ, Younis LT, et al. Correla-
tion of pharmacological 99mTc-sestamibi myocar-
dial perfusion imaging with poststenotic coronary
flow reserve in patients with angiographically inter-
mediate coronary artery stenoses. Circulation 1994;
89(5):2150–60.
[7] Bech GJ, De Bruyne B, Pijls NH, et al. Fractional
flow reserve to determine the appropriateness of an-
gioplasty in moderate coronary stenosis: a random-
ized trial. Circulation 2001;103(24):2928–34.
[8] Takagi A, Tsurumi Y, Ishii Y, et al. Clinical poten-
tial of intravascular ultrasound for physiological as-
sessment of coronary stenosis: relationship between
quantitative ultrasound tomography and pressure-
derived fractional flow reserve. Circulation 1999;
100(3):250–5.
[9] Briguori C, Anzuini A, Airoldi F, et al. Intravascular
ultrasound criteria for the assessment of the func-
tional significance of intermediate coronary artery
stenoses and comparison with fractional flow re-
serve. Am J Cardiol 2001;87(2):136–41.
[10] Abizaid AS, Mintz GS, Mehran R, et al. Long-term
follow-up after percutaneous transluminal coronary
angioplasty was not performed based on intravascu-
lar ultrasound findings: importance of lumen dimen-
sions. Circulation 1999;100(3):256–61.
[11] Mintz GS, Pichard AD, Kovach JA, et al. Impact
of preintervention intravascular ultrasound imag-
ing on transcatheter treatment strategies in coro-
nary artery disease. Am J Cardiol 1994;73(7):
423–30.
[12] Isner JM, Kishel J, Kent KM, et al. Accuracy of an-
giographic determination of left main coronary arte-
rial narrowing. Angiographic-histologic correlative
analysis in 28 patients. Circulation 1981;63(5):
1056–64.
[13] Cameron A, Kemp HG Jr, Fisher LD, et al. Left
main coronary artery stenosis: angiographic deter-
mination. Circulation 1983;68(3):484–9.
[14] Prospective randomised study of coronary artery
bypass surgery in stable angina pectoris. Second
interim report by the European Coronary Surgery
Study Group. Lancet 1980;2(8193):491–5.
[15] Fisher LD, Judkins MP, Lesperance J, et al. Repro-
ducibility of coronary arteriographic reading in the
coronary artery surgery study (CASS). Cathet Car-
diovasc Diagn 1982;8(6):565–75.
172 SIPAHI
[16] Jasti V, Ivan E, Yalamanchili V, et al. Correlations
between fractional flow reserve and intravascular ul-
trasound in patients with an ambiguous left main
coronary artery stenosis. Circulation 2004;110(18):
2831–6.
[17] Abizaid AS, Mintz GS, Abizaid A, et al. One-year
follow-up after intravascular ultrasound assessment
of moderate left main coronary artery disease in pa-
tients with ambiguous angiograms. J Am Coll Cardi-
ol 1999;34(3):707–15.
[18] Fassa AA, Wagatsuma K, Higano ST, et al. Intra-
vascular ultrasound-guided treatment for angio-
graphically indeterminate left main coronary artery
disease: a long-term follow-up study. J Am Coll Car-
diol 2005;45(2):204–11.
[19] Nakamura S, Colombo A, Gaglione A, et al. Intra-
coronary ultrasound observations during stent im-
plantation. Circulation 1994;89(5):2026–34.
[20] Kiemeneij F, Laarman G, Slagboom T. Mode of de-
ployment of coronary Palmaz-Schatz stents after
implantation with the stent delivery system: an
intravascular ultrasound study. Am Heart J 1995;
129(4):638–44.
[21] Colombo A, Hall P, Nakamura S, et al. Intracoro-
nary stenting without anticoagulation accomplished
with intravascular ultrasound guidance. Circulation
1995;91(6):1676–88.
[22] Albiero R, Rau T, Schluter M, et al. Comparison of
immediate and intermediate-term results of intravas-
cular ultrasound versus angiography-guided Pal-
maz-Schatz stent implantation in matched lesions.
Circulation 1997;96(9):2997–3005.
[23] Blasini R, Neumann FJ, Schmitt C, et al. Restenosis
rate after intravascular ultrasound-guided coronary
stent implantation. Cathet Cardiovasc Diagn 1998;
44(4):380–6.
[24] Fitzgerald PJ, Oshima A, Hayase M, et al. Final re-
sults of the Can Routine Ultrasound Influence Stent
Expansion (CRUISE) study. Circulation 2000;
102(5):523–30.
[25] Choi JW, Goodreau LM, Davidson CJ. Resource
utilization and clinical outcomes of coronary stent-
ing: a comparison of intravascular ultrasound and
angiographical guided stent implantation. Am Heart
J 2001;142(1):112–8.
[26] Orford JL, Denktas AE, Williams BA, et al. Routine
intravascular ultrasound scanning guidance of coro-
nary stenting is not associated with improved clinical
outcomes. Am Heart J 2004;148(3):501–6.
[27] Schiele F, Meneveau N, Vuillemenot A, et al. Im-
pact of intravascular ultrasound guidance in stent
deployment on 6-month restenosis rate: a multicen-
ter, randomized study comparing two strategiesd
with and without intravascular ultrasound guid-
ance. RESIST Study Group. REStenosis after Ivus
guided STenting. J Am Coll Cardiol 1998;32(2):
320–8.
[28] Mudra H, di Mario C, de Jaegere P, et al. Random-
ized comparison of coronary stent implantation
et al
under ultrasound or angiographic guidance to re-
duce stent restenosis (OPTICUS Study). Circulation
2001;104(12):1343–9.
[29] Oemrawsingh PV, Mintz GS, Schalij MJ, et al. Intra-
vascular ultrasound guidance improves angio-
graphic and clinical outcome of stent implantation
for long coronary artery stenoses: final results of
a randomized comparison with angiographic guid-
ance (TULIP Study). Circulation 2003;107(1):62–7.
[30] Casella G, Klauss V, Ottani F, et al. Impact of intra-
vascular ultrasound-guided stenting on long-term
clinical outcome: a meta-analysis of available studies
comparing intravascular ultrasound-guided and
angiographically guided stenting. Catheter Cardio-
vasc Interv 2003;59(3):314–21.
[31] Hoffmann R, Mintz GS, Mehran R, et al. Intravas-
cular ultrasound predictors of angiographic resteno-
sis in lesions treated with Palmaz-Schatz stents.
J Am Coll Cardiol 1998;31(1):43–9.
[32] de Jaegere P, Mudra H, Figulla H, et al. Intravascu-
lar ultrasound-guided optimized stent deployment.
Immediate and 6 months clinical and angiographic
results from the Multicenter Ultrasound Stenting
in Coronaries Study (MUSIC Study). Eur Heart J
1998;19(8):1214–23.
[33] Moussa I, Moses J, Di Mario C, et al. Does the spe-
cific intravascular ultrasound criterion used to opti-
mize stent expansion have an impact on the
probability of stent restenosis? Am J Cardiol 1999;
83(7):1012–7.
[34] Moses JW, Leon MB, Popma JJ, et al. Sirolimus-
eluting stents versus standard stents in patients
with stenosis in a native coronary artery. N Engl J
Med 2003;349(14):1315–23.
[35] Schofer J, Schluter M, Gershlick AH, et al. Siroli-
mus-eluting stents for treatment of patients with
long atherosclerotic lesions in small coronary ar-
teries: double-blind, randomised controlled trial
(E-SIRIUS). Lancet 2003;362(9390):1093–9.
[36] Sonoda S, Morino Y, Ako J, et al. Impact of final
stent dimensions on long-term results following siro-
limus-eluting stent implantation: serial intravascular
ultrasound analysis from the SIRIUS trial. J Am
Coll Cardiol 2004;43(11):1959–63.
[37] Fujii K, Mintz GS, Kobayashi Y, et al. Contribution
of stent underexpansion to recurrence after siroli-
mus-eluting stent implantation for in-stent resteno-
sis. Circulation 2004;109(9):1085–8.
[38] Fujii K, Carlier SG, Mintz GS, et al. Stent underex-
pansion and residual reference segment stenosis are
related to stent thrombosis after sirolimus-eluting
stent implantation: an intravascular ultrasound
study. J Am Coll Cardiol 2005;45(7):995–8.
[39] Stone GW, Hodgson JM, St. Goar FG, et al. Im-
proved procedural results of coronary angioplasty
with intravascular ultrasound-guided balloon sizing:
the CLOUT Pilot Trial. Clinical Outcomes with Ul-
trasound Trial (CLOUT) Investigators. Circulation
1997;95(8):2044–52.
 INTRAVASCULAR ULTRASOUND IN CURRENT PCI ERA
[40] Regar E, Lemos PA, Saia F, et al. Incidence of
thrombotic stent occlusion during the first three
months after sirolimus-eluting stent implantation
in 500 consecutive patients. Am J Cardiol 2004;
93(10):1271–5.
[41] Schuhlen H, Hadamitzky M, Walter H, et al. Major
benefit from antiplatelet therapy for patients at high
risk for adverse cardiac events after coronary
Palmaz-Schatz stent placement: analysis of a pro-
spective risk stratification protocol in the Intracoro-
nary Stenting and Antithrombotic Regimen (ISAR)
trial. Circulation 1997;95(8):2015–21.
[42] Ziada KM, Tuzcu EM, De Franco AC, et al. Intra-
vascular ultrasound assessment of the prevalence
and causes of angiographic ‘‘haziness’’ following
high-pressure coronary stenting. Am J Cardiol
1997;80(2):116–21.
[43] Grewal J, Ganz P, Selwyn A, et al. Usefulness of
intravascular ultrasound in preventing stenting of
hazy areas adjacent to coronary stents and its sup-
port of support spot-stenting. Am J Cardiol 2001;
87(11):1246–9.
[44] Painter JA, Mintz GS, Wong SC, et al. Serial intra-
vascular ultrasound studies fail to show evidence of
chronic Palmaz-Schatz stent recoil. Am J Cardiol
1995;75(5):398–400.
[45] Hoffmann R, Mintz GS, Dussaillant GR, et al. Pat-
terns and mechanisms of in-stent restenosis. A serial
intravascular ultrasound study. Circulation 1996;
94(6):1247–54.
[46] Lemos PA, Saia F, Ligthart JM, et al. Coronary
restenosis after sirolimus-eluting stent implantation:
morphological description and mechanistic analysis
from a consecutive series of cases. Circulation 2003;
108(3):257–60.
[47] Castagna MT, Mintz GS, Leiboff BO, et al. The con-
tribution of ‘‘mechanical’’ problems to in-stent reste-
nosis: an intravascular ultrasonographic analysis of
1090 consecutive in-stent restenosis lesions. Am
Heart J 2001;142(6):970–4.
[48] Sharma SK, Kini A, Mehran R, et al. Randomized
trial of Rotational Atherectomy Versus Balloon An-
gioplasty for Diffuse In-stent Restenosis (ROS-
TER). Am Heart J 2004;147(1):16–22.
[49] Takebayashi H, Kobayashi Y, Mintz GS, et al. In-
travascular ultrasound assessment of lesions with
173
target vessel failure after sirolimus-eluting stent im-
plantation. Am J Cardiol 2005;95(4):498–502.
[50] Cheneau E, Pichard AD, Satler LF, et al. Intravas-
cular ultrasound stent area of sirolimus-eluting
stents and its impact on late outcome. Am J Cardiol
2005;95(10):1240–2.
[51] Abizaid A, Costa MA, Blanchard D, et al. Siroli-
mus-eluting stents inhibit neointimal hyperplasia in
diabetic patients. Insights from the RAVEL Trial.
Eur Heart J 2004;25(2):107–12.
[52] Smith SC Jr, Dove JT, Jacobs AK, et al. ACC/AHA
guidelines for percutaneous coronary intervention
(revision of the 1993 PTCA guidelines)dexecutive
summary: a report of the American College of Car-
diology/American Heart Association task force on
practice guidelines (Committee to Revise the 1993
Guidelines for Percutaneous Transluminal Coro-
nary Angioplasty) endorsed by the Society for Car-
diac Angiography and Interventions. Circulation
2001;103(24):3019–41.
[53] Tuzcu EM, Schoenhagen P. Acute coronary syn-
dromes, plaque vulnerability, and carotid artery dis-
ease: the changing role of atherosclerosis imaging.
J Am Coll Cardiol 2003;42(6):1033–6.
[54] Schaar JA, De Korte CL, Mastik F, et al. Character-
izing vulnerable plaque features with intravascular
elastography. Circulation 2003;108(21):2636–41.
[55] Schaar JA, Regar E, Mastik F, et al. Incidence
of high-strain patterns in human coronary arteries:
assessment with three-dimensional intravascular
palpography and correlation with clinical presenta-
tion. Circulation 2004;109(22):2716–9.
[56] Nair A, Kuban BD, Tuzcu EM, et al. Coronary pla-
que classification with intravascular ultrasound
radiofrequency data analysis. Circulation 2002;
106(17):2200–6.
[57] Murashige A, Hiro T, Fujii T, et al. Detection of lipid-
laden atherosclerotic plaque by wavelet analysis of
radiofrequency intravascular ultrasound signals: in
vitro validation and preliminary in vivo application.
J Am Coll Cardiol 2005;45(12):1954–60.
[58] Kawasaki M, Sano K, Okubo M, et al. Volumetric
quantitative analysis of tissue characteristics of cor-
onary plaques after statin therapy using three-
dimensional integrated backscatter intravascular
ultrasound. J Am Coll Cardiol 2005;45(12):1946–53.
 Cardiol Clin 24 (2006) 175–199

Antithrombotic Therapy for Percutaneous

Coronary Intervention
Nitin Barman, MDa, Deepak L. Bhatt, MDb,*
a
Interventional Cardiology, Mount Sinai Hospital, One Gustave L. Levy Place, New York, NY 10029, USA
b
Cardiac, Peripheral, and Carotid Intervention, Department of Cardiovascular Medicine,
Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk F25, Cleveland, OH 44195, USA

The use of percutaneous coronary intervention
(PCI) in the treatment of obstructive coronary
artery disease has expanded rapidly in the past
decade. Despite extensive technologic advance-
ments in the field, pharmacotherapy has remained
a cornerstone in the overall treatment strategy.
Optimizing the ischemic complications mediated
through a complex coagulation cascade, with the
ever-present risk of bleeding from antiplatelet and
anticoagulant therapy, has a remained a challenge
for drug developers and for clinicians. In this
article, the authors discuss the evolution, current
treatment, and future landscape of pharmacother-
apy in PCI.
Platelet biology
Comprehension of platelet biology has pro-
vided the basis for the current standard of
antiplatelet therapy in PCI. Circulating platelets
traverse the vasculature in an inactivated state
until they are exposed to collagen fibrils in the
connective tissue matrix underlying normal endo-
thelial cells. Exposure to this matrix, which occurs
universally after vessel injury induced by PCI,
results in a complex series of events leading to
platelet activation and aggregation. Activated
platelets change their configuration and secrete,
among other substances, thromboxane A2 and
ADP into the local environment, resulting in am-
plification of circulating platelet activation. Paral-
lel pathways of platelet activation exist, including
* Corresponding author.
E-mail address: bhattd@ccf.org (D.L. Bhatt).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.01.005
the interaction of tissue factor (TF), which is ex-
pressed on all nonvascular cells, and factor VIIa.
This interaction results in the generation of
thrombin, the most potent of platelet activators,
and local coagulation. The final pathway of aggre-
gation of platelets, which leads to the platelet
thrombus, is mediated through the glycoprotein
IIb/IIIa receptor. Activated platelets allow for
conformational changes and upregulation of sur-
face expression of the IIb/IIIa receptor. These
conformation changes allow for fibrinogen bind-
ing and cross-linking of platelets, with the ulti-
mate result being growth and stability of the
hemostatic plug (Fig. 1). Because initial platelet
thrombus formation during PCI plays a major
role in acute and subacute periprocedural ische-
mic complications, it is no surprise that this path-
way has been a major target for pharmacotherapy
in PCI.
Antiplatelet agents
Aspirin
The use of aspirin is ubiquitous in patients
undergoing coronary intervention. By acetylating
the cyclooxygenase-1 enzyme, aspirin inhibits the
synthesis of thromboxane A2, resulting in irrevers-
ible inhibition of platelet function. Daily adminis-
tration of 30 to 50 mg of aspirin results in virtually
complete suppression of thromboxane A2 synthe-
sis by 7 to 10 days [1,2]. Justification for the use of
aspirin in PCI is based primarily on a number of
early trials that compared aspirin and dipyrida-
mole versus placebo before percutaneous translu-
minal coronary angiogplasty (PTCA). These
studies demonstrated a statistically significant
cardiology.theclinics.com
176 BARMAN & BHATT

Fig. 1. Schematic of platelet activation involving (1) a shape change in which the platelet membrane surface area is
greatly increased; (2) the secretion of proinflammatory, prothrombotic, adhesive, and chemotactic mediators that prop-
agate, amplify, and sustain the atherothrombotic process; and (3) the activation of the glycoprotein (GP) IIb/IIIa recep-
tor from its inactive form. Multiple agonists including thromboxane A2 (TXA2), ADP, thrombin, serotonin, epinephrine,
and collagen can activate the platelet and, thus, contribute toward establishing the environmental conditions necessary
for atherothrombosis to occur. COX1, cyclooxygenase-1; PAI, plasminogen activator inhibitor; PDGF, platelet-derived
growth factor; vWF, von Willebrand factor. (From Mehta SR, Yusuf S. Short- and long-term oral antiplatelet therapy in
acute coronary syndromes and percutaneous coronary intervention. J Am Coll Cardiol 2003;41(4):S79–88; with
permission.)

reduction in periprocedural ischemic complica- resistance’’ [9]. Furthermore, the biochemical

tions including abrupt vessel closure, Q wave
myocardial infarction (MI), or need for emergent
coronary artery bypass grafting (5% versus 14%)
[3,4]. Further study demonstrated that these ben-
efits were present irrespective of the use of dipyr-
imadole, which has subsequently fallen out of
favor in the treatment of patients undergoing
PCI [5]. With respect to dose, the initial trials of as-
pirin all used high doses of aspirin (O650 mg/d),
although subsequent study has suggested an ab-
sence of additional efficacy beyond 81 mg/d used
chronically [6,7]. Due to the higher dose of aspirin
used initially after PCI in more recent randomized
studies, current recommendations are for 325 mg
daily for 1 month following bare metal stent or
for 3 to 6 months following drug-eluting stent im-
plantation [8]. Despite the clear benefits of aspi-
rin in the setting of PCI, a significant number
of patients still suffer thrombotic complications.
Among other explanations, recent investigations
have demonstrated significant variability among
the population with respect to antiplatelet re-
sponse to aspirin therapy, or so-called ‘‘aspirin
finding of aspirin resistance has been demon-
strated, prospectively, to result in up to a three-
fold increase in the risk of major adverse
outcome in patients who have established car-
diovascular disease [10]. Although the complete
details are not discussed in depth in this article,
aspirin resistance exists in large part as a result
of genetic heterogeneity and as a manifestation
of the redundancy of the platelet thrombus
pathway. It is this redundancy that is the target
of the other therapies discussed in the following
section.
Platelet ADP receptor antagonists
Thienopyridine antiplatelet agents including
ticlopidine and clopidogrel are prodrugs that,
when metabolized, achieve their antiplatelet ef-
fects by inhibiting the binding of ADP to its G
protein–coupled P2Y12 purinergic receptor.
When active, the P2Y12 receptor inhibits platelet
disaggregation, an effect mediated by intracellular
cyclic AMP [11]. Inhibition of this receptor, with
 ANTITHROMBOTIC THERAPY FOR PCI
the resultant impairment in platelet aggregation,
provides the basis for the development and study
of the current and future antiplatelet agents dis-
cussed in this section.
The use of thienopyridine platelet P2Y12 re-
ceptor antagonists paralleled the introduction of
intracoronary stents to the interventional cardiol-
ogy world. Routine stenting dramatically de-
creased the risk of abrupt or threatened vessel
closure and restenosis [12]; however, it also led to
concern over possible subacute thrombotic vessel
occlusion due to the inherent thrombogenicity of
the implanted intracoronary stent. These concerns
resulted in several investigations regarding the op-
timal anticoagulation regimen after unplanned or
elective stenting. In the STARS trial (see Appen-
dix 1 for a glossary of clinical trial acronyms
that appear in this article), 1653 patients were ran-
domized to receive aspirin monotherapy, ticlopi-
dine plus aspirin, or warfarin plus aspirin after
intracoronary stenting. In this study, it was found
that treatment with aspirin plus ticlopidine com-
pared with aspirin alone or aspirin plus warfarin
resulted in a lower rate of stent thrombosis [13].
Similar results were seen in the FANTASTIC
study comparing ticlopidine plus aspirin versus
warfarin plus aspirin after routine stenting. In ad-
dition to improved efficacy with the antiplatelet
strategy, the additional finding of reduced bleed-
ing was observed [14]. Multiple additional ran-
domized trials in varying patient populations
yielded consistent results, thus solidifying the
role of dual antiplatelet therapy after PCI [15–17].
Despite the clear efficacy afforded by ticlopi-
dine in addition to aspirin or in place of full
anticoagulation after intracoronary stenting, ti-
clopidine is associated with a significant number
of adverse effects that have ultimately limited its
use. The most common serious side effect attrib-
uted to ticlopidine is severe drug-induced neutro-
penia. Occurring in 1% to 2% of patients, usually
within the first 3 months of therapy, this poten-
tially fatal side effect mandates immediate cessa-
tion of the drug [18]. A less common but
potentially more serious adverse effect of ticlopi-
dine is the development of thrombotic thrombo-
cytopenic purpura–hemolytic uremic syndrome
(TTP-HUS). Most often occurring 3 to 12 weeks
after initiation of the drug, the frequency of this
complication ranges from 1 in 1600 to 1 in 4800
patients treated. Like severe neutropenia, develop-
ment of TTP-HUS necessitates immediate discon-
tinuation of the drug and requires additional
therapy with timely plasmapheresis [19,20]. As
177
a result of these and other less specific adverse ef-
fects, ticlopidine has essentially been replaced by
clopidogrel for the prevention of subacute stent
thrombosis. The basis for this change (ie, the dra-
matically reduced incidence of side effects) was
demonstrated in the CLASSICS trial that com-
pared clopidogrel versus ticlopidine in addition
to aspirin after PCI. Clopidogrel was associated
with a 50% reduction in the incidence of the com-
bined end point of major bleeding, thrombocyto-
penia, neutropenia, or discontinuation for
noncardiac adverse events (4.6% versus 9.1%)
[21].
In addition to a more favorable safety profile,
clopidogrel may be more efficacious than ticlopi-
dine. A large meta-analysis of trials and registries
comparing clopidogrel and ticlopidine demon-
strated reduced major adverse cardiac events
with the use of clopidogrel (Fig. 2) [22]. One po-
tential mechanism for differential efficacy is re-
lated to the onset of action of platelet inhibition,
which is significantly slower with ticlopidine
than with clopidogrel [23]. Platelet activation
and distal embolization are major factors respon-
sible for PCI-related ischemic complications, and
because these events occur with initial vessel in-
strumentation, the issues of pretreatment and op-
timal loading dose of oral antiplatelet therapy
have become particularly relevant.
The PCI subset of the CURE trial gave the
initial indication that pretreatment with a 300-mg
dose of clopidogrel was clinically useful. In the
CURE trial, 12,562 patients who had acute coro-
nary syndromes (ACS) were given aspirin plus
loading and maintenance-therapy clopidogrel or
placebo [24]. In the 2658 patients undergoing
PCI who received the loading dose a median of
10 days before the procedure, the primary end
point of cardiovascular death, MI, or urgent re-
vascularization was significantly reduced in the
clopidogrel pretreatment group (4.5% versus
6.4%), with no differences in major bleeding
[25]. In the CREDO trial, 2100 patients were ran-
domized to a loading dose of 300 mg of clopidog-
rel or placebo 3 to 24 hours before elective PCI.
All subjects received 28 days of clopidogrel fol-
lowed by 11 months of therapy dictated by the
original randomization, with 40% receiving glyco-
protein IIb/IIIa inhibition. In the pretreatment
group, particularly in those who received clopi-
dogrel more than 6 hours before PCI, there was
a nearly significant reduction in 28-day death,
MI, or urgent revascularization. At 1 year, there
was a significant reduction in the combined end
178 BARMAN & BHATT

Fig. 2. Odds ratio plots with 95% confidence intervals (CIs) for the rates of death, MI, target vessel revascularization
(TVR), subacute stent thrombosis (SAT), and major adverse cardiac events (MACE) using clopidogrel compared with
ticlopidine after intracoronary stenting. (From Bhatt DL, Bertrand ME, Berger PB, et al. Meta-analysis of randomized
and registry comparisons of ticlopidine with clopidogrel after stenting. J Am Coll Cardiol 2002;39(1):9–14; with
permission.)

point of death, MI, or stroke in the loading and
long-term clopidogrel treatment group (8.5% ver-
sus 11.5%), without a significant increase in major
bleeding (Table 1). Subgroup analyses demon-
strated that longer intervals between the 300-mg
loading dose of clopidogrel and PCI may reduce
events [26].
Universal pretreatment of patients with clopi-
dogrel, however, has met with two major practical
limitations. Specifically, in the CURE trial, the
patients randomized to clopidogrel who under-
went coronary bypass grafting within 5 days of
receiving the drug had a 53% increased risk for
major bleeding compared with similar patients
receiving placebo, presumably due to prolonged
platelet inhibition [24]. This finding has led to re-
luctance to pretreat patients undergoing coronary
angiography for fear of delaying revascularization
in patients whose anatomy is more suitable for
coronary artery bypass grafting. These concerns,
however, may become less relevant because de-
creasing numbers of patients are being referred
for surgical revascularization in the era of drug-
eluting stents. The second issue limiting clopidogrel
pretreatment relates to catheterization laboratory
infrastructure and the feasibility of dosing

Table 1
One-year clinical outcomes for clopidogrel versus placebo from The Clopidogrel for The Reduction of Events During
Observation trial
Number of patients (%)
End point Clopidogrel (n ¼ 1053) Placebo (n ¼ 1063) RRR (95% CI)
Death, MI, stroke 89 (8.5) 122 (11.5) 26.9 (3.9–44.4)
Death, MI 84 (7.9) 111 (10.4) 24.0 (
0.9–42.7)
Death 18 (1.7) 24 (2.3) 24.6 (
38.9–59.1)
MI 70 (6.7) 89 (8.4) 20.8 (
8.4–42.1)
Stroke 9 (0.9) 10 (0.9) 10.0 (
21.3–24.0)
Revascularization
Any TVR 138 (13.1) 144 (13.6) 4.0 (
21.3–24.0)
Urgent TVR 21 (2.0) 23 (2.2) 8.1 (
66.1–49.1)
Any revascularization 224 (21.3) 223 (21.0)
1.1 (
21.7–16.0)
Abbreviations: CI, confidence interval; RRR, relative risk reduction; TVR, target vessel revascularization.
From Steinhubl SR, Tan WA, Foody JM, et al. Early and sustained dual oral antiplatelet therapy following percu-
taneous coronary intervention: a randomized controlled trial. JAMA 2002;288(19):2411–20; with permission.
 ANTITHROMBOTIC THERAPY FOR PCI
patients several hours to days before their proce-
dure. This issue is more directly addressed by nu-
merous studies evaluating the pharmacodynamics
of different loading doses of clopidogrel in human
subjects. Specifically, it is clear that a higher load-
ing dose of clopidogrel leads to more rapid and
long-lasting inhibition of ADP-induced ex vivo
platelet aggregation [27]. Further insight was
gained from the ISAR-REACT trial in which
low-risk elective PCI patients received a 600-mg
loading dose of clopidogrel before PCI and were
randomized to PCI with or without abciximab,
with no significant difference in ischemic out-
comes. In a prespecified subanalysis of the timing
of pretreatment, there was no difference in the
combined ischemic end point of death, MI, or ur-
gent revascularization for abciximab versus pla-
cebo among patients receiving clopidogrel 2 to 3
hours, 3 to 6 hours, 6 to 12 hours, or more than
12 hours before PCI (Fig. 3) [28]. More recently,
300-mg versus 600-mg loading doses of clopi-
dogrel were directly compared prospectively in
the ARMYDA-2 trial. Patients who had stable or
unstable angina scheduled for PCI were random-
ized to one of these two loading doses 4 to 8 hours
before their procedure. The primary end point of
death, MI, or urgent revascularization occurred
significantly less often in the group receiving
a 600-mg loading dose (4% versus 12%), al-
though this difference was solely accounted for
by a reduction in periprocedural MI [29]. The
concept of a higher loading dose was advanced
further by the recently reported ALBION trial
Fig. 3. Kaplan-Meier event curves for 30-day occur-
rence of death, MI, or urgent revascularization relative
to clopidogrel loading-dose interval from the ISAR-RE-
ACT study. (From Kandzari DE, Berger PB, Kastrati A,
et al. Influence of treatment duration with a 600-mg dose
of clopidogrel before percutaneous coronary revascular-
ization. J Am Coll Cardiol 2004;44(11):2133–6; with
permission.)
179
evaluating 103 patients who had ACS and re-
ceived a 300-mg, 600-mg, or 900-mg loading
dose of clopidogrel on presentation. Compared
with the 300-mg loading dose, the 600-mg dose
of clopidogrel demonstrated a more rapid onset
of action and higher level of inhibition while
maintaining a similar safety profile; 900 mg ap-
peared to provide slightly higher antiplatelet in-
hibition than the 600 mg loading dose [30]. The
ISAR-CHOICE study demonstrated additional
platelet inhibition with a 600-mg loading dose
versus a 300-mg loading dose, although 900
mg did not provide any further antiplatelet ef-
fect compared with the 600-mg loading dose in
this study [31]. Further investigation of higher
loading doses in the setting of PCI is ongoing.
Despite the almost universal use of dual
antiplatelet agents in PCI, periprocedural ische-
mic complications still occur, and a small per-
centage (0.4%–1.1%) of patients suffer subacute
stent thrombosis [32]. As is the case with aspirin,
variability in response to clopidogrel is an increas-
ingly recognized entity that may account for some
of these events. Laboratory measurements dem-
onstrate that up to 30% of patients may have an
inadequate antiplatelet response to standard dos-
ing of clopidogrel, a finding that correlates with
cytochrome P450 3A4 metabolic activity, al-
though this has not been correlated with clinical
events per se [33]. As such, the search for more po-
tent antiplatelet agents continues. One such agent
is prasugrel, a novel oral thienopyridine that has
more predictable and potent antiplatelet effects
[34]. The JUMBO–TIMI 26 trial, a phase II
dose-finding study of prasugrel versus clopidogrel
in patients undergoing elective or urgent PCI, sug-
gested an equivalent bleeding risk to clopidogrel,
with a trend toward more efficacy at 30 days in
the prasugrel-treated patients [35]. TRITON–
TIMI 38, a large-scale randomized trial for effi-
cacy of prasugrel versus clopidogrel in patients
who have ACS undergoing PCI, is ongoing. Be-
cause thienopyridines are prodrugs that require
metabolism to produce active metabolites, there
is an inherent delay in their effect [11]. As such,
development and investigation of nonthienopyri-
dine direct P2Y12 receptor antagonists are also
underway. AZD6140, a novel oral direct P2Y12
receptor antagonist, appears to be a more potent,
consistent, and immediate platelet antagonist [36].
The DISPERSE-2 phase II trial evaluated this
agent in patients who had ACS and found a simi-
lar rate of bleeding and similar ischemic events to
clopidogrel, although there was a higher rate of
180
dyspnea with AZD6140 [37]. Cangrelor, an analog
of the endogenous direct platelet ADP receptor
inhibitor ATP, is an intravenous drug demon-
strating more immediate antiplatelet effects. Ini-
tial studies suggest consistent, rapid, and potent
platelet inhibition and an acceptable safety profile
[36]. Larger clinical trials of this intravenous agent
are planned.
Although CREDO and PCI-CURE support
the use of clopidogrel in addition to aspirin for
a year after PCI, longer-term therapy may yield
even greater benefit. This issue is being addressed
as part of the ongoing CHARISMA trial [38]. The
issue of prolonged dual antiplatelet therapy may
be particularly relevant with drug-eluting stents,
which seem to have the potential for delayed stent
thrombosis compared with bare metal stents when
antiplatelet therapy is interrupted (such as for sur-
gery) or discontinued.
Platelet glycoprotein IIb/IIIa receptor antagonists
Although platelets can be activated by a num-
ber of different pharmacologic and mechanical
stimuli, their ability to aggregate and adhere to
disrupted endothelium is largely mediated by the
surface glycoprotein IIb/IIIa receptor, a member
of the integrin family of membrane receptors [39].
With activation, the surface expression and con-
formation of the glycoprotein IIb/IIIa receptor
changes, rendering it competent to bind a number
of different ligands. Fibrinogen, by far the domi-
nant ligand, is bound by the activated glycopro-
tein IIb/IIIa receptor at opposite ends, allowing
a cross-link to form between two adjacent plate-
lets and the endothelium [40]. This common path-
way for platelet aggregation and adherence has
served as a potentially high-yielding target in the
development of antiplatelet agents for PCI.
Table 2
Dosing and pharmacokinetic comparison of clinically available intravenous glycoprotein IIb-IIIa inhibitors
Feature Abciximab
Mechanism Chimeric antibody
Dose: loading 0.25 mg/kg  1
Dose: infusion 0.125 m/kg/min  12 h
Plasma half-life 30 min
Dose reduction in CRI None
Reversible with Yes
platelet infusion
Removed by hemodialysis No
Abbreviations: CrCl, creatinine clearance; CRI, chronic renal insufficiency.
a
FDA, approved dose for ACS.
BARMAN & BHATT
Currently, there are three intravenous glycopro-
tein IIb/IIIa inhibitors available clinically: abcixi-
mab, tirofiban, and eptifibatide. Features of the
different intravenous glycoprotein IIb/IIIa recep-
tor antagonists are summarized in Table 2. All
of these agents have been studied extensively in
large clinical trials, the results of which are dis-
cussed in the following sections.
Abciximab
Originally produced in 1985 and subsequently
approved by the Food and Drug Administration
(FDA) in 1994, abciximab is the Fab antibody
fragment of a chimeric human-murine monoclonal
antibody specific for the glycoprotein IIb/IIIa
receptor. Joined with the constant regions of
human immunoglobulin, this chimer allows for
preserved specificity with minimal antigenicity [41].
Initial animal and human pilot studies verified the
potent antiplatelet effect of abciximab with up to
93% platelet inhibition at 2 hours in patients re-
ceiving the highest dose [42,43]. Clinical efficacy
was demonstrated on completion of the EPIC trial,
the original large experience with abciximab in the
catheterization laboratory. In this study, 2099
high-risk patients undergoing PTCA or atherec-
tomy were randomized to abciximab or placebo
in addition to receiving aspirin and heparin. Pa-
tients randomized to a bolus of abciximab followed
by a 12-hour infusion had a 35% reduction in the
primary composite end point of death, MI, or ur-
gent intervention at 30 days. These benefits were
maintained at 6 months and at 3 years and were
greatest in magnitude in those who had evolving
MI [44,45]. The EPILOG trial, another large trial
solidifying the efficacy of abciximab in PCI, ran-
domized patients undergoing elective or urgent
PCI to abciximab plus standard- or low-dose
Eptifibatide Tirofiban
Cyclic heptapeptide inhibitor Nonpeptide inhibitor
180 m/kg  2 (10-min interval) 0.4 m/kg/min  30 mina
2 m/kg/min  18–24 h 0.1 m/kg/min  18–24 ha
2.5 h 2h
Load: None Infusion: 50% Load: 50% Infusion: 50%
(CrCl!50 mL/min) (CrCl!30 mL/min)
No No
Yes Yes
 ANTITHROMBOTIC THERAPY FOR PCI
adjunctive heparin or to placebo plus standard-
dose heparin. The trial was halted prematurely af-
ter enrolling 2792 patients because the composite
end point was significantly reduced in both abcixi-
mab groups compared with placebo. The treat-
ment effect was seen in high- and low-risk
individuals and was maintained at 6 months and
at 1 year [46,47]. Validation for abciximab in the
setting of intracoronary stenting came from the
EPISTENT trial, a 2399-patient randomized study
of stenting alone, abciximab plus stenting, or ab-
ciximab plus PTCA. Abciximab-treated patients
demonstrated a significantly lower incidence of
death, MI, or urgent revascularization at 30 days
and at 6 months [48,49]. In addition, the beneficial
effects of stenting plus abciximab on death and
large MI were sustained at 1 year [50]. Pooled anal-
ysis of these three large randomized trials con-
firmed that the benefits of abciximab on 30-day
death or MI are irrespective of sex or the PCI de-
vice used [51]. More important, the significant
mortality benefit observed at short-term follow-up
is durable, with an absolute risk reduction for death
of 1.4% at 3 years (Fig. 4) [52].
The concept that selective patient populations
derive a greater clinical benefit from the use of
abciximab arose early in the evaluation of the
drug. In the CAPTURE trial, for instance, 1265
patients who had medically refractory non–ST
elevation ACS were randomized to receive abcix-
imab versus placebo 18 to 24 hours before PTCA.
Fig. 4. Cumulative mortality at 3-year follow-up of ab-
ciximab-versus placebo-treated patients in the combined
EPIC, EPILOG, and EPISTENT trials: intention-to-
treat analysis. The differences were statistically signifi-
cant (hazard ratio ¼ 0.78; 95% confidence interval:
0.63–0.98; P ¼ 0.03). (From Topol EJ, Lincoff AM, Ker-
eiakes DJ, et al. Multi-year follow-up of abciximab ther-
apy in three randomized, placebo-controlled trials of
percutaneous coronary revascularization. Am J Med
2002;113(1):1–6; with permission.)
181
The benefit observed in this trial was predomi-
nantly limited to patients who had complex
lesions on angiography or elevated serum tropo-
nin levels [53]. A similar finding was observed on
analysis of the subset of patients in the EPIS-
TENT trial who had complex atherosclerotic le-
sions [54]. In addition to patients who have
complex lesions, post hoc analyses suggested
that abciximab may be of particular benefit in
those who have diabetes. In a pooled analysis of
the 1462 patients who had diabetes enrolled in
the EPIC, EPILOG, and EPISTENT trials, abcix-
imab reduced their 1-year mortality to that ob-
served in patients who did not have diabetes and
were treated with placebo (Fig. 5) [55]. To date,
however, the ISAR–SWEET trial has been the
only large prospective trial to evaluate abciximab
specifically in patients who have diabetes. In this
trial, 701 patients receiving 600 mg of clopidogrel
at least 2 hours before planned PCI were ran-
domly assigned to abciximab or placebo with hep-
arin. Although there was reduced angiographic
restenosis and target vessel revascularization in
the abciximab-treated group, there was no differ-
ence in the incidence of death or MI at 1 year in
this study [56]. A potential explanation for the re-
duced restenosis rates observed in this and a num-
ber of the other abciximab trials relates to the
effects that the drug may have on the vitronectin
receptor and subsequent neointimal proliferation
[57]. A prospective trial of 225 patients, however,
found no difference between abciximab and pla-
cebo on neointimal proliferation as measured by
Fig. 5. Kaplan-Meier curves, pooled from the EPIC,
EPILOG, and EPISTENT trials, for 1-year mortality
in diabetics and nondiabetics randomized to placebo
(PL) or abciximab (ABX). (From Bhatt DL, Marso SP,
Loncoff AM, et al. Abciximab reduces mortality in
diabetics following percutaneous coronary intervention.
J Am Coll Cardiol 2000;35(4):922–8; with permission.)
182 BARMAN & BHATT
intravascular ultrasound or quantitative angiogra-
phy [58].
One group of patients who may not derive
a detectable benefit from the routine use of
abciximab includes those who are adequately
pretreated with clopidogrel and who are at
particularly low risk. The ISAR–REACT trial
evaluated such patients. In this trial, pretreatment
with 600 mg of clopidogrel for a median of 7.4
hours seemed to negate any beneficial effects of
abciximab on death, MI, or urgent revasculariza-
tion at 30 days [28]. Evaluation of high-dose pre-
treatment in higher-risk patients and the
implications that this strategy may have on the
utility of abciximab and other glycoprotein IIb/
IIIa inhibitors is ongoing in the ISAR–REACT-2
study.
Despite the decreasing number of patients
suffering ST-segment elevation MI (STEMI) an-
nually, it remains an important subset presenting
to the catheterization laboratory. The use of ab-
ciximab as adjunctive pharmacotherapy in pa-
tients who have STEMI undergoing primary
angioplasty has also been extensively studied;
most of these studies suggest benefit of abciximab
in this setting. In the ADMIRAL trial, 300 pa-
tients were randomized to placebo or abciximab
before PTCA or stenting. The primary composite
end point of death, MI, or revascularization was
significantly lower at 30 days and at 6 months in
the abciximab group. In addition, TIMI-3 flow
rates at initial catheterization and 6-month left
ventricular ejection fraction were higher in the ab-
ciximab group. These benefits were enhanced in
the subset of patients who received abciximab ear-
lier (ie, in the ambulance or emergency room),
a finding subsequently supported by a meta-anal-
ysis of six large trials of glycoprotein IIb/IIIa in-
hibition in primary PCI [59,60]. Evaluation at
3-year follow up confirmed the durability of ben-
efit of abciximab treatment by demonstrating
a 3.1% absolute risk reduction in all-cause mor-
tality (9.1% versus 12.2%) [61]. Similar beneficial
findings were observed in the ISAR-2, RAP-
PORT, and ACE trials [62–65]. The largest trial
of abciximab in STEMI patients, the CADILLAC
trial, resulted in much less benefit with abciximab.
In this trial, 2082 patients were randomized in
a two-by-two factorial design to placebo or
abciximab and PTCA or stenting. The 6-month
composite end point of death, MI, or ischemia-
driven revascularization was lower with stenting
compared with PTCA, irrespective of abciximab
use. Angiographic restenosis was also lower with
stenting, again with or without abciximab [66].
As a result of the somewhat conflicting results
seen in the CADILLAC trial, which may have
been related to study design and the exclusion of
sicker patients, a meta-analysis was performed
of trials evaluating abciximab in patients who
had STEMI treated with fibrinolysis and primary
PCI. This analysis demonstrated significant reduc-
tions in mortality at 30 days (2.4% versus 3.4%)
and between 6 and 12 months (4.4% versus
6.2%), without an increase in significant bleeding,
a benefit seen only in patients undergoing primary
PCI but not in those receiving fibrinolysis (Fig. 6)
[67]. These data provide strong support for abcix-
imab before primary angioplasty for STEMI.
Although the efficacy of abciximab is clear,
several features may limit its use. In particular,
acute profound thrombocytopenia resulting from
antibody formation, prolonged antiplatelet effect
given the relatively long half-life, and cost have
caused concern. As such, two other glycoprotein
IIb/IIIa inhibitors, eptifibatide and tirofiban, have
been studied in the setting of PCI.
Eptifibatide
Eptifibatide, first approved by the FDA in
1998, is a nonimmunogenic cyclic heptapeptide
derived from the structure of barbourin, a platelet
glycoprotein IIb/IIIa inhibitor found in the
venom of the southeastern pigmy rattlesnake.
Initial pilot studies of the medication suggested
that a bolus dose between 135 to 180 mg/kg was
sufficient to block 80% of ADP-mediated platelet
aggregation within 15 minutes of administration,
with various continuous infusions sustaining the
effect [68,69]. The PURSUIT trial evaluated the
180-mg/kg bolus dose of eptifibatide along with
a high- or low-infusion dose versus placebo in
patients who had ACS. In this high-risk patient
population, 46% of whom had non-STEMI,
eptifibatide therapy was associated with a lower
incidence of death or nonfatal MI at 30 days
(14.2% versus 15.7%) [70]. Although the benefit
was seen across all groups, it was observed that
those undergoing early PCI derived a larger bene-
fit [71]. Eptifibatide therapy in patients specifically
undergoing PCI was addressed in the IMPACT-II
trial. In this study, 4010 patients undergoing
planned, urgent, or emergent PCI were random-
ized to placebo or a bolus of eptifibatide and
one of two infusion doses before PCI. At 30
days, the incidence of the composite end point
of death, MI, unplanned revascularization, or
 ANTITHROMBOTIC THERAPY FOR PCI 183

Fig. 6. Incidence of and odds ratio plots with 95% confidence intervals (CIs) for 6- to 12-month mortality with abcix-
imab compared with placebo from trials of primary angioplasty and fibrinolysis in the treatment of STEMI. The size of
the data markers (squares) is approximately proportional to the sample size of each treatment group. (From De Luca G,
Suryapranata H, Stone GW, et al. Abciximab as adjunctive therapy to reperfusion in acute ST-segment elevation myo-
cardial infarction: a meta-analysis of randomized trials. JAMA 2005;293(14):1759–65; with permission.)

stent placement for abrupt vessel closure was
significantly decreased in the eptifibatide plus
lower-dose-infusion group. The treatment bene-
fit observed, however, was not maintained at
6 months [72]. The proposed explanation for the
lack of durable efficacy in the IMPACT-II trial
was the fact that the bolus and infusion doses
studied were lower than those used in the PUR-
SUIT trial. Indeed, only 50% of maximal platelet
inhibition was achieved in the IMPACT-II trial
[73]. Subsequent study established the biochemical
efficacy of a novel, double-bolus strategy of eptifi-
batide before PCI [74]. This dosing strategy pro-
vided the basis for the dosing in the ESPRIT
trial that evaluated 2064 patients scheduled for
elective PCI and ultimately confirmed the efficacy
of eptifibatide before PCI. In the ESPRIT trial,
patients were randomized to double-bolus eptifi-
batide plus continuous infusion or to placebo.
The trial was terminated prematurely because
there was a large decrease in the primary end point
of death, MI, urgent revascularization, or throm-
botic bailout use of glycoprotein IIb/IIIa inhibitor
therapy within 48 hours in the eptifibatide group
compared with placebo. The results were durable
to 1 year and were consistent among all subgroups
analyzed (Fig. 7) [75–77]. The safety of eptifibatide
is well established. Although there was an
increased need for transfusion in eptifibatide-
treated patients in the PURSUIT trial, pooled
analysis has not demonstrated an increased
risk of thrombocytopenia or of hemorrhagic
stroke, the most devastating potential complica-
tion of periprocedural antithrombotic therapy
[78]. Given the findings of the ESPRIT trial
and the cumulative safety data, double-bolus
plus higher-dose continuous infusion of eptifiba-
tide has become the standard dose for patients
receiving this glycoprotein IIb/IIIa inhibitor as
adjunctive pharmacotherapy in the setting of
coronary stent implantation.
Tirofiban
The third intravenous glycoprotein IIb/IIIa
inhibitor used in clinical practice is tirofiban,
which also was approved by the FDA in 1998.
The pharmacokinetic features of this nonpeptide
competitive inhibitor of the glycoprotein IIb/IIIa
inhibitor are given in Table 2. Phase I study of the
drug identified the optimal bolus and infusion
dose to achieve significant and sustained platelet
inhibition as measured by an ex vivo platelet ag-
gregation assay [79]. It also provided the basis
for the dosing of tirofiban in the RESTORE trial,
the first large randomized trial of the drug in PCI.
184 BARMAN & BHATT

Fig. 7. Rates of death or MI at 1 year and odds ratio plots with 95% confidence intervals (CIs) in various subgroups
from the ESPRIT trial. (From O’Shea JC, Butler CE, Cantor WJ, et al. Long-term efficacy of platelet glycoprotein IIb/
IIIa integrin blockade with eptifibatide in coronary stent intervention. JAMA 2002;287(5):618–21; with permission.)

Randomizing 2139 patients who had ACS and
were undergoing PCI within 72 hours, the RE-
STORE trial evaluated tirofiban versus placebo.
The composite end point of death, MI, failed an-
gioplasty requiring bypass surgery, urgent revas-
cularization, or implantation of an intracoronary
stent because of actual or threatened abrupt clo-
sure was lower in the tirofiban group compared
with placebo at 30-day (10.3% versus 12.2%)
and at 6-month follow-up (24.1% versus
27.1%), although the differences were not statisti-
cally significant [80,81]. In the PRISM-PLUS
trial, 1915 patients who had ACS were random-
ized to receive heparin only, tirofiban only, or tir-
ofiban plus heparin. Patients underwent
angioplasty after 48 hours of drug treatment at
the discretion of the treating physician. Random-
ization to the tirofiban-only arm was discontinued
early because this group demonstrated increased
mortality at 7 days. The group treated with tirofi-
ban and heparin, however, had a significant re-
duction in the incidence of death, MI, or
refractory ischemia at 7 days (12.9% versus
17.9%), an effect that was maintained at 30 days
(18.5% versus 22.3%) and at 6 months (27.7%
versus 32.1%). The benefits of tirofiban with hep-
arin were seen in patients who were treated medi-
cally and in those who underwent angioplasty
[82]. In addition, from PRISM-PLUS, five inde-
pendent clinical risk factors that predicted benefit
from tirofiban therapy were identified. Consistent
with other trials of glycoprotein IIb/IIIa inihibi-
tors that demonstrated increasing benefit in sicker
patients, those who had 0 or 1 risk factors had no
detectable advantage from tirofiban therapy,
whereas those who had more risk factors demon-
strated a clear and increasing benefit (Fig. 8) [83].
Despite the validation of tirofiban in patients who
have ACS, the lack of a significant durable benefit
in the RESTORE trial has limited its use in coro-
nary patients undergoing planned PCI. As with
eptifibatide, issues of dosing and the limitations
of ex vivo platelet aggregation studies as an accu-
rate surrogate of in vivo inhibition served as the
potential explanation for the relatively weaker ef-
ficacy of tirofiban in the catheterization labora-
tory. The ADVANCE trial attempted to resolve
this issue. This smaller study randomized 202 pa-
tients to high-dose bolus tirofiban and infusion
versus placebo on a baseline of heparin and aspi-
rin. At 6 months, the primary composite end point
of death, MI, or target vessel revascularization
was significantly lower in the high-dose tirofiban
group (20% versus 35%). Subgroup analysis sug-
gested that the benefits of tirofiban in this study
were limited to those who had ACS and those
 ANTITHROMBOTIC THERAPY FOR PCI
Fig. 8. Composite event rate of death, MI, or refractory
ischemia at 7 days for patients from the PRISM-PLUS
trial stratified by risk factor score and treatment group.
One point assigned to risk factor score for each of the
following: age greater than 65 years, prior bypass sur-
gery, antecedent aspirin use, antecedent b-blocker use,
and ST depressions on the presenting electrocardiogram.
ARR, absolute risk reduction; NS, not significant; %
popul., percentage of patients in each risk score group;
RRR, relative risk reduction. (From Sabatine MS,
Januzzi JL, Snapinn S, et al. A risk score system for
predicting adverse outcomes and magnitude of benefit
with glycoprotein IIb/IIIa inhibitor therapy in patients
with unstable angina pectoris. Am J Cardiol
2001;88(5):488–92; with permission.)
who had diabetes [84]. At present, the use of tiro-
fiban in most catheterization laboratories remains
limited to those patients who have ACS and are
already receiving the drug.
The use of glycoprotein IIb/IIIa inhibitors in
specific circumstances also merits close consider-
ation. For instance, in saphenous vein graft
intervention, the role of these agents is much less
well established. A pooled analysis of five large
randomized trials including 627 patients undergo-
ing saphenous vein graft intervention failed to
demonstrate any benefit of glycoprotein IIb/IIIa
inhibition [85]. Although there may be a role for
these agents in saphenous vein graft interventions
using emboli protection devices, this concept re-
quires further study before it can be clearly recom-
mended. Patients who have impaired renal
function also represent a population in whom
the utility of glycoprotein IIb/IIIa inhibition is
less firmly grounded. The issue is relevant because
this condition is increasingly prevalent and these
185
patients are generally at higher risk. Dose reduc-
tion of renally cleared glycoprotein IIb/IIIa inhib-
itors has been established, as has efficacy, albeit by
way of post hoc analysis. Prospective study in this
group is still necessary. Lastly, because most trials
of glycoprotein IIb/IIIa inhibitors occurred dur-
ing a time of balloon angioplasty and bare metal
stenting, their utility in the current era of drug-
eluting stenting must be inferred. Biologically,
there is unlikely to be a difference in the setting
of bare metal versus drug-eluting stenting because
delivery systems and technique are similar; how-
ever, because repeat prospective trials are not
forthcoming, analysis of registry data is important
to evaluate this question.
Overall, the general use of glycoprotein IIb/
IIIa inhibition with PCI is well supported by the
literature. A meta-analysis of 19 large randomized
trials of glycoprotein IIb/IIIa inhibitors in a vari-
ety of different patient populations confirmed
a significant reduction in mortality at 30 days
(relative risk reduction ¼ 0.69) and at 6 months
(relative risk reduction ¼ 0.79) compared with
placebo. The reduction in the risk of mortality is
only partially explained by reduction in MI [86].
The extent of benefit is clearly related to patient
features, with high-risk patients generally deriving
a greater degree of benefit. Patients who have di-
abetes and ACS especially seem to benefit from
glycoprotein IIb/IIIa inhibitor use, presumably
because they are generally at an increased state
of platelet activation [87,88]. On the contrary,
low-risk patients undergoing elective stenting,
particularly those who have been adequately pre-
treated with clopidogrel with respect to timing
and dose, may not benefit from adjunctive glyco-
protein IIb/IIIa inhibition. As far as comparative
efficacy is concerned, a separate meta-analysis
evaluated the efficacy and safety of the three
agents. Obviously limited by the included study
designs, this study suggested superiority of abcix-
imab compared with eptifibatide and tirofiban
with respect to MI and target vessel revasculariza-
tion at the cost of increased bleeding [89]. To date,
however, only 1 prospective randomized trial ex-
ists comparing two agents. In the TARGET trial,
4809 patients were randomized to receive abcixi-
mab or tirofiban before nonemergent stent im-
plantation. Designed to test for noninferiority of
tirofiban, this study found instead that abciximab
was significantly superior in terms of the primary
end point of death, MI, or urgent revasculariza-
tion at 30 days (6% versus 7.6%) [90]. Ultimately,
the decision of which agent to use as adjunctive
186 BARMAN & BHATT
pharmacotherapy in PCI rests on the operator.
The EPISTENT and ESPRIT trials support the
use of abciximab and eptifibatide, respectively,
in the setting of elective stenting. The use of tiro-
fiban in this setting is less well established and
may be a result of suboptimal dosing.
Another intravenous agent, lamifaban, with
potentially advantageous pharmacokinetics was
developed as a possible replacement to the current
glycoprotein IIb/IIIa agents being used. The
safety and efficacy of this nonpeptide inhibitor
was evaluated in the Canadian Lamifaban Study
and in PARAGON A and B, clinical trials that
enrolled patients who had ACS. Although these
studies suggested potential benefit of this medica-
tion, there is no plan to continue development of
this agent [91–93]. In addition, no large dedicated
study of this agent in patients undergoing planned
PCI has been performed. Oral glycoprotein IIb/
IIIa inhibitors, although extensively studied, are
discussed only briefly because they currently
have no clinical utility. Initially holding promise,
multiple trials of these oral agents yielded disap-
pointing results. A meta-analysis of four large
studies of patients treated with oral glycoprotein
IIb/IIIa inhibitors for 3 to 6 months, including
more than 33,000 patients, not only demonstrated
a clear lack of efficacy but also suggested in-
creased mortality in addition to increased rates
of major bleeding with their use (Fig. 9) [94].
The mechanism of the clinical failure of these
agents is still unclear.
Anticoagulants
It has long been known that antiplatelets alone
cannot substantially reduce platelet activation in
an environment of high thrombin activity. Be-
cause vessel injury during PCI induces a substan-
tial amount of thrombin generation through TF
activation of the coagulation cascade, the con-
cominant use of antithrombotic therapy during
this procedure has been intuitive.
Unfractionated and low molecular weight heparin
Unfractionated heparin (UFH) has been used
in the setting of PCI for over 20 years to prevent
periprocedural ischemic complications. By binding
circulating antithrombin, UFH induces a confor-
mation change that results in accelerated inactiva-
tion of several coagulation factors, particularly
factor Xa. Inhibition of factor IIa, thrombin,
occurs when larger molecular weight fractions of
heparin form a tertiary complex by binding to
thrombin and to antithrombin. Given the variable
dose-response relationship of heparin (a result of
drug composition inconsistencies and patient var-
iations of antithrombin activity), monitoring of
drug effect is essential. Because the activated
partial thromboplastin time is greatly prolonged
with high-dose heparin use, the activated clotting
time (ACT) has become the standard by which the
UFH effect is measured in the catheterization
laboratory [95]. Despite its widespread use, a dra-
matic absence of large-scale, well-conducted pro-
spective studies evaluating UFH in PCI exists. In
response to this lack of evidence, a pooled analysis
of six large randomized trials of novel anticoagu-
lants, in which UFH was used in the control arm,
was conducted to characterize the efficacy of hepa-
rin in PCI and to describe the optimal ACT for its
use. In this analysis, an ACT of 350 to 375 seconds
was associated with the lowest ischemic event rate
at 7 days after PCI. Because an ACT between 325
and 350 seconds was associated with lower rates of
bleeding, it is more widely accepted as the target
ACT during coronary interventions in which
UFH is the primary anticoagulant (Fig. 10) [96].
When UFH is used adjunctively with glycoprotein
IIb/IIIa inhibitors, the target ACT is significantly
lower. A large meta-analysis including 8369 pa-
tients who had available ACT data (89% of
whom received concominant glycoprotein IIb/
IIIa inhibition) suggested that the incidence of
death, MI, or target vessel revascularization was
equivalent in patients whose ACTs were less than
256, 257 to 296, 297 to 347, or greater than 348 sec-
onds. The risk of bleeding, however, was found to
increase with increasing ACT [97]. From these
data, it is generally accepted that the goal ACT
to achieve maximal safety and efficacy, when con-
cominant glycoprotein IIb/IIIa inhibitors are
used, is between 200 and 250 seconds. With respect
to duration of therapy, it was common in the initial
era of PTCA to continue UFH for up to 24 hours
after the procedure [98]. At the cost of increased
bleeding and prolonged hospitalization, it was
originally believed that prolonged therapy might
reduce postprocedural ischemic events. With the
advent of stents and more effective long-term oral
antiplatelet agents, however, subacute ischemic
events have substantially decreased, and UFH
therapy after coronary intervention is no longer
recommended [99].
Despite continuing to be the predominant
antithrombotic used during PCI, UFH has several
major limitations that have led to the search for
 ANTITHROMBOTIC THERAPY FOR PCI 187

Fig. 9. Odds ratios and 95% confidence intervals for risk of death (A), MI (B), and urgent revascularization (C) beyond
30 days with respect to low-dose oral glycoprotein IIb/IIIa inhibitor versus aspirin alone and high-dose oral glycoprotein
IIb/IIIa inhibitor versus aspirin alone. N indicates sample size. (From Chew DP, Bhatt DL, Sapp S, et al. Increased mor-
tality with oral platelet glycoprotein IIb/IIIa antagonists: a meta-analysis of phase III multicenter randomized trials. Cir-
culation 2001;103(2):201–6; with permission.)

alternative agents. In particular, UFH has a nar-
row therapeutic index with a highly variable dose-
response relationship. It is unable to inhibit
thrombin bound to fibrin, thus allowing existing
thrombus to grow. In addition, UFH may actu-
ally promote platelet aggregation, potentially
explaining the observed rebound hypercoaguabil-
ity seen when it is discontinued. Lastly, it has been
increasingly recognized that some patients may
suffer significant heparin-induced thrombocytope-
nia with re-exposure, a potentially fatal compli-
cation that can paradoxically be associated with
systemic thrombosis. In response to these limita-
tions, the development and study of low molecu-
lar weight heparins (LMWH), fragments of UFH
prepared by chemical or enzymatic depolymeriza-
tion, has been extensive. By predominantly inhib-
iting factor Xa, as opposed to thrombin directly,
LMWH inhibits upstream activation of the co-
agulation cascade, thus acting more efficiently and
avoiding many of the limitations of UFH.
LMWHs are more easily administered, have
a more reliable anticoagulant effect, and are
associated with fewer side effects than UFH
[100,101]. As such, these agents, of which enoxa-
parin is the most common, have been evaluated
in the setting of ACS and PCI.
In an early trial comparing intravenous
LMWH to intravenous UFH in PCI, 60 patients
received a 1-mg/kg dose of enoxaparin or UFH,
188 BARMAN & BHATT
Fig. 10. (A) Relationship between minimum ACT at or
around time of device activation and death, MI, or ur-
gent revascularization (Revasc) at 7 days. (B) Relation-
ship between maximum ACT and major or minor
bleeding events at 7 days. Lowest smoothing estimate
and 95% confidence intervals indicated by solid and dot-
ted lines, respectively. Percentages represent actual event
rates observed. (From Chew DP, Bhatt DL, Lincoff AM,
et al. Defining the optimal activated clotting time during
percutaneous coronary intervention: aggregate results
from 6 randomized, controlled trials. Circulation
2001;103(7):961–6; with permission.)
with repeat boluses as needed, to maintain an
ACT greater than 300 seconds. This small study
demonstrated comparable levels of factor Xa inhi-
bition and clinical outcomes, thus providing a ba-
sis for dosing of intravenous enoxaparin in the
catheterization laboratory [102]. Although this
and other small initial studies suggested benefit,
clinical efficacy and safety of enoxaparin in the
PCI was more firmly established by the NICE reg-
istry. In the NICE-1 study, 828 patients undergo-
ing elective or urgent PCI receiving 1 mg/kg of
intravenous enoxaparin at the time of the proce-
dure were found (at 30 days) to be comparable
to historical controls receiving UFH in terms of
bleeding, death, MI, or urgent target vessel revas-
cularization. Similarly, in the setting of abciximab
therapy for elective or urgent PCI, the NICE-4
trial demonstrated that outcomes in patients
receiving a reduced dose of intravenous enoxa-
parin (0.75 mg/kg) were comparable to outcomes
of historical controls from the EPILOG and
EPISTENT trials who received UFH with abcixi-
mab [103,104]. The CRUISE trial provided more
robust, prospective randomized data that sug-
gested the efficacy and safety of enoxaparin with
glycoprotein IIb/IIIa inhibition in elective PCI.
In this trial, 261 patients receiving eptifibatide
were randomized to receive 0.75 mg/kg of intrave-
nous enoxaparin or UFH with their procedure.
Clinical outcomes at 48 hours were similar in the
two groups, with the enoxaparin group demon-
strating a slight trend toward decreased bleeding
(4.1% versus 10.5%) [105]. More recently, the
STEEPLE study, which randomized 3528 patients
undergoing elective PCI to UFH or one of two
doses of enoxaparin, supported this trend of im-
proved safety. Specifically, the enoxaparin group
receiving a 0.75-mg/kg dose at the time of PCI
had a lower major bleeding rate compared with
the UFH group (1.2% versus 2.8%) [106].
The use of subcutaneous enoxaparin became
an accepted strategy for the medical therapy for
ACS after the completion of two large random-
ized trials in the late 1990s [107,108]. To address
the issue of additional enoxaparin dosing at the
time of PCI, the NICE-3 trial evaluated 628 pa-
tients being treated with subcutaneous enoxaparin
and a glycoprotein IIb/IIIa inhibitor; 283 went on
to PCI. Patients who received a subcutaneous
enoxaparin dose between 8 and 12 hours before
PCI were given an additional intravenous bolus
of 0.3 mg/kg, whereas those who received a dose
within 8 hours were given no additional anticoa-
gulation. Again, compared with historical con-
trols receiving heparin and a glycoprotein IIb/
IIIa inhibitor, there was no significant difference
in the incidence of death, MI, or bleeding related
to non–coronary artery bypass grafting [109].
Evaluating this prospectively, the INTERACT
study randomized patients who had ACS and
who received eptifibatide and aspirin to subcuta-
neous enoxaparin or to UFH for 48 hours. In
the two thirds of patients who ultimately pro-
ceeded to coronary angiography, nearly half un-
derwent PCI. Patients in the enoxaparin group
suffered significantly lower rates of bleeding and
recurrent ischemic events at 48 and 96 hours. In
addition, there was a lowering of the composite
end point of death and reinfarction at 30 days in
the group receiving enoxaparin (5.0% versus
9.0%) (Fig. 11) [110]. Although upstream use of
LMWH in ACS is accepted, the strategy of
 ANTITHROMBOTIC THERAPY FOR PCI
Fig. 11. Kaplan-Meier curves for the composite out-
come of death or nonfatal myocardial reinfarction dur-
ing the first 30 days after randomization in patients
treated with UFH versus enoxaparin. (From Goodman
SG, Fitchett D, Armstrong PW, et al. Randomized eval-
uation of the safety and efficacy of enoxaparin versus
unfractionated heparin in high-risk patients with
non-ST-segment elevation acute coronary syndromes re-
ceiving the glycoprotein IIb/IIIa inhibitor eptifibatide.
Circulation 2003;107(2):238–44; with permission.)
continued LMWH therapy after PCI is much less
well established. In a number of studies evaluating
this question, in terms of reduced restenosis and
subacute stent thrombosis, particularly in the set-
ting of dual antiplatelet therapy, LMWH has not
seemed to provide any additional benefit [111–
115]. It is therefore not routinely recommended.
Despite a more predictable anticoagulant dose
response, which may obviate the need for moni-
toring in most cases, a rapid point-of-care moni-
toring system would still likely improve the safety
and efficacy of LMWH in the catheterization
laboratory. Particularly in patients who have
weight extremes or renal insufficiency, such a test
might also guide timing of safe vascular sheath
removal and administration of protamine for
clinical bleeding (although protamine only reverses
w40% of enoxaparin activity). Indeed, the
ELECT trial, a nonrandomized observational
study of 445 patients, aimed to assess the predictive
value of the ENOX test, a surrogate measure of
anti-Xa activity. Patients in this trial who had an
ENOX time within the proposed target range of
250 to 450 seconds had a 4% combined ischemic
and bleeding event rate compared with 7.2% in pa-
tients outside of this range [116]. Despite FDA ap-
proval, however, this test of enoxaparin effect is
unlikely to receive widespread clinical acceptance
until it is validated in prospective trials.
The SYNERGY trial demonstrated the non-
inferiority of enoxaparin versus UFH in more
189
than 10,000 patients who had ACS generally
treated with an early invasive strategy. A signif-
icant proportion underwent PCI while on enox-
aparin, and the rate of thrombotic complications
was similar to that seen with UFH. A strategy of
switching from enoxaparin to UFH or vice versa
was associated with an increased hazard of
bleeding [117].
Despite the limitation of readily quantifying
effect, LMWH remains a reasonable choice for
anticoagulation in patients undergoing PCI. It
should be understood, however, that most clinical
trials establishing the efficacy of this class of
agents have used enoxaparin or dalteparin. Cur-
rently, there are seven clinically evaluated
LWMHs, each with different ratios of factor Xa
to IIa inhibition, and further investigation is
necessary to justify their individual use.
Direct thrombin inhibitors
Direct thrombin inhibitors represent another
major class of antithrombotics studied in the
setting of PCI. These agents do not require
binding to antithrombin to exert their anticoagu-
lant effect because they directly inhibit the cata-
lytic site on thrombin. Direct thrombin inhibitors
inhibit soluble and clot-bound thrombin, a major
potential advantage over heparin [118]. Other ad-
vantages over heparin include the fact that these
agents do not activate platelets, do not bind
plasma proteins, and are resistant to neutraliza-
tion by platelet factors [119]. These features and
the noncompetitive and reversible nature by
which direct thrombin inhibitors binds thrombin
allow for predictable, intense anticoagulation for
a controlled duration of time.
Initial studies of hirudin, a naturally occurring
direct thrombin inhibitor, yielded promising re-
sults. In the HELVETICA trial, 1141 patients
who had unstable angina undergoing PCI were
randomized to receive intravenous hirudin with
or without subcutaneous therapy for 3 days or
heparin therapy with the procedure. At 7 days,
patients receiving hirudin were significantly less
likely to suffer a cardiac event, although this
difference was not maintained at 7-month fol-
low-up [120]. In addition, subgroup analysis of
two large trials of hirudin as medical therapy for
ACS suggested that in those undergoing PCI, hir-
udin significantly reduced the risk of MI at 30
days compared with heparin. In these same trials,
however, it was found that the risk of major
190 BARMAN & BHATT
bleeding was significantly higher with hirudin
[121,122].
Bivalirudin, previously known as hirulog, rep-
resents one member in the family of subsequent-
generation, synthetic direct thrombin inhibitors.
Although other agents in this class have been
studied and approved in a variety of noncardiac
conditions including heparin-induced thrombocy-
topenia and venous thromboembolic disease,
bivalirudin boasts the most data in the arena of
ACS and PCI. Initial use of bivalirudin as
a primary anticoagulant in patients who have
unstable angina undergoing PTCA was based on
the cumulative results from the BAT, CACHET,
and REPLACE-1 trials. These trials suggested
that bivalirudin reduced major adverse cardiac
ischemic events and partially reduced major
bleeding complications compared with heparin
[123–125]. The potential uncoupling of periproce-
dural ischemic events and bleeding risk was very
encouraging and provided further basis for the
REPLACE-2 study. In this large study, 6010 pa-
tients undergoing coronary stenting were random-
ized to receive heparin with planned glycoprotein
IIb/IIIa inhibition or bivalirudin with provisional
glycoprotein IIb/IIIa inhibitor therapy. Nearly
all patients received clopidogrel pretreatment,
and only 7% of patients in the bivalirudin arm
required provisional glycoprotein IIb/IIIa use,
primarily for angiographic complications. Statisti-
cally designed to establish noninferiority for the
primary end point of death, MI, urgent revascu-
larization, or major in-hospital bleeding, bivaliru-
din satisfied all prespecified criteria to make this
claim. Although there was a nonsignificant reduc-
tion in postprocedural non-STEMI in the heparin
plus glycoprotein IIb/IIIa inhibitor group com-
pared with the bivalirudin group (5.8% versus
6.6%), this difference did not translate into a mor-
tality benefit at 1-year follow-up. In addition to
similar efficacy, bivalirudin was found to be asso-
ciated with a significant reduction in protocol-
defined major bleeding (2.4% versus 4.1%)
[126,127].
The increased safety of a PCI anticoagulation
strategy using bivalirudin as opposed to UFH
may be even greater in patients who have im-
paired renal function because bivalirudin clear-
ance is comparatively less dependent on renal
clearance. A recent pooled analysis of three trials
including 5035 patients undergoing PCI evaluated
this concept. In this study, the absolute benefit in
terms of bleeding and ischemic risk of bivalirudin
over heparin was 2.2%, 5.8%, 7.7%, and 14.4%
in patients who had normal, mildly, moderately,
or severely reduced renal function, respectively
(Fig. 12) [128]. Overall, REPLACE-2, along with
various trials preceding it, effectively established
bivalirudin as an acceptable alternative to UFH
with routine glycoprotein IIb/IIIa inhibition in
elective PCI. Especially in patients considered
low to moderate risk for periprocedural ischemic
complications or high risk for periprocedural
bleeding complications, this agent should be
strongly considered. The ongoing ACUITY trial
is examining the specific role of bivalirudin in
high-risk ACS patients managed in an aggressive,
contemporary approach, whereas the ongoing
HORIZONS trial is evaluating bivalirudin in the
setting of primary PCI of STEMI patients [129].
Novel antithrombin agents
The advancement in pharmacotherapy during
PCI over the last several years has been impressive.
The promise of continued improvements in efficacy
and safety exists as next-generation compounds in
existing drug classes and entirely new classes of
agents continue to be developed. One such class
includes the direct factor Xa inhibitors, which by
exerting their effect upstream in the coagulation
cascade may provide more amplified inhibition of
thrombin. Fondaparinux, an existing indirect in-
hibitor of factor Xa acting by way of antithrombin,
showed safety and efficacy in the ASPIRE trial of
Fig. 12. Odds ratio plots for the quadruple end point of
death, MI, urgent revascularization, and major hemor-
rhage in patients treated with bivalirudin versus heparin,
stratified by renal function. (From Chew DP, Bhatt DL,
Kimball W, et al. Bivalirudin provides increasing benefit
with decreasing renal function: a meta-analysis of ran-
domized trials. Am J Cardiol 2003;92(8):919–23; with
permission.)
 ANTITHROMBOTIC THERAPY FOR PCI
350 patients undergoing urgent or elective PCI
[130]. The 20,000-patient OASIS-5 trial of ACS
showed favorable ischemic outcomes with fonda-
parinux versus enoxaparin in ACS, with less bleed-
ing. There was, however, a concern of possibly
increased rates of catheter-associated thrombus
with fondaparinux [131]. The OASIS-6 trial in pa-
tients who have STEMI will provide more informa-
tion about fondaparinux in PCI. A large number of
direct factor Xa inhibitors have been developed, in-
cluding DX-9065a and otamixaban. XaNADU-
ACS, a phase II trial of DX-9065a versus heparin
in 402 patients who have ACS in whom glycopro-
tein IIb/IIIa inhibition and early PCI were encour-
aged, demonstrated a nonsignificant trend toward
improved efficacy and less bleeding [132]. Similarly,
XaNADU-PCI evaluated this agent in elective PCI
patients and found consistent results with respect to
safety and efficacy while providing rationale for the
recommended dosing regimen [133]. It does not ap-
pear, however, that clinical development of this
compound will proceed. SEPIA-PCI, a dose-rang-
ing phase II study of otamixaban versus UFH in
nonurgent PCI patients, has completed enrollment.
In aggregate, these studies will provide a solid foun-
dation on which larger scale clinical trials of direct
factor Xa inhibitors can be performed.
Another major class of novel antithrombotics
under current investigation is the TF inhibitors.
Exposure of coagulation factors to TF present in
the subendothelial layer during the plaque rupture
and vessel injury seen with ACS and PCI is
a critical step in the subsequent generation of
thrombin and the platelet-rich thrombus [134].
Applying the same general concept that upstream
coagulation cascade inhibition may lead to ampli-
fied anticoagulation, two TF antagonists have
currently been developed. Sunol-cH36, a chimeric
monoclonal antibody directed against TF, medi-
ates its anticoagulant effect by inhibiting binding
of factor X to the TF:VIIa complex, a critical
proximal step in the coagulation cascade. PROX-
IMATE–TIMI 27, the first trial of this agent in
humans, evaluated the safety, pharmacokinetics,
and dosing of this agent in 26 patients who had
stable coronary artery disease. Sunol-cH36 was
found to have dose-dependent anticoagulation ef-
fects at the cost of only a slight increase in the in-
cidence of minor nonplatelet count–dependent
mucosal bleeding [135]. Larger trials of safety
and efficacy using this agent are being planned.
The other agent in the class of TF inhibitors
that is currently being evaluated is the recombi-
nant nematode-associated protein c2 (rNAPc2).
191
Originally isolated from the hematophageous
nematode hookworm, Anclyostoma caninum, this
agent, like sunol-cH36, also blocks the TF:VIIa
enzymatic complex [136]. The recently completed
phase IIa dose-confirmation portion of the AN-
THEM–TIMI 32 study conducted in 203 patients
who had ACS treated with rNAPc2 or placebo in
addition to standard antithrombotic therapy dem-
onstrated a dose-related inhibition of thrombin
generation. In addition, major and minor bleeding
rates were not statistically different between the two
treatment groups (4.3% in rNAPc2 group versus
2.5% in placebo group) [137]. The follow-up
heparin-replacement portion of this study will
further evaluate the efficacy and safety of rNAPc2
by reducing the dose of and ultimately replacing
UFH in 50 to 100 patients being treated for ACS.
Summary
The extensive clinical study of antiplatelet and
anticoagulant therapy in PCI documented in this
review has not only justified the standard of care
but also provided the framework for the intro-
duction of novel therapeutic agents such as those
discussed previously. Although the evolution to
this current standard has been relatively rapid and
there have been substantial advancements made in
this arena, there remains a clinically relevant rate
of ischemic and bleeding complications. This fact,
coupled with ongoing advancements in the un-
derstanding of mechanisms of vascular injury and
thrombosis with ACS and PCI, continues to drive
researchers and clinicians to develop new classes
of agents with improved pharmacokinetic, effi-
cacy, and safety profiles. Currently, patient- and
lesion-specific characteristics guide the interven-
tionalist in deciding which specific antithrombotic
and antiplatelet agents to use before, during, and
after PCI. In the future, pharmacogenomics, the
science of examining genetic variations that dic-
tate response to drug therapy, may play a role in
the choice of anticoagulation for any given
patient. Until that time, however, the continued
execution of well-designed clinical trials of exist-
ing and novel pharmacotherapeutic agents and
adherence to the findings of such trials is neces-
sary to optimize the outcomes in the growing
population of patients undergoing PCI.
Appendix 1. Glossary of clinical trial acronyms
ACEdAbciximab and Carbostent Evaluation
trial
192 BARMAN & BHATT
ACUITYdAcute Catheterization and Urgent
Intervention Triage strategY
ADMIRALdAbciximab before Direct angio-
plasty and stenting in Myocardial Infarction
Regarding Acute and Long-term follow-up
ADVANCEdADditive Value of tirofiban Ad-
ministered with the high-dose bolus in the
preventioN of ischemic Complications dur-
ing high-risk coronary artEry angioplasty
ALBIONdAssessment of the best Loading
dose of clopidogrel to Blunt platelet activa-
tion, Inflammation, and Ongoing Necrosis
ANTHEMdAnticoagulation with Napc2 To
Help Eliminate Mace
ARMYDAdAntiplatelet therapy for Reduc-
tion of MYocardial Damage during
Angioplasty
ASPIREdArixtra Study in Percutaneous
coronary Intervention: a Randomized
Evaluation
BATdBivalirudin Angioplasty Trial
CACHETdComparison of Abciximab Com-
plications with Hirulog ischemic Events
Trial
CADILLACdControlled Abciximab and De-
vice Investigation to Lower Late Angio-
plasty Complications
CAPTUREdC7E3 AntiPlatelet Therapy in
Unstable REfractory angina
CHARISMAdClopidogrel for High Athero-
thrombotic Risk and Ischemic Stabilization,
Management and Avoidance
CLASSICSdCLopidogrel ASpirin Stent In-
ternational Cooperative Study
CREDOdClopidogrel for the Reduction of
Events During Observation
CRUISEdCoronary Revascularization Using
Integrilin and Single-bolus Enoxaparin
CUREdClopidogrel in Unstable angina to
prevent Recurrent Events
DISPERSE-2dDose Confirmation and Feasi-
bility Study of AZD614O þ Acetyl Salicylic
Acid (ASA) Compared with Clopidogrel þ
ASA in Patients with Non-ST Segment Ele-
vation Acute Coronary Syndromes
ELECTdEvaLuating Enoxaparin Clotting
Times
EPICdEvaluation of c7E3 for Prevention of
Ischemic Complications
EPILOGdEvaluation in PTCA to Improve
Long-term Outcome with abciximab Glyco-
protein IIb/IIIa blockade
EPISTENTdEvaluation of Platelet IIb/IIIa
Inhibitor for STENTing trial
ESPRITdEnhanced Suppression of the Plate-
let IIb/IIIa Receptor with Integrilin Therapy
FANTASTICdFull ANTicoagulation versus
ASpirin and TIClopidine
HELVETICAdHirudin in a European reste-
nosis prevention triaL VErsus heparin
Treatment In ptCA patients
HORIZONSdHarmonizing Outcomes with
RevascularIZatiON and Stents
IMPACTdIntegrilin to Manage Platelet Ag-
gregation to prevent Coronary Thrombosis
INTERACTdINTegrelin and Enoxaparin
Randomized assessment of Acute Coronary
syndrome Treatment
ISAR–CHOICEdIntracoronary Stenting and
Antithrombotic Regimen–Choose between
three High Oral doses for Immediate Clopi-
dogrel Effect
ISAR–REACTdIntracoronary Stenting and
Antithrombotic Regimen–Rapid Early Ac-
tion for Coronary Treatment
ISAR–SWEETdIntracoronary Stenting and
Antithrombotic Regimen–is abciximab a Su-
perior Way to Eliminate Elevated Throm-
botic risk in diabetics
JUMBO–TIMI 26dJoint Use of Medications
to Block Platelets Optimally
NICEdNational Investigators Collaborating
on Enoxaparin
OASISdOrganization to Assess Strategies for
Ischemia Syndromes
PARAGONdPlatelet IIb/IIIa Antagonism
for the Reduction of Acute coronary syn-
drome events in the Global Organization
Network
PRISM–PLUSdPlatelet Receptor Inhibition
for ischemic Syndrome Management in Pa-
tients Limited to very Unstable signs and
Symptoms
PROXIMATEdPROXimal Inhibition of co-
agulation using a Monoclonal Antibody to
TissuE factor (Sunol cH36)
PURSUITdPlatelet glycoprotein IIb/IIIa in
Unstable angina: Receptor Suppression Us-
ing Integrilin Therapy
RAPPORTdReoPro And Primary PTCA Or-
ganization and Randomized Trial
REPLACEdRandomized Evaluation in PCI
Linking Angiomax to reduced Clinical
Events
RESTOREdRandomized Efficacy Study of
Tirofiban for Outcomes and REstenosis
SEPIA–PCIdStudy to Evaluate the Pharma-
codynamics, the safety and tolerability, and
 ANTITHROMBOTIC THERAPY FOR PCI
the pharmacokinetics of several Intravenous
regimens of the factor XA inhibitor otamix-
aban, in comparison to intravenous unfrac-
tionated heparin, in subjects undergoing
nonurgent Percutaneous Coronary
Intervention
STARSdSTent Anticoagulation Restenosis
Study
STEEPLEdSafeTy and Efficacy of Enoxa-
parin in Percutaneous coronary intervention
patients, an internationaL randomized
Evaluation
SYNERGYdSuperior Yield of the New strat-
egy of Enoxaparin, Revascularization, and
GlYcoprotein IIb/IIIa inhibitors
TARGETddo Tirofiban And ReoPro Give
similar Efficacy outcomes Trial
TIMIdThrombolysis In Myocardial
Infarction
TRITON–TIMI 38dTRial to assess Improve-
ment in Therapeutic Outcomes by optimiz-
ing platelet inhibitioN with prasugrel
XaNADUdFirst Experience with Direct, Se-
lective Factor XA Inhibition in Patients
with Non-ST-elevation Acute Coronary
Syndromes
References
[1] Patrono C, Ciabattoni G, Patrignani P, et al. Clin-
ical pharmacology of platelet cyclooxygenase inhi-
bition. Circulation 1985;72(6):1177–84.
[2] Patrignani P, Filabozzi P, Patrono C. Selective
cumulative inhibition of platelet thromboxane pro-
duction by low-dose aspirin in healthy subjects.
J Clin Invest 1982;69(6):1366–72.
[3] Barnathan ES, Schwartz JS, Taylor L, et al. Aspirin
and dipyridamole in the prevention of acute coro-
nary thrombosis complicating coronary angio-
plasty. Circulation 1987;76(1):125–34.
[4] Schwartz L, Bourassa MG, Lesperance J, et al. As-
pirin and dipyridamole in the prevention of reste-
nosis after percutaneous transluminal coronary
angioplasty. N Engl J Med 1988;318(26):1714–9.
[5] Lembo NJ, Black AJ, Roubin GS, et al. Effect of
pretreatment with aspirin versus aspirin plus dipyr-
idamole on frequency and type of acute complica-
tions of percutaneous transluminal coronary
angioplasty. Am J Cardiol 1990;65(7):422–6.
[6] Peters RJ, Mehta SR, Fox KA, et al. Effects of as-
pirin dose when used alone or in combination with
clopidogrel in patients with acute coronary syn-
dromes: observations from the Clopidogrel in
Unstable angina to prevent Recurrent Events
(CURE) study. Circulation 2003;108(14):1682–7.
193
[7] Mufson LH. A randomized trial of aspirin in
PTCA: effect of high dose versus low dose aspirin
on major complications and restenosis. J Am Coll
Cardiol 1988;11:236A.
[8] Smith SCJ, Feldman TE, Hirshfeld JW Jr, et al.
ACC/AHA/SCAI 2005 guideline update for percu-
taneous coronary interventiondsummary article:
a report of the American College of Cardiology/
American Heart Association Task Force on Practice
Guidelines (ACC/AHA/SCAI Writing Committee
to Update the 2001 Guidelines for Percutaneous
Coronary Intervention). Circulation 2006;113:1–20.
[9] Mason PJ, Jacobs AK, Freedman JE. Aspirin resis-
tance and atherothrombotic disease. J Am Coll
Cardiol 2005;46(6):986–93.
[10] Gum PA, Kottke-Marchant K, Welsh PA, et al. A
prospective, blinded determination of the natural
history of aspirin resistance among stable patients
with cardiovascular disease. J Am Coll Cardiol
2003;41(6):961–5.
[11] Sharis PJ, Cannon CP, Loscalzo J. The antiplatelet
effects of ticlopidine and clopidogrel. Ann Intern
Med 1998;129(5):394–405.
[12] Altmann DB, Racz M, Battleman DS, et al. Re-
duction in angioplasty complications after the in-
troduction of coronary stents: results from a
consecutive series of 2242 patients. Am Heart J
1996;132(3):503–7.
[13] Leon MB, Baim DS, Popma JJ, et al. A clinical trial
comparing three antithrombotic-drug regimens af-
ter coronary-artery stenting. Stent Anticoagulation
Restenosis Study Investigators. N Engl J Med
1998;339(23):1665–71.
[14] Bertrand ME, Legrand V, Boland J, et al. Ran-
domized multicenter comparison of conventional
anticoagulation versus antiplatelet therapy in un-
planned and elective coronary stenting. The full
anticoagulation versus aspirin and ticlopidine
(FANTASTIC) study. Circulation 1998;98(16):
1597–603.
[15] Schomig A, Neumann FJ, Kastrati A, et al. A ran-
domized comparison of antiplatelet and anticoagu-
lant therapy after the placement of coronary-artery
stents. N Engl J Med 1996;334(17):1084–9.
[16] Schomig A, Neumann FJ, Walter H, et al. Coro-
nary stent placement in patients with acute myo-
cardial infarction: comparison of clinical and
angiographic outcome after randomization to
antiplatelet or anticoagulant therapy. J Am Coll
Cardiol 1997;29(1):28–34.
[17] Urban P, Macaya C, Rupprecht HJ, et al. Ran-
domized evaluation of anticoagulation versus anti-
platelet therapy after coronary stent implantation
in high-risk patients: the Multicenter Aspirin and
Ticlopidine Trial after Intracoronary Stenting
(MATTIS). Circulation 1998;98(20):2126–32.
[18] Neumann FJ, Hall D, Schomig A. Neutropenia
with ticlopidine plus aspirin. Lancet 1997;349(9064):
1552–3.
194 BARMAN & BHATT
[19] Steinhubl SR, Tan WA, Foody JM, et al. Incidence
and clinical course of thrombotic thrombocytope-
nic purpura due to ticlopidine following coronary
stenting. EPISTENT Investigators. Evaluation of
Platelet IIb/IIIa Inhibitor for Stenting. JAMA
1999;281(9):806–10.
[20] Bennett CL, Davidson CJ, Raisch DW, et al.
Thrombotic thrombocytopenic purpura associated
with ticlopidine in the setting of coronary artery
stents and stroke prevention. Arch Intern Med
1999;159(21):2524–8.
[21] Bertrand ME, Rupprecht HJ, Urban P, et al. Dou-
ble-blind study of the safety of clopidogrel with and
without a loading dose in combination with aspirin
compared with ticlopidine in combination with as-
pirin after coronary stenting: the Clopidogrel Aspi-
rin Stent International Cooperative Study
(CLASSICS). Circulation 2000;102(6):624–9.
[22] Bhatt DL, Bertrand ME, Berger PB, et al. Meta-
analysis of randomized and registry comparisons
of ticlopidine with clopidogrel after stenting. J Am
Coll Cardiol 2002;39(1):9–14.
[23] Quinn MJ, Fitzgerald DJ. Ticlopidine and clopi-
dogrel. Circulation 1999;100(15):1667–72.
[24] Yusuf S, Zhao F, Mehta SR, et al. Effects of clopi-
dogrel in addition to aspirin in patients with acute
coronary syndromes without ST-segment eleva-
tion. N Engl J Med 2001;345(7):494–502.
[25] Mehta SR, Yusuf S, Peters RJ, et al. Effects of pre-
treatment with clopidogrel and aspirin followed
by long-term therapy in patients undergoing percu-
taneous coronary intervention: the PCI-CURE
study. Lancet 2001;358(9281):527–33.
[26] Steinhubl SR, Berger PB, Mann JT, et al. Early and
sustained dual oral antiplatelet therapy following
percutaneous coronary intervention: a randomized
controlled trial. JAMA 2002;288(19):2411–20.
[27] Muller I, Seyfarth M, Rudiger S, et al. Effect of
a high loading dose of clopidogrel on platelet func-
tion in patients undergoing coronary stent place-
ment. Heart 2001;85(1):92–3.
[28] Kandzari DE, Berger PB, Kastrati A, et al. Influ-
ence of treatment duration with a 600-mg dose of
clopidogrel before percutaneous coronary revascu-
larization. J Am Coll Cardiol 2004;44(11):2133–6.
[29] Patti G, Colonna G, Pasceri V, et al. Randomized
trial of high loading dose of clopidogrel for reduc-
tion of periprocedural myocardial infarction in pa-
tients undergoing coronary intervention: results
from the ARMYDA-2 (Antiplatelet therapy for
Reduction of MYocardial Damage during Angio-
plasty) study. Circulation 2005;111(16):2099–106.
[30] Montalescot G. Assessment of the Best Loading
Dose of Clopidogrel to Blunt Platelet Activation,
Inflammation, and Ongoing Necrosis (ALBION)
trial. Presented at the 12th EuroPCR. Paris, France,
May 24–27, 2005.
[31] von Beckerath N, Taubert D, Pogatsa-Murray G,
et al. Absorption, metabolization, and antiplatelet
effects of 300-, 600-, and 900-mg loading doses of
clopidogrel: results of the ISAR-CHOICE (Intra-
coronary Stenting and Antithrombotic Regimen:
Choose Between 3 High Oral Doses for Immediate
Clopidogrel Effect) trial. Circulation 2005;112(19):
2946–50.
[32] Moreno R, Fernandez C, Hernandez R, et al.
Drug-eluting stent thrombosis: results from
a pooled analysis including 10 randomized studies.
J Am Coll Cardiol 2005;45(6):954–9.
[33] Nguyen TA, Diodati JG, Pharand C. Resistance to
clopidogrel: a review of the evidence. J Am Coll
Cardiol 2005;45(8):1157–64.
[34] Niitsu Y, Jakubowski JA, Sugidachi A, et al. Phar-
macology of CS-747 (prasugrel, LY640315),
a novel, potent antiplatelet agent with in vivo
P2Y12 receptor antagonist activity. Semin Thromb
Hemost 2005;31(2):184–94.
[35] Wiviott SD, Antman EM, Winters KJ, et al. Ran-
domized comparison of prasugrel (CS-747,
LY640315), a novel thienopyridine P2Y12 antago-
nist, with clopidogrel in percutaneous coronary
intervention: results of the Joint Utilization of Med-
ications to Block Platelets Optimally (JUMBO)-
TIMI 26 trial. Circulation 2005;111(25):3366–73.
[36] van Giezen JJ, Humphries RG. Preclinical and clin-
ical studies with selective reversible direct P2Y12
antagonists. Semin Thromb Hemost 2005;31(2):
195–204.
[37] Cannon CP. The DISPERSE 2 trial: safety, tolera-
bility, and preliminary efficacy of AZD6140, the
first oral, reversible ADP receptor antagonist, com-
pared with clopidogrel in patients with non-ST seg-
ment elevation acute coronary syndrome. Presented
at the 51st American Heart Association Scientific
Sessions. Dallas (TX), November 13–16, 2005.
[38] Bhatt DL, Topol EJ. Clopidogrel added to aspirin
versus aspirin alone in secondary prevention and
high-risk primary prevention: rationale and design
of the Clopidogrel for High Atherothrombotic
Risk and Ischemic Stabilization, Management,
and Avoidance (CHARISMA) trial. Am Heart J
2004;148(2):263–8.
[39] Hynes RO. Integrins: a family of cell surface recep-
tors. Cell 1987;48(4):549–54.
[40] Plow EF, Ginsberg MH. Cellular adhesion: GPIIb-
IIIa as a prototypic adhesion receptor. Prog
Hemost Thromb 1989;9:117–56.
[41] Coller BS. A new murine monoclonal antibody re-
ports an activation-dependent change in the con-
formation and/or microenvironment of the
platelet glycoprotein IIb/IIIa complex. J Clin In-
vest 1985;76(1):101–8.
[42] Coller BS, Scudder LE. Inhibition of dog platelet
function by in vivo infusion of F(ab’)2 fragments
of a monoclonal antibody to the platelet glycopro-
tein IIb/IIIa receptor. Blood 1985;66(6):1456–9.
[43] Ellis SG, Tcheng JE, Navetta FI, et al. Safety and
antiplatelet effect of murine monoclonal antibody
 ANTITHROMBOTIC THERAPY FOR PCI
7E3 Fab directed against platelet glycoprotein IIb/
IIIa in patients undergoing elective coronary angio-
plasty. Coron Artery Dis 1993;4(2):167–75.
[44] Use of a monoclonal antibody directed against the
platelet glycoprotein IIb/IIIa receptor in high-risk
coronary angioplasty. The EPIC Investigation.
N Engl J Med 1994;330(14):956–61.
[45] Topol EJ, Ferguson JJ, Weisman HF, et al. Long-
term protection from myocardial ischemic events in
a randomized trial of brief integrin beta3 blockade
with percutaneous coronary intervention. EPIC In-
vestigator Group. Evaluation of Platelet IIb/IIIa
Inhibition for Prevention of Ischemic Complica-
tion. JAMA 1997;278(6):479–84.
[46] Platelet glycoprotein IIb/IIIa receptor blockade
and low-dose heparin during percutaneous coro-
nary revascularization. The EPILOG Investiga-
tors. N Engl J Med 1997;336(24):1689–96.
[47] Lincoff AM, Tcheng JE, Califf RM, et al. Sustained
suppression of ischemic complications of coronary
intervention by platelet GP IIb/IIIa blockade with
abciximab: one-year outcome in the EPILOG trial.
Evaluation in PTCA to Improve Long-term Out-
come with abciximab GP IIb/IIIa blockade. Circu-
lation 1999;99(15):1951–8.
[48] Randomised placebo-controlled and balloon-
angioplasty-controlled trial to assess safety of
coronary stenting with use of platelet glycopro-
tein-IIb/IIIa blockade. The EPISTENT Investiga-
tors. Evaluation of Platelet IIb/IIIa Inhibitor for
Stenting. Lancet 1998;352(9122):87–92.
[49] Lincoff AM, Califf RM, Moliterno DJ, et al. Com-
plementary clinical benefits of coronary-artery
stenting and blockade of platelet glycoprotein
IIb/IIIa receptors. Evaluation of Platelet IIb/IIIa
Inhibition in Stenting Investigators. N Engl J
Med 1999;341(5):319–27.
[50] Topol EJ, Mark DB, Lincoff AM, et al. Outcomes
at 1 year and economic implications of platelet gly-
coprotein IIb/IIIa blockade in patients undergoing
coronary stenting: results from a multicentre rand-
omised trial. EPISTENT Investigators. Evaluation
of Platelet IIb/IIIa Inhibitor for Stenting. Lancet
1999;354(9195):2019–24.
[51] Cho L, Topol EJ, Balog C, et al. Clinical benefit of
glycoprotein IIb/IIIa blockade with abciximab is
independent of gender: pooled analysis from
EPIC, EPILOG and EPISTENT trials. Evaluation
of 7E3 for the Prevention of Ischemic Complica-
tions. Evaluation in Percutaneous Transluminal
Coronary Angioplasty to Improve Long-Term
Outcome with Abciximab GP IIb/IIIa blockade.
Evaluation of Platelet IIb/IIIa Inhibitor for Stent-
ing. J Am Coll Cardiol 2000;36(2):381–6.
[52] Topol EJ, Lincoff AM, Kereiakes DJ, et al. Multi-
year follow-up of abciximab therapy in three ran-
domized, placebo-controlled trials of percutaneous
coronary revascularization. Am J Med 2002;
113(1):1–6.
195
[53] Randomised placebo-controlled trial of abciximab
before and during coronary intervention in refrac-
tory unstable angina: the CAPTURE Study. Lan-
cet 1997;349(9063):1429–35.
[54] Cura FA, Bhatt DL, Lincoff AM, et al. Pro-
nounced benefit of coronary stenting and adjunc-
tive platelet glycoprotein IIb/IIIa inhibition in
complex atherosclerotic lesions. Circulation 2000;
102(1):28–34.
[55] Bhatt DL, Marso SP, Lincoff AM, et al. Abciximab
reduces mortality in diabetics following percutane-
ous coronary intervention. J Am Coll Cardiol 2000;
35(4):922–8.
[56] Mehilli J, Kastrati A, Schuhlen H, et al. Random-
ized clinical trial of abciximab in diabetic patients
undergoing elective percutaneous coronary inter-
ventions after treatment with a high loading dose
of clopidogrel. Circulation 2004;110(24):3627–35.
[57] Tam SH, Sassoli PM, Jordan RE, et al. Abcixi-
mab (ReoPro, chimeric 7E3 Fab) demonstrates
equivalent affinity and functional blockade of gly-
coprotein IIb/IIIa and alpha(v)beta3 integrins.
Circulation 1998;98(11):1085–91.
[58] Acute platelet inhibition with abciximab does not
reduce in-stent restenosis (ERASER study). The
ERASER Investigators. Circulation 1999;100(8):
799–806.
[59] Montalescot G, Barragan P, Wittenberg O, et al.
Platelet glycoprotein IIb/IIIa inhibition with coro-
nary stenting for acute myocardial infarction. N Engl
J Med 2001;344(25):1895–903.
[60] Montalescot G, Borentain M, Payot L, et al. Early
vs late administration of glycoprotein IIb/IIIa
inhibitors in primary percutaneous coronary inter-
vention of acute ST-segment elevation myocardial
infarction: a meta-analysis. JAMA 2004;292(3):
362–6.
[61] Three-year duration of benefit from abciximab in
patients receiving stents for acute myocardial in-
farction in the randomized double-blind ADMI-
RAL study. Eur Heart J 2005;26(23):2520–3.
[62] Neumann FJ, Kastrati A, Schmitt C, et al. Effect of
glycoprotein IIb/IIIa receptor blockade with abcix-
imab on clinical and angiographic restenosis rate
after the placement of coronary stents following
acute myocardial infarction. J Am Coll Cardiol
2000;35(4):915–21.
[63] Brener SJ, Barr LA, Burchenal JE, et al. Random-
ized, placebo-controlled trial of platelet glycopro-
tein IIb/IIIa blockade with primary angioplasty
for acute myocardial infarction. ReoPro And Pri-
mary PTCA Organization and Randomized Trial
(RAPPORT) Investigators. Circulation 1998;
98(8):734–41.
[64] Antoniucci D, Rodriguez A, Hempel A, et al. A
randomized trial comparing primary infarct artery
stenting with or without abciximab in acute myo-
cardial infarction. J Am Coll Cardiol 2003;42(11):
1879–85.
196 BARMAN & BHATT
[65] Antoniucci D, Migliorini A, Parodi G, et al. Abcix-
imab-supported infarct artery stent implantation
for acute myocardial infarction and long-term sur-
vival: a prospective, multicenter, randomized trial
comparing infarct artery stenting plus abciximab
with stenting alone. Circulation 2004;109(14):
1704–6.
[66] Stone GW, Grines CL, Cox DA, et al. Comparison
of angioplasty with stenting, with or without abcix-
imab, in acute myocardial infarction. N Engl J Med
2002;346(13):957–66.
[67] De Luca G, Suryapranata H, Stone GW, et al.
Abciximab as adjunctive therapy to reperfusion in
acute ST-segment elevation myocardial infarction:
a meta-analysis of randomized trials. JAMA
2005;293(14):1759–65.
[68] Phillips DR, Scarborough RM. Clinical pharma-
cology of eptifibatide. Am J Cardiol 1997;80(4A):
11B–20B.
[69] Harrington RA, Kleiman NS, Kottke-Marchant K,
et al. Immediate and reversible platelet inhibition
after intravenous administration of a peptide
glycoprotein IIb/IIIa inhibitor during percutane-
ous coronary intervention. Am J Cardiol 1995;
76(17):1222–7.
[70] Inhibition of platelet glycoprotein IIb/IIIa with
eptifibatide in patients with acute coronary syn-
dromes. The PURSUIT Trial Investigators. Plate-
let Glycoprotein IIb/IIIa in Unstable Angina:
Receptor Suppression Using Integrilin Therapy.
N Engl J Med 1998;339(7):436–43.
[71] Kleiman NS, Lincoff AM, Flaker GC, et al. Early
percutaneous coronary intervention, platelet inhi-
bition with eptifibatide, and clinical outcomes in
patients with acute coronary syndromes. PUR-
SUIT Investigators. Circulation 2000;101(7):
751–7.
[72] Randomised placebo-controlled trial of effect of
eptifibatide on complications of percutaneous
coronary intervention: IMPACT-II. Integrilin to
Minimise Platelet Aggregation and Coronary
Thrombosis-II. Lancet 1997;349(9063):1422–8.
[73] Phillips DR, Teng W, Arfsten A, et al. Effect of
Ca2 þ on GP IIb-IIIa interactions with integrilin:
enhanced GP IIb-IIIa binding and inhibition of
platelet aggregation by reductions in the concentra-
tion of ionized calcium in plasma anticoagulated
with citrate. Circulation 1997;96(5):1488–94.
[74] Tcheng JE, Talley JD, O’Shea JC, et al. Clinical
pharmacology of higher dose eptifibatide in percu-
taneous coronary intervention (the PRIDE study).
Am J Cardiol 2001;88(10):1097–102.
[75] Novel dosing regimen of eptifibatide in planned
coronary stent implantation (ESPRIT): a rando-
mised, placebo-controlled trial. Lancet 2000;
356(9247):2037–44.
[76] O’Shea JC, Hafley GE, Greenberg S, et al. Platelet
glycoprotein IIb/IIIa integrin blockade with eptifi-
batide in coronary stent intervention: the ESPRIT
trial: a randomized controlled trial. JAMA 2001;
285(19):2468–73.
[77] O’Shea JC, Buller CE, Cantor WJ, et al. Long-term
efficacy of platelet glycoprotein IIb/IIIa integrin
blockade with eptifibatide in coronary stent inter-
vention. JAMA 2002;287(5):618–21.
[78] Dasgupta H, Blankenship JC, Wood GC, et al.
Thrombocytopenia complicating treatment with
intravenous glycoprotein IIb/IIIa receptor inhibi-
tors: a pooled analysis. Am Heart J 2000;140(2):
206–11.
[79] Kereiakes DJ, Kleiman NS, Ambrose J, et al. Ran-
domized, double-blind, placebo-controlled dose-
ranging study of tirofiban (MK-383) platelet IIb/
IIIa blockade in high risk patients undergoing cor-
onary angioplasty. J Am Coll Cardiol 1996;27(3):
536–42.
[80] Effects of platelet glycoprotein IIb/IIIa blockade
with tirofiban on adverse cardiac events in patients
with unstable angina or acute myocardial infarc-
tion undergoing coronary angioplasty. The RE-
STORE Investigators. Randomized Efficacy
Study of Tirofiban for Outcomes and REstenosis.
Circulation 1997;96(5):1445–53.
[81] Gibson CM, Goel M, Cohen DJ, et al. Six-month
angiographic and clinical follow-up of patients pro-
spectively randomized to receive either tirofiban or
placebo during angioplasty in the RESTORE trial.
Randomized Efficacy Study of Tirofiban for Out-
comes and Restenosis. J Am Coll Cardiol 1998;
32(1):28–34.
[82] Inhibition of the platelet glycoprotein IIb/IIIa re-
ceptor with tirofiban in unstable angina and non-
Q-wave myocardial infarction. Platelet Receptor
Inhibition in Ischemic Syndrome Management in
Patients Limited by Unstable Signs and Symptoms
(PRISM-PLUS) Study Investigators. N Engl J
Med 1998;338(21):1488–97.
[83] Sabatine MS, Januzzi JL, Snapinn S, et al. A risk
score system for predicting adverse outcomes and
magnitude of benefit with glycoprotein IIb/IIIa in-
hibitor therapy in patients with unstable angina
pectoris. Am J Cardiol 2001;88(5):488–92.
[84] Valgimigli M, Percoco G, Barbieri D, et al. The
additive value of tirofiban administered with the
high-dose bolus in the prevention of ischemic com-
plications during high-risk coronary angioplasty:
the ADVANCE Trial. J Am Coll Cardiol 2004;
44(1):14–9.
[85] Roffi M, Mukherjee D, Chew DP, et al. Lack of
benefit from intravenous platelet glycoprotein IIb/
IIIa receptor inhibition as adjunctive treatment
for percutaneous interventions of aortocoronary
bypass grafts: a pooled analysis of five randomized
clinical trials. Circulation 2002;106(24):3063–7.
[86] Karvouni E, Katritsis DG, Ioannidis JP. Intrave-
nous glycoprotein IIb/IIIa receptor antagonists re-
duce mortality after percutaneous coronary
interventions. J Am Coll Cardiol 2003;41(1):26–32.
 ANTITHROMBOTIC THERAPY FOR PCI
[87] Boersma E, Harrington RA, Moliterno DJ, et al.
Platelet glycoprotein IIb/IIIa inhibitors in acute
coronary syndromes: a meta-analysis of all major
randomised clinical trials. Lancet 2002;359(9302):
189–98.
[88] Roffi M, Chew DP, Mukherjee D, et al. Platelet
glycoprotein IIb/IIIa inhibitors reduce mortality in
diabetic patients with non-ST-segment-elevation
acute coronary syndromes. Circulation 2001;104(23):
2767–71.
[89] Brown DL, Fann CS, Chang CJ. Meta-analysis of
effectiveness and safety of abciximab versus eptifi-
batide or tirofiban in percutaneous coronary inter-
vention. Am J Cardiol 2001;87(5):537–41.
[90] Topol EJ, Moliterno DJ, Herrmann HC, et al.
Comparison of two platelet glycoprotein IIb/IIIa
inhibitors, tirofiban and abciximab, for the preven-
tion of ischemic events with percutaneous coronary
revascularization. N Engl J Med 2001;344(25):
1888–94.
[91] Theroux P, Kouz S, Roy L, et al. Platelet mem-
brane receptor glycoprotein IIb/IIIa antagonism
in unstable angina. The Canadian Lamifiban Study.
Circulation 1996;94(5):899–905.
[92] International randomized, controlled trial of lami-
fiban (a platelet glycoprotein IIb/IIIa inhibitor),
heparin, or both in unstable angina. The PARA-
GON Investigators. Platelet IIb/IIIa Antagonism
for the Reduction of Acute coronary syndrome
events in a Global Organization Network. Circula-
tion 1998;97(24):2386–95.
[93] Randomized placebo-controlled trial of titrated in-
travenous lamifiban for acute coronary syndromes.
Circulation 2002;105(3):316–21.
[94] Chew DP, Bhatt DL, Sapp S, et al. Increased mor-
tality with oral platelet glycoprotein IIb/IIIa antag-
onists: a meta-analysis of phase III multicenter
randomized trials. Circulation 2001;103(2):201–6.
[95] Hirsh J, Warkentin TE, Shaughnessy SG, et al.
Heparin and low-molecular-weight heparin:
mechanisms of action, pharmacokinetics, dosing,
monitoring, efficacy, and safety. Chest 2001;
119(1 Suppl):64S–94S.
[96] Chew DP, Bhatt DL, Lincoff AM, et al. Defining
the optimal activated clotting time during percuta-
neous coronary intervention: aggregate results
from 6 randomized, controlled trials. Circulation
2001;103(7):961–6.
[97] Brener SJ, Moliterno DJ, Lincoff AM, et al. Rela-
tionship between activated clotting time and ische-
mic or hemorrhagic complications: analysis of 4
recent randomized clinical trials of percutaneous
coronary intervention. Circulation 2004;110(8):
994–8.
[98] Ferguson JJ, Dohmen P, Wilson JM, et al. Results
of a national survey on anticoagulation for PTCA.
J Invasive Cardiol 1995;7(5):136–41.
[99] Rabah M, Mason D, Muller DW, et al. Heparin af-
ter percutaneous intervention (HAPI): a prospective
197
multicenter randomized trial of three heparin regi-
mens after successful coronary intervention. J Am
Coll Cardiol 1999;34(2):461–7.
[100] Hirsh J, Raschke R. Heparin and low-molecular-
weight heparin: the Seventh ACCP Conference on
Antithrombotic and Thrombolytic Therapy. Chest
2004;126(3 Suppl):188S–203S.
[101] Warkentin TE, Greinacher A. Heparin-induced
thrombocytopenia: recognition, treatment, and
prevention: the Seventh ACCP Conference on
Antithrombotic and Thrombolytic Therapy. Chest
2004;126(3 Suppl):311S–37S.
[102] Rabah MM, Premmereur J, Graham M, et al. Use-
fulness of intravenous enoxaparin for percutaneous
coronary intervention in stable angina pectoris. Am
J Cardiol 1999;84(12):1391–5.
[103] Young JJ, Kereiakes DJ, Grines CL. Low-molecu-
lar-weight heparin therapy in percutaneous coro-
nary intervention: the NICE 1 and NICE 4 trials.
National Investigators Collaborating on Enoxa-
parin Investigators. J Invasive Cardiol 2000;
12(Suppl E):E14–8 [discussion: E25–8].
[104] Kereiakes DJ, Grines C, Fry E, et al. Enoxaparin
and abciximab adjunctive pharmacotherapy during
percutaneous coronary intervention. J Invasive
Cardiol 2001;13(4):272–8.
[105] Bhatt DL, Lee BI, Casterella PJ, et al. Safety of
concomitant therapy with eptifibatide and enoxa-
parin in patients undergoing percutaneous coro-
nary intervention: results of the Coronary
Revascularization Using Integrilin and Single bo-
lus Enoxaparin Study. J Am Coll Cardiol 2003;
41(1):20–5.
[106] Montalescot G. STEEPLE: Safety and Efficacy of
Enoxaparin in Percutaneous Coronary Intervention
Patients, an International Randomized Evaluation.
Presented at the 27th European Society of Cardiology
Congress. Stockholm, Sweden, September 3–7, 2005.
[107] Cohen M, Demers C, Gurfinkel EP, et al. A com-
parison of low-molecular-weight heparin with
unfractionated heparin for unstable coronary ar-
tery disease. Efficacy and Safety of Subcutaneous
Enoxaparin in Non-Q-Wave Coronary Events
Study Group. N Engl J Med 1997;337(7):447–52.
[108] Antman EM, McCabe CH, Gurfinkel EP, et al.
Enoxaparin prevents death and cardiac ischemic
events in unstable angina/non-Q-wave myocardial
infarction. Results of the Thrombolysis In Myocar-
dial Infarction (TIMI) 11B trial. Circulation 1999;
100(15):1593–601.
[109] Ferguson JJ, Antman EM, Bates ER, et al. Com-
bining enoxaparin and glycoprotein IIb/IIIa antag-
onists for the treatment of acute coronary
syndromes: final results of the National Investiga-
tors Collaborating on Enoxaparin-3 (NICE-3)
study. Am Heart J 2003;146(4):628–34.
[110] Goodman SG, Fitchett D, Armstrong PW, et al.
Randomized evaluation of the safety and efficacy
of enoxaparin versus unfractionated heparin in
198 BARMAN & BHATT
high-risk patients with non-ST-segment elevation
acute coronary syndromes receiving the glycopro-
tein IIb/IIIa inhibitor eptifibatide. Circulation
2003;107(2):238–44.
[111] Faxon DP, Spiro TE, Minor S, et al. Low molecu-
lar weight heparin in prevention of restenosis after
angioplasty. Results of Enoxaparin Restenosis
(ERA) Trial. Circulation 1994;90(2):908–14.
[112] Cairns JA, Gill J, Morton B, et al. Fish oils and
low-molecular-weight heparin for the reduction of
restenosis after percutaneous transluminal coro-
nary angioplasty. The EMPAR Study. Circulation
1996;94(7):1553–60.
[113] Grassman ED, Leya F, Fareed J, et al. A random-
ized trial of the low-molecular-weight heparin cer-
toparin to prevent restenosis following coronary
angioplasty. J Invasive Cardiol 2001;13(11):723–8.
[114] Zidar JP. Low-molecular-weight heparins in coro-
nary stenting (the ENTICES trial). ENoxaparin
and TIClopidine after Elective Stenting. Am J Car-
diol 1998;82(5B):29L–32L.
[115] Batchelor WB, Mahaffey KW, Berger PB, et al. A
randomized, placebo-controlled trial of enoxaparin
after high-risk coronary stenting: the ATLAST
trial. J Am Coll Cardiol 2001;38(6):1608–13.
[116] Moliterno DJ, Hermiller JB, Kereiakes DJ, et al. A
novel point-of-care enoxaparin monitor for use
during percutaneous coronary intervention. Re-
sults of the Evaluating Enoxaparin Clotting Times
(ELECT) Study. J Am Coll Cardiol 2003;42(6):
1132–9.
[117] Ferguson JJ, Califf RM, Antman EM, et al. Enox-
aparin vs unfractionated heparin in high-risk
patients with non-ST-segment elevation acute cor-
onary syndromes managed with an intended early
invasive strategy: primary results of the SYN-
ERGY randomized trial. JAMA 2004;292(1):
45–54.
[118] Weitz JI, Hudoba M, Massel D, et al. Clot-bound
thrombin is protected from inhibition by heparin-
antithrombin III but is susceptible to inactivation
by antithrombin III-independent inhibitors. J Clin
Invest 1990;86(2):385–91.
[119] Eitzman DT, Chi L, Saggin L, et al. Heparin neu-
tralization by platelet-rich thrombi. Role of platelet
factor 4. Circulation 1994;89(4):1523–9.
[120] Serruys PW, Herrman JP, Simon R, et al. A com-
parison of hirudin with heparin in the prevention
of restenosis after coronary angioplasty. HELVE-
TICA Investigators. N Engl J Med 1995;333(12):
757–63.
[121] Mehta SR, Eikelboom JW, Rupprecht HJ, et al. Ef-
ficacy of hirudin in reducing cardiovascular events
in patients with acute coronary syndrome undergo-
ing early percutaneous coronary intervention. Eur
Heart J 2002;23(2):117–23.
[122] Roe MT, Granger CB, Puma JA, et al. Comparison
of benefits and complications of hirudin versus hep-
arin for patients with acute coronary syndromes
undergoing early percutaneous coronary interven-
tion. Am J Cardiol 2001;88(12):1403–6, A6.
[123] Bittl JA, Strony J, Brinker JA, et al. Treatment with
bivalirudin (Hirulog) as compared with heparin
during coronary angioplasty for unstable or postin-
farction angina. Hirulog Angioplasty Study Inves-
tigators. N Engl J Med 1995;333(12):764–9.
[124] Lincoff AM, Kleiman NS, Kottke-Marchant K,
et al. Bivalirudin with planned or provisional abcix-
imab versus low-dose heparin and abciximab dur-
ing percutaneous coronary revascularization:
results of the Comparison of Abciximab Complica-
tions with Hirulog for Ischemic Events Trial (CA-
CHET). Am Heart J 2002;143(5):847–53.
[125] Lincoff AM, Bittl JA, Kleiman NS, et al. Compar-
ison of bivalirudin versus heparin during percuta-
neous coronary intervention (the Randomized
Evaluation of PCI Linking Angiomax to Reduced
Clinical Events [REPLACE]-1 trial). Am J Cardiol
2004;93(9):1092–6.
[126] Lincoff AM, Bittl JA, Harrington RA, et al. Biva-
lirudin and provisional glycoprotein IIb/IIIa block-
ade compared with heparin and planned
glycoprotein IIb/IIIa blockade during percutane-
ous coronary intervention: REPLACE-2 random-
ized trial. JAMA 2003;289(7):853–63.
[127] Lincoff AM, Kleiman NS, Kereiakes DJ, et al.
Long-term efficacy of bivalirudin and provisional
glycoprotein IIb/IIIa blockade vs heparin and
planned glycoprotein IIb/IIIa blockade during per-
cutaneous coronary revascularization: REPLACE-
2 randomized trial. JAMA 2004;292(6):696–703.
[128] Chew DP, Bhatt DL, Kimball W, et al. Bivalirudin
provides increasing benefit with decreasing renal
function: a meta-analysis of randomized trials.
Am J Cardiol 2003;92(8):919–23.
[129] Stone GW, Bertrand M, Colombo A, et al. Acute
Catheterization and Urgent Intervention Triage
strategY (ACUITY) trial: study design and ratio-
nale. Am Heart J 2004;148(5):764–75.
[130] Mehta SR, Steg PG, Granger CB, et al. Random-
ized, blinded trial comparing fondaparinux with
unfractionated heparin in patients undergoing con-
temporary percutaneous coronary intervention:
Arixtra Study in Percutaneous Coronary Interven-
tion: a Randomized Evaluation (ASPIRE) Pilot
Trial. Circulation 2005;111(11):1390–7.
[131] Yusuf S. An international, randomized, double-
blind study evaluating the efficacy and safety of
fondaparinux versus enoxaparin in the acute treat-
ment of unstable angina/non ST-segment elevation
MI acute coronary syndromes. Presented at the
27th European Society of Cardiology Congress.
Stockholm, Sweden, September 3–7, 2005.
[132] Alexander JH, Yang H, Becker RC, et al. First ex-
perience with direct, selective factor Xa inhibition
in patients with non-ST-elevation acute coronary
syndromes: results of the XaNADU-ACS Trial.
J Thromb Haemost 2005;3(3):439–47.
 ANTITHROMBOTIC THERAPY FOR PCI
[133] Alexander JH, Dyke CK, Yang H, et al. Initial ex-
perience with factor-Xa inhibition in percutaneous
coronary intervention: the XaNADU-PCI Pilot.
J Thromb Haemost 2004;2(2):234–41.
[134] Toschi V, Gallo R, Lettino M, et al. Tissue factor
modulates the thrombogenicity of human athero-
sclerotic plaques. Circulation 1997;95(3):594–9.
[135] Morrow DA, Murphy SA, McCabe CH, et al. Po-
tent inhibition of thrombin with a monoclonal
antibody against tissue factor (Sunol-cH36): results
of the PROXIMATE-TIMI 27 trial. Eur Heart J
2005;26(7):682–8.
199
[136] Stassens P, Bergum PW, Gansemans Y, et al. Anti-
coagulant repertoire of the hookworm Ancylostoma
caninum. Proc Natl Acad Sci U S A 1996;93(5):
2149–54.
[137] Giugliano RP, Wiviott SD, Morrow DA, et al.
Addition of a tissue-factor/factor VIIa inhibitor
to standard treatments in nste-acs managed with
an early invasive strategy: results of the phase 2
ANTHEM-TIMI 32 double-blind randomized
clinical trial. Presented at the 51st American
Heart Association Scientific Sessions. Dallas
(TX), November 13–6, 2005.
 Cardiol Clin 24 (2006) 201–215
Coronary Interventional Devices: Balloon,
Atherectomy, Thrombectomy and Distal
Protection Devices
Samin K. Sharma, MD, FACC*, Victor Chen, MBChB, FRACP
Cardiac Catheterization Laboratory, Cardiovascular Institute, Mount Sinai Hospital,
Box 1030, One Gustave Levy Place, New York, NY 10029, USA
Atherosclerosis is a complex disease in which
cholesterol deposition, inflammation, calcifica-
tion, and thrombus formation play major roles.
Rupture of high-risk vulnerable plaques rich in
cholesterol and plaque erosion are responsible for
occlusive coronary thrombosis and acute coro-
nary events by exposing highly thrombogenic,
cholesterol-rich material to the bloodstream.
Platelet adhesion and aggregation and activation
of the coagulation cascade are induced, which
results in thrombus formation. Newly formed
thrombus may further impair blood flow and
a spectrum of syndromes may ensue, from being
minimal or asymptomatic to causing accelerated
or unstable angina pectoris, myocardial infarction
(acute coronary syndrome [ACS]), or even sudden
death. Treatment of coronary or graft lesions with
associated thrombus (suspected or angiographi-
cally evident) is challenging for an intervention-
alist, because interventions in these cases are
associated with an increased incidence of pro-
cedural complications (distal embolization,
abrupt occlusion, slow or no-reflow, need for
emergent bypass surgery, and death). As an
attempt to decrease complications, strategies for
removing thrombus from the vessel before and
during the percutaneous intervention are devel-
oped (eg, thrombectomy and distal protection
devices). In other cases atherosclerotic plaque is
chronic and bulky, with a high fibrous and
calcium content making it resistant to routine
* Corresponding author.
E-mail address: samin.sharma@msnyuhealth.org
(S.K. Sharma).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.02.002
dilatation and expansion by balloons or stents.
These hard and complex lesions must be modified
by debulking and plaque modification adjunctive
devices (eg, percutaneous transluminal coronary
angioplasty (PTCA), atherectomy, atherotomy,
and laser) before stenting to attain good acute re-
sults with low complications, decreased rates of
stent thrombosis, and perhaps lower restenosis.
Although coronary stent implantation remains
the mainstay and ultimate step for percutaneous
treatment of most coronary lesions, some adjunc-
tive devices may be required for lesion prepara-
tion to facilitate stent deployment and optimal
expansion and prevent distal embolization [1,2].
Atherectomy, atherotomy, laser, and other
debulking devices
With increased operator experience and im-
proved device technology there has been a constant
growth in the number of complex lesions (ie, diffuse
lesions, calcified lesions, nondilatable rigid lesions,
ostial lesions, bifurcations, and chronic total
occlusions) attempted by interventionalists with
use of drug-eluting stents (DES), despite the fact
that data are lacking in their effectiveness. Pro-
cedural complexity, in-hospital complications, and
long-term outcome remain major concerns of DES
in these complex lesions because of large or re-
sistant plaque burdens leading to stent underex-
pansion, which possibly increases the chances of
stent thrombosis and restenosis. Stent deployment
is sometimes difficult in such complex calcified
lesions and is prone to problems such as stent
entrapment, stent stripping, underexpansion, and
cardiology.theclinics.com
202 SHARMA & CHEN
balloon rupture, with attendant risks of major
coronary dissection and acute in-hospital compli-
cations. Optimal stent expansion and final minimal
lumen diameter remain important predictors of
restenosis even after DES implantation [3]. Uni-
form rather than nonuniform stent expansion guar-
antees the homogenous drug delivery and diffusion
into the lesion by maintaining adequate distance
between stent struts [4]. Pretreatment of complex
lesions with high-pressure noncompliant balloon
inflation before DES is an option but is not always
successful and may be limited by plaque rigidity,
plaque shift, snow plough into side branch, exten-
sive vessel dissection, and even vessel perforation.
Incomplete stent apposition has been reported to
be a predictor for late stent thrombosis with DES.
On the other hand, high-pressure coronary stent
implantation in postmortem histopathologic
studies has been shown to cause arterial medial
disruption, break in internal elastic lamina, or lipid
core penetration by stent struts and may induce in-
creased arterial inflammation associated with
increased neointimal growth (Fig. 1) [5].
Studies using intravascular ultrasound (IVUS)
have suggested that neointimal area was predicted
by the degree of underlying plaque burden (plaque
area) before percutaneous coronary intervention
(PCI) and the minimal lumen area achieved after
PCI [6]. Prati and colleagues [7] demonstrated
with IVUS analysis that late loss (degree of inter-
nal hyperplasia) has a direct correlation with the
amount of residual plaque burden after stent im-
plantation, which suggests that debulking before
stenting might reduce restenosis. Approaching
complex lesions is challenging, and lesion prepara-
tion with plaque modification (debulking) tech-
niques is suggested as an important adjunct to
stenting, which maintains optimal integrity of
Fig. 1. (A) Degree of intimal hyperplasia in relation to vessel wall contact with struts. (B) Degree of inflammation re-
lated to stent struts contacts.
the DES delivery platform and polymer by allow-
ing uniform and optimal stent expansion. Routine
use of IVUS in addition to good angiographic vi-
sualization is suggested in selecting the appropri-
ate plaque modification strategy. Unlike balloon
angioplasty or stenting, which widens the coro-
nary lumen merely by displacing atherosclerotic
plaque, atherectomy or atherotomy techniques
widen the lumen by actually removing tissue
from the vessel wall or scoring the plaque in a con-
trolled manner. These plaque modification tech-
niques can be classified into following groups:
(1) atherectomy, which removes the plaque using
either directional coronary atherectomy (DCA)
(Guidant Corp., Indianapolis, Indiana) or rota-
tional atherectomy (RA) (Boston Scientific
Corp., Natick, Massachusetts); (2) atherotomy,
which cuts (scores) the plaque using low pressure
injury using either cutting balloon (CB) (Boston
Scientific Corp.) or the FX miniRAIL Balloon
(Guidant Corp.); (3) laser angioplasty using exci-
mer laser ablation (Spectranetics Corp., Colorado
Springs, Colorado); (4) other innovative newer de-
bulking devices.
Directional coronary atherectomy
The technique of DCA (ie, removing obstruc-
tive tissue by a catheter-based excision technique
using a nose cone, a metal cutter and housing, and
balloon inflation) was approved by the US Food
and Drug Administration in 1990. DCA is effec-
tive in removing fibrotic, noncalcified plaque,
particularly in aorto-ostial, branch ostial, bifur-
cation, and bulky eccentric lesions in proximal
large vessels (size R3 mm). Stand-alone DCA
has been shown to yield better acute and long-
term angiographic results than plain balloon
 CORONARY INTERVENTIONAL DEVICES 203

angioplasty when optimally performed; however,
late restenosis still remains high because DCA
does not eliminate arterial remodeling and chronic
recoil. Coronary stents reduce angiographic reste-
nosis by inhibiting acute recoil and chronic
arterial remodeling. The combined approach of
debulking followed by stenting has been suggested
to overcome the limitations of each device and
improve clinical outcomes. Data from several
registries suggested the feasibility, safety, and
efficacy of performing DCA before stenting,
demonstrating low restenosis rates and acceptable
acute results in these complex coronary lesions [8].
The encouraging results seen in these registries led
to two large, prospective, randomized clinical tri-
als that compared DCA before stenting with
stenting alone: the Atherectomy Before MULTI-
LINK Improves Lumen Gain and Clinical Out-
comes (AMIGO) trial and the Debulking and
Stenting In Restenosis Elimination (DESIRE)
trial. Long-term (8-month follow-up) results in
the AMIGO trial that randomized 753 patients
to either DCA followed by stenting or stenting
alone showed no difference in angiographic or
clinical restenosis between the two groups, with
a trend toward lower restenosis rates after optimal
debulking. The cumulative major adverse cardiac
events (MACE) at 30 days after procedure was
slightly higher in the DCA-treated patients [9].
The DESIRE trial randomized 500 patients to
IVUS-guided DCA followed by stenting or stenting
alone. Despite the achievement of a lower loss index
at follow-up in the DCA þ stent group (0.34 versus
0.41 mm), this trial also failed to show a clinical
benefit at 6-month follow-up in this group.
Based on the available data, DCA currently is
used sparingly in noncalcified ostial left anterior
descending and large bifurcation lesions to pro-
vide acceptable acute results by reducing plaque
shift and improving late angiographic outcome in
conjunction with stenting (Fig. 2). The use of
DCA has been limited for various reasons, includ-
ing poor tracking in sharply angulated lesions, dif-
ficult feasibility for calcified lesions, long
procedure time, risks of perforation, and Dotter
effects of the bulky device.
Rotational atherectomy
High-speed RA has been used preferentially in
the treatment of heavily calcified, ostial, and
undilatable coronary artery lesions. Such lesions
are usually associated with lower success and
higher complication rates with conventional stent-
ing because of difficulties in stent delivery and
expansion. High-pressure, noncompliant balloon
inflation for predilatation occasionally may suc-
ceed but is often insufficient to overcome vessel
wall/plaque resistance and may result in acute
recoil, arterial dissections, and perforation. The

Fig. 2. DCA and stent of ostial left anterior descending coronary artery lesion. (A) Angiogram before intervention
shows severe stenosis: 90%–95% obstruction of ostial left anterior descending. (B) 3.5–4.0 mm Flexicut device in left
anterior descending coronary artery. (C) An angiogram after 15 cuts at 40 psi. (D) The final angiogram shows good re-
sults after 3.5/18 mm Cypher stent at 12 atm, with residual obstruction of !10% and no dissections. (E) Interventional
site with no restenosis (residual !30%) on follow-up angiography 3 years later performed for left circumflex lesion.
204 SHARMA & CHEN
technique of RA involves plaque ablation and
pulverization by an abrasive diamond coated bur
that rotates at approximately 150,000 to 160,000
rpm. Rotablation causes differential cutting, with
selective ablation of diseased, inelastic calcified
stenosis and produces excellent acute procedural
results with relatively low complication rates. The
abraded plaque is pulverized into microparticles
5 to 10 mm in diameter that pass through the
coronary microcirculation and ultimately undergo
phagocytosis in the liver, spleen, and lung. Despite
the high acute procedural success rates, RA is
plagued by restenosis when used as a stand-alone
treatment, perhaps because of chronic arterial
recoil and adverse remodeling. It has been re-
ported that successful RA of severely calcified or
nondilatable coronary lesions facilitates stent de-
livery and expansion and reduces plaque shift,
which decreases side-branch closure [10,11].
Several nonrandomized, retrospective studies
showed improved procedural success rates and
a trend toward lower restenosis in calcified lesions
with use of RA before stenting versus stenting
alone [12–14]. These findings led to two random-
ized trials that examined the effect of debulking
with RA before stenting versus stenting alone.
In the Effects of Debulking on Restenosis
(EDRES) trial, 150 patients were randomized to
stenting alone versus RA with stenting. Although
acute gain was identical in both arms, there was
reduction in 6-month binary angiographic reste-
nosis in the RA þ stenting arm [15]. In the larger
Stent Implantation Post Rotational Atherectomy
Fig. 3. SPORT trial: acute results and follow-up events at 6–8 months.
(SPORT) trial (Fig. 3), 750 patients were random-
ized to receive either PTCA or RA before stent-
ing. The procedural success rate was better in
the RA group; however, there were no differences
in angiographic or clinical endpoints between the
two groups at 6 to 8 months follow-up [16]. These
results may be explained by selection bias in ex-
cluding severely calcified lesions to be randomized
in the trial. A prospective, randomized trial that
compared RA with PTCA before stenting in pa-
tients with chronic total occlusion reported that
the strategy of debulking before stenting yielded
significantly lower angiographic restenosis rates
at follow-up [17].
Another set of lesions in which RA has proved
beneficial is bifurcation lesions. A major issue
with stenting of bifurcation lesions is the snow
plough effect and plaque shift into the side
branch, which causes severe narrowing or occlu-
sion of the side branch, especially where there is
pre-existing ostial disease. Rotablation allows
definitive controlled debulking of bulky, sharply
angulated calcific bifurcation lesions along with
the creation of a channel/gutter to allow the
subsequent safe passage of balloons or stents
and reduce side-branch closure [18,19].
In summary, RA is recommended before
stenting in patients with severely calcified lesions,
undilatable lesions, chronic total occlusions, and
bifurcation lesions (Fig. 4). Use of this treatment
modality requires a high level of training and ex-
perience, however, to achieve good results consis-
tently with low rates of complications.
 CORONARY INTERVENTIONAL DEVICES 205

Fig. 4. RA and CB of left anterior descending /D1/D2 bifurcation calcified lesions. (A) Right anterior oblique cranial
projection before intervention shows severe stenosis: 80%–90% obstruction of mid- left anterior descending, 95% ob-
struction of distal left anterior descending, 80%–90% D1, and 70%–80% D2. (B) A 1.75-mm and then 2.15-mm Rota-
blator bur is passed through left anterior descending stenosis. (C) FXminiRAIL 2.5/10 mm atherotomy in D1. (D) CB
2.5/10 mm atherotomy in D2. (E) Final angiogram shows good results, with 10% residual obstruction after debulking,
dilatation, and stenting (3.5/18 mm Cypher stent) left anterior descending, 10%–30% obstruction of ostial D1, and
!10% obstruction of D2. (F) Follow-up angiography at 6 months reveals patent PCI sites in left anterior descending
/D1/D2.

Cutting balloon angioplasty of the target lesion is achieved at low-pressure

The CB device uses three to four longitudinally
mounted microsurgical atherotomes on the sur-
face of a noncompliant balloon. The device allows
precise scoring of atheromatous plaque and sev-
ering continuity of the elastic and fibrotic compo-
nents of the vessel wall, achieving lumen gain
mainly by plaque compression than by vessel wall
expansion (Fig. 5 PLUS) [20]. Optimal dilatation
Fig. 5. Mechanisms of lumen enlargement with CB an-
gioplasty versus conventional balloon angioplasty.
inflations, which potentially minimizes intimal
injury and subsequent neointimal proliferation.
Although randomized, clinical trials that com-
pared the CB with conventional PTCA showed
no differences in acute or long-term clinical and
angiographic outcomes, its novel mechanism of
action makes CB an attractive tool for lesion
preparation before stenting with DES in lesions
with high elastic recoil (ostial, bifurcation, or un-
dilatable mildly calcified lesions) [21]. In the
Restenosis Reduction by Cutting Balloon Evalua-
tion III (REDUCE III) trial, 521 patients were
randomized to CB before ICS (260 patients) or
PTCA before stenting (261 patients) with IVUS
versus angiographic guidance. At 6-month fol-
low-up, a significantly lower rate of angiographic
restenosis was seen with the CB-before-stenting
group, mainly in the IVUS-guided group [22].
Use of CB is recommended in ostial, bifurca-
tion, or undilatable mildly calcified lesions before
stenting. As a caveat, use of CB PTCA is
associated with slightly higher incidence of perfo-
ration compared with PTCA. Deliverability also
may be an issue with CB PTCA in sharply
angulated and tortuous lesions.
206 SHARMA & CHEN
FXminiRAIL balloon angioplasty
The FXminiRAIL PTCA catheter (Fig. 6) is an
innovative novel approach of lesion modification
that applies longitudinal force–focused balloon
angioplasty. The catheter consists of an integrated
wire outside a semi-compliant dilating balloon
and a short monorail with the guidewire lumen lo-
cated distal to the balloon. This unique dual wire
design allows balloon inflation against the stan-
dard coronary guidewire and the integrated exter-
nal wire, which provides concentrated
longitudinal scoring of atheromatous plaque
along the two wires and creates an expansion
plane at low inflation pressures, which facilitates
dilatation of resistant and elastic lesions. The
safety and efficacy of stand-alone FXminiRAIL
PTCA, including its use in mild-to-moderate calci-
fied lesions, have been reported. The efficacy of
this technique of plaque modification before stent-
ing is being evaluated in the PreFX Registry, and
the preliminary results show a trend toward some-
what better stent expansion after the FXmini-
RAIL PTCA [23].
Fig. 6. FXminiRAIL catheter.
FXminiRAIL PTCA atherotomy may be in-
dicated in ostial and undilatable chronic lesions
before stenting. By virtue of its design, FXmini-
RAIL is more flexible and deliverable than the CB
and can be advanced easily into angulated and less
accessible lesions.
Excimer laser coronary angioplasty
First applications of laser technology to the
cardiovascular system were in the 1980s. Excimer
laser has been used before stenting for aorto-ostial
and vein graft lesion with high procedural success
(O90%) but no reduction in restenosis and high
procedural device and equipment cost. Currently,
excimer laser coronary angioplasty has a limited
use in lesion modification before stenting.
Other innovative plaque modification devices
Other innovative plaque excision devices, such
as the SilverHawk System (Fox Hollow Technolo-
gies, Inc., Redwood City, California), AngioSculpt
scoring balloon (Angioscore, Inc., Alameda,
 CORONARY INTERVENTIONAL DEVICES
California), orbital atherectomy (Cardiovascular
Systems, Inc., Santa Clara, California), and Car-
dio-Path (Pathway Medical Technologies, Inc.,
Redmond, Washington), are being studied in
complex lesions with favorable outcomes. Final
results of these trials are awaited before any definite
conclusion can be drawn about their effectiveness.
Thrombectomy devices
In recent years research has demonstrated that
restoration of normal coronary flow in the infarct-
related artery is not necessarily equivalent to the
restoration of normal myocardial perfusion
through the coronary microcirculation. After
conventional primary PCI for ST elevation myo-
cardial infarction (MI) with stent implantation
and IIb/IIIa blockade, the normal myocardial
perfusion expressed on angiography by the tissue
myocardial perfusion grade three is seen in only
one third of patients. In the other two thirds of
cases, impaired microcirculatory perfusion is ob-
served (tissue myocardial perfusion grade 2-0),
accompanied by only partial (30%–70%) or no
resolution (!30%) of ST segment elevation in
electrocardiography.
Complete resolution of ST segment elevation
(O70%) in resting electrocardiography is recog-
nized as a good indicator of restoration of normal
myocardial perfusion. Accordingly, patients with
impaired microcirculation have increased early
and late mortality, greater irreversible myocardial
injury, and, consequently, higher incidence of
adverse remodeling of the left ventricle, with
higher rates of mortality, congestive heart failure,
and arrhythmia. One of the main causes of
inadequate myocardial reperfusion despite resto-
ration of epicardial flow in the infarct-related
artery is embolization of distal artery, side
branches, or microcirculation by embolic material
that consists of fragmented thrombus, platelet/
platelet–leukocyte aggregates, and fragmented
plaque released in the course of fibrinolytic
therapy or primary PCI. Other reasons include
increased microcirculatory resistance caused by
neutrophil obstruction, arteriolar constriction,
capillary necrosis, progressive myocardial dam-
age, and edema after prolonged ischemia or as
a result of reperfusion injury. In extreme cases
these phenomena may lead to abrogation of
normal rates of epicardial flow despite removal
of mechanical obstruction in the infarct-related
artery (no reflow) and subsequent attended de-
scribed adverse sequelae. Thrombus removal may
207
ameliorate the rate of these adverse sequelae
(Fig. 7). Use of glycoprotein (GP) IIb/IIIa inhibi-
tor before PCI has also shown to reduce thrombus
burden and subsequent complications [24].
Currently, there are several systems for percu-
taneous intracoronary thrombectomy. The most
widely studied systems are the AngioJet (Possis
Medical, Minneapolis, Minnesota) and X-Sizer
(eV3) devices. The transluminal extraction cathe-
ter is no longer used. Recently, several newer
intracoronary thrombectomy devices have been
introduced, such as the Export/Transport Cathe-
ter (Medtronic, Minneapolis, Minnesota), Diver
CE (Invatec Inc., eV3, Plymouth, Minnesota),
Pronto (Vascular Solutions, Inc., Minneapolis,
Minnesota), Reinspiration (Kerberos Proximal
Solutions Inc., Cupertino, California), and simple
standard guiding catheters. All these systems
differ considerably in construction and principles
of operation.
Several small studies have shown improved
outcomes in patients who have acute myocardial
infarction (AMI) and large thrombus burdens
treated with thrombectomy during primary PCI.
These initial results in small patient populations
were encouraging but so far have not been
confirmed in large randomized studies.
Rheolytic thrombectomy with the AngioJet
(Possis Medical) is an effective method for
removing thrombus by applying the Venturi-
Bernoulli vacuum principle. The device (Fig. 8)
is a 5-F double lumen, highly flexible catheter
that uses a 0.014-in guidewire. Six high-speed
saline jets create a low-pressure region at the tip
(approximately
760 mm Hg), which acts to
pull thrombus and extract it from the vessel. In
approaching a lesion with a thrombotic filling de-
fect, the AngioJet device is advanced proximal to
Fig. 7. Coronary intervention of thrombotic lesions:
pathophysiology of no-reflow and treatments.
208 SHARMA & CHEN
Fig. 8. AngioJet thrombectomy catheter.
the lesion and then activated and slowly advanced
at approximately 0.5 to 1 mm per second. Repeat
passes are performed with angiography after each
pass until there is no further improvement in the
angiographic appearance of the thrombotic lesion
(three to five passes on average). Stent implanta-
tion can be performed safely when most of the
thrombus has been removed. The device has
proved effective in thrombus-containing lesions
in the peripheral and coronary circulation and re-
ceived US Food and Drug Administration ap-
proval after showing superiority to urokinase
infusion in the VEGAS II Trial [25].
Small studies using AngioJet in AMI have
been encouraging. A recently completed multi-
center randomized trial of AngioJet Rheolytic
Thrombectomy in Patients Undergoing Primary
Angioplasty for Acute Myocardial Infarction
(AIMI) failed to show any benefit of AngioJet
compared with conventional treatment (Fig. 9)
[26]. This failure has put the prospects of AngioJet
in the treatment of patients who have AMI in se-
vere doubt; however, the serious limitations of the
Fig. 9. AngioJet catheter thrombectomy in patients un-
dergoing primary angioplasty for AMI.
study have to be recognized, most importantly,
absent or minimal thrombus reported in approxi-
mately 25% of patients. Baseline TIMI flow-3 was
statistically more frequently observed in the con-
trol group (27% compared with 19%; P ! 0.05).
Taking this into consideration, it seems that fur-
ther studies are necessary to finally elucidate the
use of Angiojet during primary PCI for AMI
with large thrombus burden.
We analyzed our outcomes with or without the
AngioJet thrombectomy catheter during primary
PCI in patients who have AMI with high-grade
thrombus (Rgrade 3, as per TIMI classification)
(Box 1) [27]. Ninety-five consecutive patients who
had AMI and large thrombotic burden who un-
derwent primary stenting with AngioJet (n ¼ 52)
and without AngioJet (n ¼ 43) were analyzed
for the epicardial and microvascular flow and
correlated with 30-day MACE and 1-year
Box 1. Thrombus grades: TIMI thrombus
grade
 Grade 0: No thrombus
 Grade 1: Possible thrombus (mural
opacities)
 Grade 2: Small thrombus (<0.5l normal
lumen diameter)
 Grade 3: Medium thrombus (0.5–1.5l
normal lumen diameter)
 Grade 4: Large thrombus (>1.5l normal
lumen diameter)
 Grade 5: Recent thrombotic occlusion
(fresh thrombus with dye stasis and
delayed washout)
 Grade 6: Chronic total occlusion
(smooth, abrupt, and with no dye
stasis and brisk flow)
 CORONARY INTERVENTIONAL DEVICES 209

survival. Results showed that use of the AngioJet Table 1

thrombectomy catheter during primary stenting AngioJet thrombectomy catheter during primary stent-
of patients who have AMI with high-grade throm- ing of patients with high-grade thrombus
bus resulted in better epicardial and microvascular Non-
flow, which resulted in improved short- and long- AngioJet AngioJet
term outcomes (Table 1). Our findings are in con- (n ¼ 52) (n ¼ 43) P
trast with the lack of benefit of AngioJet use in the Age (y) 62 G 14 66 G 15 NS
AIMI trial, perhaps because of the different exclu- Anterior MI (%) 65 70 NS
sion criteria (in the AIMI trial, all patients were Pain-to-door time (hr) 14 G 3 12 G 4 NS
included, regardless of the thrombus grade, Door-to-balloon 183 G 32 169 G 62 NS
whereas our trial included only patients with time (min)
high-grade thrombus). LVEF (%) 44 G 10 42 G 10 NS
TIMI before PCI (%) 0.66 G 0.98 0.93 G 1.14 0.24
It has been demonstrated that thrombectomy
Thrombus grade (%) 4.6 4.3 NS
with X-Sizer (Fig. 10) before stent implantation TIMI after PCI (%) 2.7 2.2 0.007
during primary PCI for AMI effectively decreases TMP grade (mean) 2.16 G 10 1.53 G 1.1 0.006
thrombus mass in the culprit lesion, which allows TMP grade 2/3 (%) 66 43 0.01
restoration of TIMI- 3 flow in a large proportion CTFC (mean) 22 G 13 28 G 11 0.02
of patients and prevents slow-flow, no-reflow, and 30-day MACE (%) 6 11 0.12
distal embolization, as measured by improved 1-year mortality (%) 7.4 14.6 0.08
myocardial perfusion by angiography and im- Event-free survival 86 74 0.05
proved ST segment elevation resolution at 60 min- at 1 year (%)
utes after PCI. The system’s helix cutter is housed Abbreviations: CTFC, corrected TIMI frame count;
within an atraumatic catheter tip. Fully assembled LVEF, left ventricle ejection fraction; TMP, TIMI myo-
and completely disposable, the X-Sizer catheter cardial perfusion; TIMI, thrombolysis in myocardial
system can be set up, used, and disposed of in infarction.
a matter of minutes (Fig. 11). The X-Sizer catheter
combines vacuum technology with a patented he- catheter decreases its ability to cope with tortu-
lix cutter housed in the tip of a small catheter. osity and excessive calcification proximal to the
When engaged, the system creates a powerful vac- culprit lesion. Its large profile prevents crossing of
uum effect designed to capture and remove occlu- tight lesions and limits its use with arteries with
sive material. The Archimedes screw is designed to reference diameter O2.5 mm. The presence of the
grab thrombus on contact and quickly draw it in, cutting blade in conjunction with difficult anat-
shearing and removing it. A randomized trial, the omy increases the risk of vessel perforation.
X-Tract trial, which compared X-sizer and con- Numerous old and emerging thrombectomy
ventional treatment, failed to show any difference devices work mainly on the principle of aspira-
in 30-day MACE but did show reduced large MI tion: Transluminal Extraction Catheter (Interven-
(Table 2). tional Therapeutics, Minneapolis, Minnesota),
Technical limitations of X-Sizer must be ac- Rescue (Boston Scientific Corp.), Export/Trans-
knowledged. The significant rigidity of the port Catheter (Medtronic), Pronto Extraction

Fig. 10. AngioJet thrombectomy for thrombotic LCx lesion in AMI. LCx, left circumflex.
210 SHARMA & CHEN
Fig. 11. The X-Sizer catheter.
Catheter (Vascular Solutions, Inc.), Diver CE
(Invatec Inc.), and Reinspiration (Kerberos Prox-
imal Solutions Inc.). Trials involving these aspi-
ration catheters are small, with variable success,
but they may play a role in the catheterization
laboratory because of ease of use and deliverabil-
ity, reduced bulkiness, and fast setup (Tables 2
and 3).
Distal protection devices
Percutaneous intervention of diseased saphe-
nous vein grafts (SVGs) is known to be associated
with a high periprocedural rate of complications
[28]. Distal protection systems are based on either
occlusive distal balloons (PercuSurge GuardWire)
or filter-based devices. Two distal protection de-
vices (GuardWire and Filter WireEX) have
proved to be clinically beneficial for PCI of SVG
lesions. The hope that distal protection may im-
prove results of PCI of thrombotic lesions is based
on the fact that these systems provide protection
from distal embolization during each balloon
inflation and stent implantation, while thrombec-
tomy is performed only before stent implantation.
Distal protection during SVG PCI using a balloon
occlusion and aspiration system (the PercuSurge
GuardWire) has been shown in a large random-
ized trial to reduce 30-day MACE rates by 42%
(Saphenous vein graft Angioplasty Free of Emboli
Randomized, SAFER Trial) (Table 2) [29]. Com-
pared with balloon occlusion systems, filter-based
distal protection devices may be simpler to use
and allow antegrade perfusion during the proce-
dure, which reduces ischemia time and facilitates
intervention in patients with poor left ventricular
function. The absolute clinical efficacy of distal fil-
ters as an adjunct to PCI for diseased SVGs has
not been proven (Fig. 12). The relative efficacy
of two distal protection devices was compared in
a randomized trial (see Table 2). The FIRE trial
was a large, multicenter trial of PCI in diseased
SVGs in which 650 consecutive patients at 65
US and Canadian centers were randomized 1:1
to intervention with distal protection using the
BSC/EPI FilterWire EX versus the PercuSurge
GuardWire. The FilterWire consists of a micropo-
rous polyurethane net attached to a self-expand-
ing nitinol ring anchored distally to a 0.014-in
guidewire over which PCI is performed. Random-
ization was stratified by use of IIb/IIIa inhibitors.
The primary endpoint was the composite rate of
death, mL (CPK-MB 3 normal values), coro-
nary artery bypass graft, or target lesion revascu-
larization at 30 days. The mean graft age was 11
G 8 years. The trial showed equal efficacy of
both distal protection devices [30].
Promising results with PercuSurge GuardWire
or Filterwire in SVG have not yet been confirmed
in the larger randomized studies of AMI, such as
the Enhanced Myocardial Efficacy and Removal
by Aspiration of Liberated Debris (EMERALD)
study. Despite the fact that thrombotic and
plaque debris were found in aspirates of 76% of
patients, no differences were found between stud-
ied groups in angiographically assessed myocar-
dial reperfusion, ST segment elevation resolution,
or infarct size measured by isotope scan at 30
days. The results of the EMERALD trial have
impeached seriously the concept of mechanical
cardioprotection of the microcirculation during
primary PCI for AMI [31]. The PercuSurge
Table 2
Completed trails using embolic protection in coronary intervention
Clinical Management Results (intervention
Trial name syndrome Device No. patients strategy Other agents used Endpoint vs. control)
SAFE Elective SVG GuardWire 105 Registry In-hospital MACE 5% 99%
intervention final TIMI
(low thrombus grade 3 flow
burden)

CORONARY INTERVENTIONAL DEVICES

SAFER Elective SVG GuardWire vs. no 801 Prospective, GP IIb/IIIa inhibitor 30-day MACE 9.6% 16.5% P ¼ 0.004
intervention distal protection randomized trial in approximately 58%
FilterWire Elective SVG Filter Wire 60 (phase 1) 248 Registry GP IIb/IIIa inhibitor 30-day MACE 21% (phase 1)
during SVG intervention (phase 2) in 30%–50% 11% (phase 2)
stenting
FIRE Elective SVG Filter Wire vs. 651 Prospective, GP IIb/IIIa inhibitor 30-day MACE 10% vs. 12%
intervention GuardWire randomized trial in approximately
(low thrombus 52% lesions
burden)
X-TRACT SVG (70%) or X-Sizer vs. 50 (phase 1) 797 Prospective, GP IIb/IIIa inhibitor 30-day MACE, 17% vs. 17%, P ¼ NS
thrombus-rich no DPD (phase 2) randomized trial in approximately Large AMI 5% vs. 10%, P ¼ 0.002
native vessel 76% lesions in (CKMB O8X
(30%) both arms ULN)
PRIDE SVG TriActive vs. Filter 631 Prospective, GP IIb IIIa inhibitor 30-day MACE 11% vs. 12%, P ¼ NS
Wire/GuardWire randomized trial in approximately 54%
in both arms
CAPTIVE SVG Cardio Shield vs. 652 Prospective, Uncertain, pending 30-day MACE 10% vs. 12%, P ¼ NS
GuardWire randomized trial full trial publication
Abbreviations: CKMB, creatine kinase isoenzyme MB; DPD, distales protection device; SVG, saphenous vein graft; ULN, upper limit of normal.

211
 212
Table 3
Ongoing trials using embolic protection in coronary intervention
Clinical Management Results (intervention
Trial name syndrome Device No. patients strategy Other agents used Endpoint vs. control)
PROXIMAL SVG Proximal Proxis 600 Randomized to Proxis or 30-day MACE Completed equal
intervention protection device other DPD efficacy
DEAR-MI AMI Thrombectomy Pronto 200 Randomized to Pronto ST-segment Ongoing

SHARMA & CHEN

thrombectomy vs. resolution LV
no treatment function recovery
RULE-SVG PCI to SVG Filter Rubicon 60 Randomized to Rubicon 30-day MACE Ongoing
vs. no protection
SPIDER PCI to SVG Filter SpideRX 770 Randomized to 30-day MACE Ongoing
SpideRX or GuardWire
GUARD PCI to SVG Filter Angioguard 800 Randomized to AngioGuard 30-day MACE Ongoing
or GuardWire
AMEthyst PCI to SVG Filter Interceptor 600 Randomized to Interceptor 30-day MACE Ongoing
or GuardWire
Abbreviation: DPD, distales protection device.
 CORONARY INTERVENTIONAL DEVICES 213

Fig. 12. Filterwire in SVG to left circumflex (LCx) lesion.

GuardWire system has important limitations. As-
piration of thromboembolic material is performed
via an ordinary perfusion catheter without the
thrombus fragmentation, which is in contradic-
tion to the AngioJet or X-Sizer. This approach
may result in the inability to remove large frag-
ments of thrombotic debris. The GuardWire bal-
loon is inflated 3 to 5cm beyond the occlusion
site and as such has no ability to prevent distal
embolization of side branches that originate in
between.
Two new combined embolic protection and
thrombectomy devices have been introduced for
SVG interventions (see Tables 2 and 3). The
Proxis-device (Velocimed, Minneapolis, Minne-
sota) consists of a short, flexible catheter (internal
diameter, 0.058 in) attached to a hypertube-catheter
shaft with a short circumferential balloon at the
distal tip. The device is introduced through an 8-F
guiding catheter and advanced in the proximal
part of the occluded artery. The balloon at the tip
of the device is inflated at 2 atm; wire crossing of
the coronary occlusion, balloon dilatation, and
stent placement are performed through the device
under total proximal blockade of the vessel. After
withdrawal of dilatation or stent balloons, aspira-
tion of debris is performed and continuses when
coronary flow is restored by deflation of the device
balloon. Temporary proximal vessel occlusion and
aspiration are repeated during each step of the PCI
procedure. The preliminary results suggest that the
device is effective for aspiration of embolic
elements.
Another distal protection device based on
balloon occlusion and aspiration, the TriActive
System, was evaluated in the Protection During
Saphenous Vein Graft Intervention to Prevent
Distal Embolization (PRIDE) study against other
distal protection devices. The PRIDE study
compared outcomes with the TriActive System
(Kensey Nash Corp., Exton, Pennsylvania), a bal-
loon-protection flush and extraction device, with
an embolic protection group during treatment of
SVGs. The incidence of major adverse cardiac
events at 30 days was 11.2% for the TriActive
group and 10.1% for the control group (P ¼
0.65; P ¼ 0.02 for noninferiority). Slightly higher
vascular complications were reported in the Tri-
Active group [32].
Summary
Percutaneous treatment of thrombus-contain-
ing lesions is associated with higher complication
rates compared with nonthrombotic lesions. Ad-
junct devices, such as thrombectomy or distal
protection, are commonly used as part of the
interventional procedure along with the liberal use
of GP IIb/IIIa inhibitors or vasodilators (Fig. 13).
Percutaneous treatment of SVG using distal pro-
tection (PercuSurge or filter devices) is routinely
recommended with stent implantation to improve
short- and long-term results. Despite achieving
TIMI-3 flow, myocardial perfusion in AMI
Fig. 13. Treatment options for thrombotic lesions (SVG
or native).
214 SHARMA & CHEN
remains suboptimal in a significant number of
patients, resulting in larger final infarct size. Ef-
fective removal of thrombi before stenting
theoretically may reduce distal embolization of
thrombus, which could improve myocardial per-
fusion and salvage. Randomized studies do not
support routine use of thrombectomy devices or
distal protection devices with primary PCI in all
patients who have AMI for the reduction of major
adverse cardiac events. Simple manual aspiration
with easy-to-use catheters to extract atherothrom-
botic material from target lesions, which restores
flow and increases patency rate of an infarct-re-
lated artery before stenting, may be the hypothet-
ical option for selected patients during primary
PCI.
References
[1] Moses JW, Leon MB, Popma JJ, et al. Sirolimus-
eluting stents versus standard stents in patients
with stenosis in a native coronary artery. N Engl
J Med 2003;349:1315–23.
[2] Stone GW, Ellis SG, Cox D, et al. A polymerbased,
paclitaxel-eluting stent in patients with coronary ar-
tery disease. N Engl J Med 2004;350:221–31.
[3] Fujii K, Mintz GS, Kobayashi Y, et al. Contribution
of stent underexpansion to recurrence after siroli-
mus-eluting stent implantation for in-stent resteno-
sis. Circulation 2004;109:1085–8.
[4] Takebayashi H, Mintz G, Carlier S, et al. Nonuni-
form strut distribution correlates with more neointi-
mal hyperplasia after sirolimus-eluting stent
implantation. Circulation 2004;110:3430–4.
[5] Farb A, Sangiorgi G, Carter A, et al. Pathology of
acute and chronic coronary stenting in humans. Cir-
culation 1999;99:44–52.
[6] Takeda Y, Tsuchikane E, Kobayashi T, et al. Effect
of plaque debulking before stent implantation on in-
stent neointimal proliferation: a serial 3-dimensional
intravascular ultrasound study. Am Heart J 2003;
146:175–82.
[7] Prati F, Mario C, Moussa I, et al. In-stent neointi-
mal proliferation correlates with the amount of re-
sidual plaque burden outside the stent: an
intravascular ultrasound study. Circulation 1999;
99:1011–4.
[8] Moussa I, Moses J, Di Mario C, et al. Stenting after
optimal lesion debulking (SOLD) registry: angio-
graphic and clinical outcome. Circulation 1998;98:
1604–9.
[9] Stankovic G, Colombo A, Bersin R, et al. Compar-
ison of directional coronary atherectomy and stent-
ing versus stenting alone for the treatment of de
novo and restenotic coronary artery narrowing.
Am J Cardiol 2004;93:953–8.
[10] Reifart N, Vandormael M, Krajcar M, et al. Ran-
domized comparison of angioplasty of complex cor-
onary lesions at a single center: excimer laser,
rotational atherectomy, and balloon angioplasty
comparison (ERBAC) study. Circulation 1997;96:
91–8.
[11] Kini A, Marmur J, Duvvuri S, et al. Rotational athe-
rectomy: improved procedural outcome with evolu-
tion of technique and equipment. Single-center
results of first 1,000 patients. Catheter Cardiovasc
Interv 1999;46:305–11.
[12] Chung CM, Nakamura S, Tanaka K, et al. Stenting
alone versus debulking and debulking plus stent in
branch ostial lesions of native coronary arteries.
Heart Vessels 2004;19:213–20.
[13] Cavusoglu E, Kini A, Marmur J, et al. Current sta-
tus of rotational atherectomy. Catheter Cardiovasc
Interv 2004;62:485–98.
[14] Kini A, Kim M, Das S, et al. Efficacy of drug-eluting
stents in calcified lesions. J Am Coll Cardiol 2005;
45(Suppl A):65A.
[15] Dunn B. The effects of debulking on restenosis
(EDRES). J Saudi Heart Assn 1998;10:55.
[16] Buchbinder M, Fortuna R, Sharma S, et al. Debulk-
ing prior to stenting improves acute outcomes: early
results from the SPORT trial. J Am Coll Cardiol
2000;35(Suppl A):8A.
[17] Tsuchikane E, Suzuki T, Asakura Y, et al. Debulk-
ing of chronic total occlusions with rotational or
directional atherectomy before stenting trial: a multi-
center randomized study [abstract]. J Am Coll Car-
diol 2004;43:59A.
[18] Sharma SK. Simultaneous kissing drug-eluting stent
technique for percutaneous treatment of bifurcation
lesions in large-size vessels. Catheter Cardiovasc
Interv 2005;65:10–6.
[19] Tan R, Kini A, Shalouh E, et al. Optimal treatment
of nonaorto ostial coronary lesions in large vessels:
acute and long-term results. Catheter Cardiovasc
Interv 2001;54:283–8.
[20] Hara H, Nakamura M, Asahara T, et al. Intravascu-
lar ultrasonic comparison of mechanisms of vasodi-
latation of cutting balloon angioplasty versus
conventional balloon angioplasty. Am J Cardiol
2002;89:1253–6.
[21] Mauri L, Bonan R, Weiner BH, et al. Cutting bal-
loon angioplasty for the prevention of restenosis: re-
sults of the cutting balloon global randomized trial.
Am J Cardiol 2002;90:1079–83.
[22] Ozaki Y, Suzuki T, Yamaguchi T, et al. Can intra-
vascular ultrasound guided cutting balloon angio-
plasty before stenting be a substitute for drug
eluting stent? Final results of the prospective
randomized multicenter trial comparing cutting bal-
loon with balloon angioplasty before stenting
(Reduce III) [abstract]. J Am Coll Cardiol 2004;
43:82A.
[23] Ischinger TA, Solar RJ, Hitzke E. Improved out-
come with novel device for low-pressure PTCA in
de novo and in-stent lesions. Cardiovasc Radiat
Med 2003;4:2–6.
 CORONARY INTERVENTIONAL DEVICES
[24] Zhao XQ, Théroux P, Snapinn SM, et al. Intracoro-
nary thrombus and platelet glycoprotein IIb/IIIa re-
ceptor blockade with tirofiban in unstable angina or
non–Q-wave myocardial infarction: angiographic
results from the PRISM-PLUS trial (platelet recep-
tor inhibition for ischemic syndrome management
in patients limited by unstable signs and symptoms).
Circulation 1999;100:1609–15.
[25] Kuntz R, Baim D, Cohen D, et al. A trial comparing
rheolytic thrombectomy with intracoronary uroki-
nase for coronary and vein graft thrombus (the
Vein Graft AngioJet Study [VeGAS 2]). Am J
Cardiol 2002;89:326–30.
[26] Ali A, on behalf of the AIMI Investigators. AngioJet
rheolytic thrombectomy in patients undergoing pri-
mary angioplasty for acute myocardial infarction.
Presented at the Transcatheter Cardiovascular Ther-
apeutics conference. Washington, DC, September
2004.
[27] Tamburrino F, Kini A, Gupta S, et al. The improved
outcome with AngioJet thrombectomy catheter dur-
ing primary stenting in acute myocardial infarction
patients with high-grade thrombus. Am J Cardiol
2005;96(Suppl 7A):76H.
215
[28] Hong MK, Mehran R, Dangas G, et al. Creatine ki-
nase-MB enzyme elevation following successful sa-
phenous vein graft intervention is associated with
late mortality. Circulation 1999;100:2400–5.
[29] Baim DS, Wahr D, George B, et al. Randomized
trial of a distal embolic protection device during per-
cutaneous intervention of saphenous vein aorto-cor-
onary bypass grafts. Circulation 2002;105:1285–90.
[30] Stone G, Rogers C, Hermiller J, et al. Randomized
comparison of distal protection with a filter-based
catheter and a balloon occlusion and aspiration sys-
tem during percutaneous intervention of diseased
saphenous vein aorto-coronary bypass grafts.
Circulation 2003;108:548–53.
[31] Stone GW, Webb J, Cox DA, et al. Distal microcir-
culatory protection during percutaneous coronary
intervention in acute ST-segment elevation myocar-
dial infarction: a randomized controlled trial.
JAMA 2005;293:1063–72.
[32] Carrozza J, Mumma M, Breall J, et al. Randomized
evaluation if the TriActiv balloon-protection flush
and extraction system for the treatment of saphe-
nous vein graft disease. J Am Coll Cardiol 2005;46:
1677–83.
 Cardiol Clin 24 (2006) 217–231
Drug-Eluting Stents
Jeffrey J. Popma, MD*, Mark Tulli, MD
Department of Internal Medicine (Cardiovascular Division), Brigham and Women’s Hospital,
75 Francis Street, Boston, MA 02115, USA
Percutaneous coronary intervention (PCI) has
evolved dramatically over the past 25 years.
Stand-alone balloon angioplasty has been re-
placed with the use of coronary stents because
of the near elimination of emergency coronary
artery bypass surgery (CABG) [1] and marked re-
ductions in restenosis associated with the use of
coronary stents [2,3]. With the introduction of
dozens of balloon-expandable stents during the
mid- and late 1990s, progressive improvements
in the crossing profile, deliverability, and metallic
composition occurred, albeit with little change in
the overall occurrence of stent restenosis [4–6].
Clinical restenosis associated with stent use was
particularly frequent (up to 30%) in patients
who had small vessels, long lesions, and diabetes
mellitus [7].
A breakthrough occurred in early 2000 with
the development of stents that eluted pharmacol-
ogy agents directly into the vessel wall by means
of a controlled release from a durable polymer
coating. Various drug-eluting stents (DES) were
developed, each varying with its delivery plat-
form, polymer coating (or absence of coating),
and drug selected for elution. This article de-
scribes the clinically available and late develop-
mental drug-eluting stent programs targeted for
treating patients who have coronary artery
disease.
Sirolimus-eluting stents
The CYPHER stent (Cordis Corporation,
Miami Lakes, Florida) is composed of balloon-
expandable stainless steel, a durable copolymer
* Corresponding author.
E-mail address: jpopma@partners.org (J.J. Popma).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.001
mixture of polyethylene–covinyl acetate (PEVA)
and poly-n-butyl methacrylate (PBMA), and a si-
rolimus, which is a G1 cell cycle inhibitor that also
has potent anti-inflammatory properties. Using
the commercially available, slow-release polymer
formulation, 90% of the sirolimus contained on
the stent is released within 30 days [8]. The
CYPHER stent was approved for clinical use in
Europe in April 2002 and in the United States in
May 2003. This stent is used in approximately
50% of DES procedures worldwide.
First in man studies
Much has been learned about the late-term (4-
year) safety and efficacy of the CYPHER stent
from the first 45 patients with focal native vessel
disease who were treated with slow-release (re-
lease over 30 days) and fast-release (release over
7 to 10 days) sirolimus-eluting stents (SES) [9–
11]. Serial evaluation at 4 months [9], 1 year
[12], 2 years [13], and 4 years [11] showed a min-
imal degree (less than 0.10 mm) of intimal hyper-
plasia within the stent at each of these follow-up
intervals, and no late untoward toxicities, such as
acquired late aneurysms. Although the target ves-
sel revascularization (TLR) rate for the entire co-
hort was 10% at 2 years [13], there were no
TLRs or stent thromboses (STs) that developed
between 2- and 4-year follow-up [11]. In-stent
late lumen loss was slightly greater at 4 years in
the fast-release group (0.41 G 0.49 mm) than
in the slow-release group (0.09 G 0.23 mm). Sim-
ilarly, intravascular ultrasound showed a slightly
higher degree of volume obstruction in the fast-
release group (9.1%) compared with the slow-re-
lease group (5.7%) [11], suggesting that release
kinetics may play an important role in the efficacy
of SES.
cardiology.theclinics.com
218 POPMA & TULLI
RAVEL
The RAVEL trial was a randomized study of
238 patients who had single, focal lesions located
in larger native coronary arteries and who were
assigned to treatment with SES or a bare metal
stent (BMS). It was the first controlled study to
show a marked reduction in clinical and angio-
graphic restenosis in patients treated with SES
[14]. There were no cases of binary angiographic
restenosis in patients treated with SES compared
with a 26.6% incidence in patients assigned to
treatment with BMS (P ! .001) [15]. One analysis
suggested that early and late side branch patency
improved with the use of SES [16]. Volumetric in-
travascular ultrasound analysis found a 90% re-
duction in intimal hyperplasia in those treated
with SES [17]. Clinical benefits associated with
the use of SES were sustained up to 3 years later,
with 93.7% of patients assigned to treatment with
SES being free from TLR compared with 75.0%
in patients assigned to treatment with BMS
(P ! .001) [18]. Three-year major adverse cardiac
events (MACE) developed in 15.8% of patients in
the SES group and in 33.1% of patients in the
BMS group (P ¼ .002) [18].
SIRIUS
The SIRIUS trial was a randomized compar-
ison of 1058 patients who had complex coronary
artery disease that included lesion lengths between
15 and 30 mm and vessel diameters between
2.5 mm and 3.5 mm and who were assigned
to treatment with SES or BMS [8]. The overall
complexity of patients enrolled in the study was
confirmed by the frequent presence of diabetes
mellitus (26%), long lesions (lesion length
14.4 mm), and smaller vessel size (mean diame-
ter 2.80 mm). The primary end point was target vessel
failure, defined as a composite of death from car-
diac causes, myocardial infarction (MI), or target
vessel revascularization within 270 days. These
events were reduced from 21.0% in patients as-
signed to BMS compared with 8.6% in those as-
signed to SES (P ! .001). The need for TLR
was reduced from 16.6% in those assigned to
treatment with BMS compared with 4.1% in those
assigned to SES (P ! .001). These benefits were
consistent across all patient and lesion subsets.
Although there were no differences in the occur-
rence of death or MI, at 1-year follow-up, the ab-
solute difference in TLR continued to increase
(20% in patients assigned to treatment with
BMS and 4.9% in patients assigned to treatment
with SES; P ! .001) [19].
The clinical benefit in this study was caused by
the profound reduction in intimal hyperplasia
within the SES compared with a BMS. Angio-
graphic restudy was obtained in 701 patients
enrolled in the SIRIUS trial 240 days after stent
placement [20]. Patients treated with SES had
lower rates of binary (greater than 50% follow-up
diameter stenosis) angiographic restenosis within
the segment (8.9% versus 36.3% with BMS;
P ! .001) and within the stent (3.2% versus
35.4% with BMS, P ! .001). SES had less late
lumen loss within the treated segment (0.24 G
0.47 mm versus 0.81 G 0.67 with BMS, P !
.001), within the stent (0.17 mm G 0.44 versus
1.00 G 0.70 mm, respectively; P ! .001), and
within its 5 mm proximal (0.16 G 0.48 mm versus
0.32 G 0.63 mm, respectively; P ! .001) and distal
(0.04 G 0.24 mm versus 0.24 G 0.61 mm, respec-
tively; P ! .001) edges 8 months after stent place-
ment. The frequency of late aneurysms was
similar in the two groups.
E-SIRIUS
The European extension to SIRIUS, known as
E-SIRIUS, enrolled 352 patients who had single
native coronary lesions with a vessel diameter 2.5
to 3.0 mm and lesion length 15 to 32 mm. Patients
were assigned randomly to treatment with SES or
BMS [21]. Although the frequency of diabetes was
less in the E-SIRIUS study than in the SIRIUS
study, the mean vessel diameter was slightly
smaller (2.55 G 0.37 mm), and the mean lesion
length was longer (15.0 G 6.0 mm) than in pa-
tients enrolled in SIRIUS. Fewer patients assigned
to treatment with SES experienced a 9 month ma-
jor adverse cardiac event (MACE) (8.0% versus
22.6% in patients assigned to BMS, P ¼ .0002).
This was primarily because of a reduction in the
need for TLR (4.0% versus 20.9% in patients as-
signed to BMS, P ! .0001). The binary angio-
graphic restenosis rate 8 months later was
reduced significantly in the SES group (5.9% ver-
sus 42.3% in the BMS group; P ¼ .0001). Because
of meticulous attention to deployment technique,
including the avoidance of balloon inflations at
the stent margins, the occurrence of edge resteno-
sis was reduced in the E-SIRIUS study compared
with the SIRIUS study.
 DRUG-ELUTING STENTS
C-SIRIUS
The Canadian SIRIUS study enrolled 100 pa-
tients at eight Canadian sites with similar inclusion
criteria to the SIRIUS and E-SIRIUS studies [22].
Angiographic restenosis developed in one (2.3%) pa-
tient assigned to treatment with SES and in 23 (52%)
patients assigned to treatment with a BMS (52.3%,
P ! .001). Two (4%) patients required TLRs by
270 days in the SES group compared with 18% in
patients assigned to treatment with BMS (P ¼ .05).
Stent restenosis
Stent restenosis is an increasingly uncommon
cause of recurrent ischemia after stent placement,
owing to the more frequent use of DES in patients
undergoing PCI. Yet when stent restenosis does
occur, treatment options include the use of brachy-
therapy, angioplasty with conventional or cutting
balloons, or DES placement [23–25]. Clinical event
rates were low in a series of 25 patients with relatively
focal stent restenosis who were treated with one or
two SES [23], whereas patients with more complex
in-stent restenosis had higher events rates, often relat-
ing to thrombosis within the stent, despite a minimal
degree of intimal hyperplasia within the stent [26].
In the Sirolimus-eluting stent for In-Stent
Restenosis (SISR) trial, 384 patients were assigned
randomly to treatment with vascular brachyther-
apy using gamma or beta sources or to placement
of one or more SES [27]. Patients were included if
the stent restenosis occurred in vessels with a di-
ameter between 2.5 and 3.5 mm, and the stent
restenosis was between 15 and 40 mm in length.
The primary endpoint, target vessel failure, was
reduced from 21.6% in patients assigned to treat-
ment with brachytherapy to 12.4% in patients as-
signed to treatment with DES (P ¼ .023). At the
time of follow-up angiography, there was a trend
toward improvement in the analysis segment bi-
nary angiographic restenosis rate in patients
treated with the SES (19.5% versus 29.5% in the
brachytherapy treated patients, P ¼ .067). There
were two late stent thromboses in patients as-
signed to treatment with SES. These data suggest
that long-term dual antiplatelet therapy may be
needed to prevent late stent thrombosis.
SIRIUS registries and subset analyses
Acute coronary syndromes
As part of the Rapamycin-Eluting Stent Eval-
uated At Rotterdam Cardiology Hospital
219
(RESEARCH) registry, 198 patients presenting
with an acute coronary syndrome and treated with
a SES were compared with a control group of 301
patients with a similar presentation who were
treated with BMS [28]. The 30-day major adverse
cardiac event rate was similar in both groups
(6.1% in SES patients versus 6.6% in control pa-
tients, P ¼ .8), including the occurrence of stent
thrombosis (0.5% in SES patients versus 1.7%
in control patients, P ¼ .4). Using the same RE-
SEARCH registry, the late clinical outcomes in
patients with ST segment myocardial infarction
(STEMI) were compared in 186 consecutive pa-
tients treated with SES and 183 patients treated
with BMS [29]. Stent thrombosis did not occur
in any patient treated with the SES but occurred
in 1.6% of patients treated with BMS. Those
treated with SES had a reduced incidence of com-
bined adverse events at 300 days significantly
(9.4% compared with 17% in patients treated
with BMS, P ¼ .02), primarily because of a lower
risk of TLR in patients treated with SES (1.1%
compared with 8.2% in patients treated with
BMS, P ¼ .01) [29]. The data support the use of
SES in patients who present with an acute coro-
nary syndrome, including STEMI.
Diabetic patients
In the SIRIUS trial, the 279 patients who had
a history of diabetes mellitus were treated with
SES or BMS [30]. Late (270 day) TLR was re-
duced in diabetic patients from 22.3% with
BMS to 6.9% with SES (P ! .001), and late
MACE was reduced from 25% with BMS to
9.2% with SES (P ! .001). Although SES was
markedly effective in reducing in stent intimal hy-
perplasia, edge restenosis often occurred in these
diabetic patients, because of either balloon injury
not covered with SES or oversizing of the SES in
diffusely diseased vessels.
Small coronary arteries
In a multi-center, prospective trial, 257 pa-
tients were randomized to treatment with SES
or BMS for lesions in native coronary arteries
no more than 2.75 mm [31]. The mean vessel
diameter was 2.2 G 0.28 mm, and the lesion
length was 11.8 G 6.2 mm. The primary endpoint,
8-month binary in-segment restenosis rate, oc-
curred in 9.8% of patients in the SES group and
in 53.1% patients in the BMS group (risk reduc-
tion (RR), 0.18; P ! .001). Late MACE occurred
in 9.3% patients assigned to treatment with SES
220 POPMA & TULLI
and in 31.3% patients assigned to treatment with
BMS (RR, 0.30; P ! .001), primary because of re-
ductions in TLR (7% versus 21.1%, respectively;
RR, 0.33; P ¼ .002) and MI (1.6% versus 7.8%,
respectively; RR, 0.20; P ¼ .04).
Other patient and lesion subsets
The beneficial effect of SES over BMS has been
demonstrated in numerous complex subsets, in-
cluding patients with chronic total occlusion [32],
saphenous vein graft disease [33,34], left main
coronary artery disease [35–38], bifurcation
lesions [39–41], and diffuse disease requiring
four or more stents [42]. Direct stenting with SES
is also as effective as SES placement after
redilation [43].
Durable polymer, paclitaxel-eluting stents
The TAXUS stent (Boston Scientific, Natick,
Massachusetts) is comprised of a stainless steel
stent platform, a polyolefin polymer derivative,
and a microtubular stabilizing agent paclitaxel,
with two-phase 30-day polymeric release kinetics
that provide its antiproliferative effect [44]. Pacli-
taxel is an inhibitor of microtubules that prevents
cell division at the M phase. Paclitaxel release is
completed within 30 days of implantation, al-
though a substantial portion (greater than 90%)
of the paclitaxel remains within the polymer
indefinitely.
The TAXUS-I trial
The TAXUS-I Trial was a prospective, ran-
domized study of 61 patients with de novo or
restenotic lesions who received a paclitaxel-eluting
stent (PES) or BMS [45]. There were no cases of
angiographic restenosis 6 months after the proce-
dure in patients treated with PES, compared with
a 10% restenosis rate in patients treated with
a BMS. The low composite MACE rates reported
at 1-year follow-up (3.2% in patients treated with
PES compared with 10.0% in patients with BMS)
were maintained at 2 and 3 years, with no addi-
tional MACE in either treatment group 1 year
after implantation [46].
The TAXUS-II trial
TAXUS-II was a large randomized trial in 536
patients with native vessel coronary artery disease
who were assigned randomly to treatment with
BMS (NIR Express; Boston Scientific, Natick,
Massachusetts) or PES using either a slow (SR)-
or moderate release (MR)- polymeric coating
[47]. The primary study endpoint, 6-month in-
stent net percent volume obstruction measured
by intravascular ultrasound, was significantly
lower for PES (7.9% SR and 7.8% MR) than
for SES (23.2% and 20.5%, P ! .0001 for
both). Angiographic restenosis was reduced from
17.9% in the BMS cohort to 2.3% in the
PES-SR cohort (P ! .0001) and from 20.2% in
the BMS cohort to 4.7% in the PES-MR cohort
(P ¼ .0002). A serial intravascular ultrasound
study demonstrated that the marked reduction
in in-stent restenosis with SR or MR stents was
not associated with increased edge stenosis at
6-month follow-up IVUS [48]. Compared with
BMS, there was a reduction in late lumen loss at
the distal edge with TAXUS stents [48]. Twelve-
month MACE was also lower in the TAXUS-
SR (10.9%) and TAXUS-MR (9.9%) groups
than in controls (22.0% and 21.4%, respectively),
in large part because of a reduction in TLR in
patients who were treated with PES [47].
TAXUS-III
Twenty-eight patients who had complex in-
stent restenosis were treated with PES (NIR
platform) in the TAXUS-III study [49]. Although
no subacute stent thromboses occurred within the
first 12 months, the MACE rate was 29%, includ-
ing six patients who required TLR. The high
recurrence rate was attributable to the develop-
ment of angiographic restenosis occurring in re-
gions treated with a PES but also developing in
gaps segments between PES and in BMS that
were placed in as bailout to PES. These findings
underscore the need for meticulous technique for
the placement of DES for in-stent restenosis
with coverage of all regions of balloon injury
with PES.
TAXUS-IV
The TAXUS-IV trial randomly assigned 1314
patients with complex, native coronary stenoses to
treatment with BMS or SR PES [44]. Similar to
the SIRIUS trial, patients enrolled in the
TAXUS-IV study had complex lesions that were
between 10 and 28 mm in length and located in
vessels with a diameter between 2.5 and 3.75 mm.
The frequency of diabetes mellitus (24.2%), re-
ference vessel diameter (2.75 mm) mean lesion
length (13.4 mm), and stent length (21.8 mm)
 DRUG-ELUTING STENTS
were similar in the two treatment groups, and
comparable to the lesion characteristics in patients
enrolled in the SIRIUS trial. The primary study
endpoint, 9-month target vessel revascularization
(TVR), was reduced from 12.0% in patients as-
signed to BMS to 4.7% in patients assigned to PES
(P ! .001). TLR occurred in 11.3% of patients in
the BMS group and in 3.0% of patients in the PES
group (P ! .001). The 9-month composite rates of
death from cardiac causes or MI (4.7% and 4.3%, re-
spectively) and stent thrombosis (0.6% and 0.8%,
respectively) were similar in the group that received
a paclitaxel-eluting stent and the group that re-
ceived a bare metal stent. The reductions in
TLR and MACE were sustained 1 year after
the procedure [50].
Nine-month angiographic follow-up was per-
formed in 732 patients and demonstrated a reduc-
tion from 0.92 G 0.58 mm in patients treated with
BMS to 0.39 G 0.50 in patients treated with
a TAXUS stent (P ! .001). Binary angiographic
restenosis, defined as a greater than 50% diameter
follow-up stenosis, occurred in 26.6% of pa-
tients with BMS and 7.9% of patients with PES
(P ! .001). An IVUS substudy analysis in 170
patients found a uniform suppression of intimal
hyperplasia along the length of the stent, with re-
duction in neointimal volume in patients treated
with the TAXUS stent (18 G 18 mm3) compared
with patient treated with a BMS (41 G 23 mm3,
P ! .001) [51]. Incomplete stent apposition at
9 months was similar in patients treated with the
TAXUS stent (3.0%) and those treated with BMS
(4.0%, P ¼ .12).
TAXUS-V
The TAXUS-V trial was designed to evaluate
the angiographic and clinical outcomes of 1156
patients with very complex lesions who were
treated with SR PES or BMS [52]. Patients in
this study were required to have native coronary
stenoses in vessel diameters between 2.25 and
4.0 mm and lesion lengths between 10 and 46 mm
and multiple stents. The vessel diameter (2.69 G
0.57 mm), lesion lengths (17.2 G 9.2 mm), and
stent lengths per lesion (1.38 G 0.58 mm) indi-
cated a more complex subset of patients. The pri-
mary endpoint, 9-month TVR, was reduced from
17.3% in patients assigned to BMS to 12.1% in
those assigned to PES (P ! .001). The rates of
death and MI were similar in both groups. Angio-
graphic restenosis was reduced from 33.9% in those
treated with BMS to 18.9% in patients assigned to
221
treatment with PES (P ! .001). Patients treated
with 2.25 mm stents had a lower rate of angio-
graphic restenosis with PES (31.2%) than those
treated with BMS (49.4%, P ¼ .01). Although this
study demonstrated a reduction in events associated
with the use of PES over BMS, the overall higher
event rate in these patients suggested that further im-
provements in clinical outcomes could be obtained.
TAXUS-VI
The TAXUS-VI trial was designed to evaluate
clinical and angiographic outcomes in 448 pa-
tients with complex, long stenoses who were
assigned randomly to treatment with one or
more MR PES or BMS [53]. The primary end-
point of the study, 9-month TVR, was reduced
by 53% from 19.4% in patients assigned to
BMS to 9.1% of patients assigned PES (P ¼
.0027). In-stent angiographic restenosis also was
reduced from 32.9% in the patients assigned to
treatment with BMS to 9.1% in the patients as-
signed to treatment with PES (P ! .0001). TLR
was reduced from 18.9% in patients assigned to
treatment with BMS to 6.8% in patients assigned
to treatment with PES (P ¼ .0001).
TAXUS registries and subset analyses
The TAXUS stent is also effective in reducing
the angiographic and clinical recurrence in di-
abetic patients. Medically treated diabetes was
present in 318 (24%) of the 1314 patients enrolled
in the TAXUS-IV trial; of these, 105 patients
required insulin therapy [54]. Diabetic patients
treated with the PES stent had an 81% lower
rate of 9-month binary angiographic restenosis
(6.4%) than patients treated with BMS (34.5%,
P ! .0001). Diabetic patients also had a 65%
lower rate of 12-month TLR (7.4%) than patients
treated with a bare metal stent (20.9%, P ¼
.0008). Insulin-requiring diabetic patients assigned
to treatment with PES had an 82% lower angio-
graphic restenosis rate (7.7%) than patients as-
signed to treatment with BMS (42.9%, P ¼
.0065). Similarly, 1-year TLR rates were lowered
by 68% in insulin-requiring patients assigned to
treatment with a TAXUS stent (6.2% versus 19.4%
in BMS-assigned patients, P ¼ .07). There were
no differences in the occurrence of death or MI
in the two groups.
The TAXUS trials have demonstrated specific
benefits in women [55] and in patients who have
acute coronary syndromes [56], including ST
222 POPMA & TULLI
segment elevation MI [57,58], saphenous vein
graft disease [59], left anterior descending artery
stenoses [60], bifurcation lesions [61], chronic total
occlusion [62], and in those undergoing direct
stenting [63].
Randomized comparisons of sirolimus-eluting
stents and paclitaxel-eluting stent
A series of randomized studies has provided
head-to-head comparisons of the SES and PES in
various complex lesions subsets. Although pa-
tients in these trials were assigned randomly to
treatment with PES or SES, they have varied with
the frequency of angiographic follow-up, angio-
graphic methodology used for the analysis, and
the methods used for adjudication of clinical
events. Given the impact of the oculostenotic
reflex on the occurrence of late revascularizations,
studies that have had rigorous event adjudication
with sufficient sample sizes to justify the conclu-
sions take precedent over smaller ones with less
rigorously defined endpoints.
REALITY
The REALITY trial was a prospective, ran-
domized clinical trial performed at 90 hospitals in
Europe, Latin America, and Asia that included
1386 patients with one or two de novo lesions in
small (2.25 to 3.00 mm) caliper vessels who were
randomized to treatment with a SES or PES [64].
Diabetes mellitus was present in 28.0% of pa-
tients, The primary endpoint, 8-month in-lesion
binary restenosis, occurred in 9.6% of patients as-
signed to treatment with SES and 11.1% of pa-
tients assigned to treatment with PES (RR 0.84,
P ¼ .31), despite a lower in-stent late loss in pa-
tients assigned to treatment with SES (0.09 mm)
than in those assigned to treatment with PES
(0.31 mm, P ! .001). Patients treated with SES
and PES had similar 1-year frequencies of
MACE (10.7% in SES patients and 11.4% in
PES patients, P ¼ .73) and TLR (6.0% in SES
and 6.1% in PES, P O .99). The results of this
study suggested that the beneficial effects on inti-
mal hyperplasia observed with SES did not trans-
late into beneficial effects on clinical outcomes.
The frequency of binary (and clinical restenosis)
appears related to the standard deviation of the
late lumen loss measurement and the rightward
skewedness of the distribution histogram ra-
ther than to the absolute value of late lumen loss
alone [65].
SIRTAX
The SIRTAX trial, a trial of 1012 patients who
were assigned randomly to treatment with SES or
PES evaluated all comers to the catheterization
laboratory [66]. The primary study endpoint,
a composite of 9-month MACE, occurred in
6.2% of patients in the SES group and 10.8% of
patients in the PES group (hazard ratio [HR]
0.56, P ¼ .009), primarily because of reductions
in TLR in patients assigned to treatment with
SES (4.8% versus 8.3% in patients in the PES
group, P ¼ .03). There were no differences in
death, MI, or subacute stent thrombosis in the
two groups. The binary angiographic restenosis
rate in the 53.4% of patients with angiographic
follow-up was 6.6% in patients treated with SES
and 11.7% in patients treated with PES (P ¼ .02).
ISAR-Diabetes
In the ISAR-Diabetes trial, 250 diabetic pa-
tients were assigned randomly to treatment with
PES or SES. The primary study endpoint, in-
segment late luminal loss, was segment restenosis
(at least 50% diameter stenosis) occurred in
16.5% of the PES patients and in 6.9% of the
SES patients (P ¼ .03). TLR was required in
12.0% of the PES patients and in 6.4% of the
SES patients (P ¼ .13).
TAXI trial
In a smaller series, 202 patients were assigned
randomly to treatment with a PES or SES [67].
The incidence of late (mean 7 month) MACE
was 4% with the PES and 6% with the SES
(P ¼ .8). Similarly, the need for target lesion re-
vascularization was very low in both groups (1%
with PES and 3% with SES).
ISAR-DESIRE
The ISAR-DESIRE study was a study of the
effect of balloon angioplasty, SES, or PES in 300
patients with angiographically significant in-stent
restenosis [68]. The primary endpoint, 6-month in-
segment angiographic restenosis, was 44.6% in the
balloon angioplasty group, 14.3% in the SES
group (P ! .001 versus balloon angioplasty), and
21.7% in the PES group (P ¼ .001 versus balloon
angioplasty) [68]. The incidence of target vessel re-
vascularization was 33.0% in the balloon angio-
plasty group, 8.0% in the SES group (P ! .001
versus balloon angioplasty), and 19.0% in the
PES group (P ¼ .02 versus balloon angioplasty)
 DRUG-ELUTING STENTS
[68]. There was a trend for better outcomes in
patients treated with SES compared with PES.
Meta-analyses of comparative trials
In aggregate, many of the comparative studies
were too small to make definitive conclusions
about the comparative benefit of SES or PES
stents. A subsequent meta-analysis compared the
clinical and angiographic outcomes of SES and
PES in six randomized head-to-head clinical trials
that included 3669 patients [69]. TLR occurred
less often in patients treated with SES (5.1%)
compared with PES (7.8%) (P ¼ .001). Angio-
graphic restenosis occurred less often in patients
assigned to SES (9.3%) compared with PES
(13.1%) (P ¼ .001). Event rates were similar
with SES and PES for stent thrombosis (SES
0.9%, PES 1.1%, P ¼ .62), death (1.4% and
1.6%, respectively; P ¼ .56), and the composite
of death and MI (4.9% and 5.8%, respectively;
P ¼ .23). Although this meta-analysis suggests
restenosis superiority of SES over PES, it needs
to be viewed in the context of the variable end-
points, sample size, and adjudication processes
that were used to formulate the analysis. Larger
randomized clinical trials will be useful in provid-
ing additional analyses for the comparative
studies.
Other ongoing drug-eluting stents programs
Endeavor
The Endeavor stent (Medtronic, Minneapolis,
Minnesota) uses a cobalt chromium stent plat-
form, a durable, antithrombotic, phoshorylcho-
line (PC)-encapsulated coating, and another G1
cell cycle inhibitor, zotarolimus, which elutes
from the PC coating over several days [70]. Cobalt
chromium alloys provide the potential for being
stronger than stainless steel with a higher density
that allows for similar radiopacity with thinner
(0.0036’’) stent filaments. The potential advantage
of phosphorylcholine is that it lessens platelet ad-
hesion to the metal surface, and is noninflamma-
tory on long-term vascular compatibility studies.
Zotarolimus delivered in this manner reduced inti-
mal hyperplasia by up to 40% in a porcine model
of stent injury [71].
The Endeavor I study was the first clinical
study to evaluate the safety and feasibility of the
Endeavor stent system for treating symptomatic
coronary artery disease. It enrolled 100 patients at
eight centers in Australia and New Zealand [70].
223
The procedure and device deployment success
rates were 100%. At 12 months, in-stent late
lumen loss was 0.61 G 0.44 mm; in-segment late
lumen loss was 0.43 G 0.44 mm, and neointimal hy-
perplasia volume was 14.2 G 11.8 mm3 (corre-
sponding to a percent volume obstruction of
9.7% G 8.5%). The binary angiographic reste-
nosis rates at 4 and 12 months were 2.1% and
5.4%, respectively, and the pattern of neointimal
hyperplasia was greatest within the stent and
not at the stent edges. The cumulative incidence
of MACE was 1% at 30 days and 2% at 4 and
12 months.
The Endeavor-II trial enrolled 1197 patients
who had a single coronary stenosis in a vessel
diameter between 2.25 and 3.50 mm and a lesion
length between 14 and 27 mm. Patients were
assigned randomly to receive an Endeavor stent
or BMS [72]. Patients were included with complex
coronary disease similar to the SIRIUS and
TAXUS-IV studies, including the presence of dia-
betes (20.1%), smaller vessel diameters (2.75 mm),
and longer lesions (lesion length, 14.2 mm). The
primary endpoint of 9-month target vessel failure
was reduced from 15.1% with BMS to 8.00%
with the Endeavor stent (P ! .0005). The rate
of MACE was reduced from 14.4% with BMS
to 7.3% with the Endeavor stent (P ! .0005).
TLR was 4.6% with Endeavor stent compared
with 11.8% with BMS (P ! .005). The rate of
stent thrombosis was 0.5% with the Endeavor
ABT drug-eluting stent, not different from 1.2%
with bare metal stent. In-stent late loss was re-
duced from 1.03 G 0.59 to 0.62 G 0.47 (P !
.0001), and in-segment late loss was reduced
from 0.71 G 0.61 to 0.36 G 0.47 (P ! .0001),
with the Endeavor ABT drug-eluting stent. The
rate of in-segment angiographic binary restenosis
was reduced from 35.0% to 13.7% with the En-
deavor stent (P ! .0001). There was no evidence
for edge stenosis, coronary aneurysm formation,
or late acquired malposition by intravascular ul-
trasound imaging. The Endeavor III trial com-
pared patients treated with SES and an
Endeavor stent, and the Endeavor IV trial will
compare patients treated with PES and an En-
deavor stent.
The PISCES program
The Conor stent (Conor Medsystems, Menlo
Park, California) is comprised of intrastrut wells
with an erodable polymer that is designed specif-
ically for drug delivery with programmable
224 POPMA & TULLI
pharmacokinetics [73]. The PISCES-I study in-
cluded 244 patients who were treated with
a bare metal Conor stent or one of six different re-
lease formulations that varied in dose (10 or
30 mg) and elution release kinetics, direction, and
duration (5, 10, and 30 days) [73]. The lowest in-
stent late loss (0.38 mm, P ! .01; and 0.30 mm,
P ! .01) and volume obstruction (8%, P ! .01;
and 5%, P ! 0.01) were observed with the 10
and 30 mg doses in the 30-day release groups re-
spectively, whereas the highest in-stent late loss
(0.88 mm), volume obstruction (26%), and reste-
nosis rate (11.6%) were observed in the BMS
group [73]. This stent is being evaluated in a ran-
domized study compared with the TAXUS stent.
Everolimus-eluting durable polymer stent
The Xience stent (Guidant Corporation, Santa
Clara, California) is comprised of the Vision co-
balt chromium stent (Guidant Corp., Japan),
a durable polymer coating, and everolimus,
a sirolimus analog that has immunosuppressive
and antiproliferative effects. This SPIRIT-First
was a first-in-man single blind randomized trial
that compared the safety and efficacy of the
Xience stent with BMS in 56 patients who had
de novo coronary lesions [74]. The in-stent late
loss and percentage diameter stenosis at 1 year
were 0.24 mm and 18%, respectively, in the
Xience group and 0.84 mm and 37% in the
BMS group (P ! .001). Significantly less neointi-
mal hyperplasia was observed in patients treated
with the Xience stent (neointimal volume obstruc-
tion, 10% G 7% versus 28% G 12% in patients
treated with BMS, P ! .001). The overall
MACE rate was 15.4% in the everolimus arm
and 21.4% in the bare stent arm.
SPIRIT II is a 300-patient prospective, single
blind, European, randomized noninferiority trial
comparing the Xience stent with the TAXUS
paclitaxel-eluting stent in patients who have
native coronary artery disease. SPIRIT III is
a 1380-patient global clinical trial evaluating the
Xience Stent. It will include randomization in
1002 patients who will receive either the Xience
stent or PES. Recruitment for this trial has been
completed.
Biolimus A-9-eluting stent
The Stealth drug eluting stent program (Bio-
sensors, Singapore) includes the balloon-expand-
able S stent, a bioresorbable polymer coating, and
Biolimus A-9, a sirolimus analog with potent
inhibition of intimal hyperplasia. A randomized
pilot study in 42 patients using everolimus rather
than Biolimus A-9 compared with a BMS found
a low in-stent late lumen loss (0.10 G 0.22 mm
compared with 0.85 G 0.32 mm in patients treated
with a BMS, P ! .0001) [75]. There was no in-
stent restenoses in patients treated with an evero-
limus-eluting stent. A larger multi-center trial with
the Stealth stent is planned in the United States.
Failed or minimally effective drug-eluting stent
programs
Actinomycin-D-eluting stent
The ACTinomycin-eluting stent Improves
Outcomes by reducing Neointimal hyperplasia
(ACTION) trial randomly assigned 360 patients
to receive an actinomycin-eluting stent (AES)
stent that eluted 2.5 or 10 mg/cm2 of actinomycin
D or BMS [76]. The in-stent late lumen loss and
the proximal and distal edges were higher in
both AES groups than in the BMS group and re-
sulted in higher 6-month and 1-year MACE
(34.8% and 43.1% in the 2.5 and 10 mg/cm2 group
versus 13.5% in the BMS group), driven exclu-
sively by TLR without excess death or myocardial
infarction.
Paclitaxel spray stents
Three trials have evaluated the benefit of the
delivery of a spray coating of paclitaxel that elutes
from the stent over 24 to 48 hours. The ASPECT
trial randomly assigned 177 patients to treatment
with low-dose paclitaxel (3.1 m/mm2), high-dose
paclitaxel (3.1 m/mm2), or BMS [77,78]. Patients
treated with high-dose paclitaxel had a signifi-
cantly lower binary restenosis rate than control-
treated patients (4% versus 27%, P ! .001)
[77]. There was a stepwise reduction in intimal hy-
perplasia within the stented segment in patients
who were treated with high-dose paclitaxel [77].
In the European evaLUation of the pacliTaxel
Eluting Stent (ELUTES) pilot clinical trial, the
safety and efficacy of the V-Flex Plus (Cook
Group Inc., Bloomington, Indiana) coronary
stents that were spray coated on the abluminal
surface with escalating doses of paclitaxel were
evaluated [79]. Binary angiographic restenosis de-
creased from 20.6% in the lowest dose paclitaxel
group to 3.2% in the highest paclitaxel dose group
(P ¼ .056).
The DELIVER trial was a larger randomized,
multi-center clinical evaluation that evaluated the
 DRUG-ELUTING STENTS
nonpolymer-based paclitaxel-coated stent com-
pared with BMS in 1043 patients who had focal
de novo coronary lesions [80]. Although in-stent
late lumen loss was lower with the DELIVER
stent (0.81 mm versus 0.98 mm for BMS, P ¼
.003), there we no significant differences in the pri-
mary study endpoint in patients treated with
DELIVER (11.9%) and BMS (14.5%; P ¼ .12).
These findings suggest that the release kinetics
may be an important determinant in the delivery
of paclitaxel.
Paclitaxel derivative-eluting polymeric sleeve
A multi-sleeve drug delivery coronary stent
(QuaDS-QP-2, Boston Scientific) contained up to
4000 mg of a taxol-derived lipophilic microtubule
inhibitor (QP2) designed to delivery high quanti-
ties of drug to the vessel wall. Following a pilot
study in 32 patients showing a low degree of inti-
mal hyperplasia [81], the Study to COmpare
REstenosis rate between QueST and QuaDS-
QP2 (SCORE) trial was performed to compare
the 7-hexanoyltaxol (QP2)-eluting stents (qDES)
with BMS in 266 patients [82]. The qDES showed
a 68% reduction in neointimal growth within the
stent (P ! .0001) and a significantly lower angio-
graphic restenosis rate with the qDES (6.4% ver-
sus 36.9% in BMS patients, P ! .001). The
program was terminated, however, because of an
unacceptable 10.2% subacute stent thrombosis
rate [82] and delayed restenosis 12 months after
the procedure [83], potentially because of the non-
reabsorbable polymer alone, which may have in-
duced chronic inflammation [84].
Outstanding issues
The use of DES has reduced the occurrence of
restenosis after stent placement dramatically, but
numerous potential issues have been addressed
incompletely from current clinical data, primarily
related to the lingering issue of stent thrombosis,
the infrequent occurrence of stent restenosis, and
the cost-effectiveness of the DES.
Early stent thrombosis
There does not appear to be an increased
propensity for early (30 days to 12 months) stent
thrombosis when patients are treated with ex-
tended dual antiplatelet therapy with SES (2 to 3
months of antiplatelet therapy) or PES (6 months
of antiplatelet therapy) compared with BMS [85].
225
In an analysis of 2512 patients treated with BMS,
SES, or PES, there were no differences in stent
thrombosis among the three stents [86]. Bifurca-
tion stenting in the setting of acute MI was an in-
dependent risk factor for angiographic ST in the
entire population (P ! .001). Stent thrombosis
was associated with a 30-day mortality of 15%
and a nonfatal MI rate of 60% [86].
Two meta-analyses confirmed the absence of
early incremental risk for stent thrombosis asso-
ciated with the use of DES over BMS. A meta-
analysis of 3817 patients that reviewed eight
randomized trials compared the stent thrombosis
rate in patients treated with BMS and PES and
found no incremental risk with PES within 12
months after stent placement [85]. Another meta-
analysis of 10 randomized trials that included
5066 patients followed for 6 to 12 months after
the procedure found no major differences in the
stent thrombosis rate between SES and BMS [87].
In contrast to patients taking the prescribed
duration of dual antiplatelet therapy, premature
discontinuation of dual antiplatelet therapy is
associated with a higher rate of stent thrombosis
[88–90]. In a prospective observational cohort
study of 2229 consecutive patients who underwent
successful implantation of SES or PES, aspirin
was continued indefinitely and clopidogrel or
ticlopidine for at least 3 months after sirolimus-
eluting and for at least 6 months after paclitaxel-
eluting stent implantation [90]. Stent thrombosis
occurred in 1.3% of patients by 9-month follow-
up, resulting in a high (45%) mortality rate. Pre-
dictors of stent thrombosis included premature
discontinuation of antiplatelet therapy (HR, 89.8,
P ! .001), renal failure (HR, 6.49, P ! .001), bi-
furcation lesions (HR, 6.42, P ! .001), diabetes
(HR, 3.71, P ¼ .001), and a lower ejection fraction
(HR, 1.09, P ! .001 for each 10% decrease).
These findings have important implications for
patients undergoing noncardiac surgery within
the first 3 to 6 months after DES placement.
Late stent thrombosis
With the accumulation of longer-term follow-
up with patients undergoing treatment with DES,
there has been the identification of a small number
of patients who develop stent thrombosis late
(greater than 30 to 180 days) following DES
placement, most commonly following the discon-
tinuation aspirin or clopidogrel. These events
have occurred with both SES and PER [91–93].
In a registry of 2006 patients treated with either
226 POPMA & TULLI
PES or SES, late angiographic stent thrombosis,
defined as angiographically proven stent throm-
bosis associated with acute symptoms more than
30 days (average duration of follow-up, 1.5 years),
there were eight angiographically confirmed stent
thromboses in eight patients [91], including three
patients treated with SES (at 2, 25, and 26
months) and five patients treated with PES (at 6,
7, 8, 11, and 14.5 months) [91]. The development
of stent thromboses appeared to be related to dis-
continuation of dual antiplatelet therapy after
DES placement. The potential for developing late
stent thrombosis appears related to delayed endo-
thelialization of the stent [94] or an inflammatory
reaction to the durable polymer coating [84,95,96]
and delayed healing [97].
The effect of late-acquired incomplete stent
apposition on stent thrombosis has not been
demonstrated in late outcome studies. Repeat
coronary arteriography and IVUS performed on
13 patients who received SES and showed in-
complete stent apposition at 6 months found no
further vascular remodeling was observed in the
vessel segment with incomplete stent apposition
[10]. In the TAXUS-II trial, predictive factors of
late-acquired incomplete stent apposition (ISA)
were lesion length, unstable angina, and absence
of diabetes. No stent thrombosis occurred in the
patients diagnosed with ISA over a period of 12
months [98].
Predictors of restenosis after drug-eluting stents
Although clinical restenosis is an uncommon
finding after DES placement, numerous factors
contribute to its occurrence, including lesion
length, stent length, vessel diameter, the presence
of diabetes, and underexpansion of the DES [99–
101]. In an angiographic substudy of SIRIUS,
a multiple regression model predicting late angio-
graphic outcome found that the stented lesion
length and excess stent length were associated
with absolute increases in late lumen renarrowing
[100]. Delayed healing within the vessel wall with
DES may predispose to late aneurysm formation,
and isolated aneurysms have been reported with
SES [102,103] and PES [104].
Cost-effectiveness of drug-eluting stents
With the initial introduction of DES, there was
concern that the increase in price and quantity of
stents used during PCI would have a detrimental
impact on hospital budgets and overall resource
allocation [105]. It was appreciated that DES
reduced the morbidity associated with restenosis,
but had little effect on the overall mortality of pa-
tients undergoing revascularization. As a result,
cost-effectiveness studies were needed to appropri-
ately weigh the expenditures associated with the
use of DES with other potentially more effective
morbidity and mortality reducing therapies. In the
RAVEL trial, medical costs were tracked in 238
patients for 1 year after the procedure using re-
source allocation and Dutch unit cost methods
[106]. The higher initial procedural costs associ-
ated with the use of SES were nearly offset by
cost savings that occurred after the procedure be-
cause of reductions in the need for repeat revascu-
larizations in those assigned to SES. A similar cost
and resource allocation analysis was performed in
patients enrolled in the SIRIUS trial that included
1058 patients with complex coronary stenoses as-
signed to PCI with SES or BMS [107]. Although
the initial hospital costs increased by $2881 per pa-
tient in those assigned to SES, the follow-up costs
were reduced by $2571 dollars per patient in those
assigned to SES, mostly because of reduction in
the need for TLR. Accordingly, the incremental
cost-effectiveness ratio for SES was $1650 dollars
per repeat revascularization event avoided or
$27,540 per quality adjusted year of life gained.
These savings become even more profound as the
prices of DES fall with increased availability and
competition. Ongoing studies will evaluate the
cost benefit of DES or CABG in patients with
multi-vessel disease.
Drug-eluting stents have become an integral
component of PCI in patients with single and
multi-vessel coronary artery disease. The dramatic
reductions in 1-year clinical events associated with
the use of DES have been sustained with late follow-
up studies, albeit at the expense of a small (0.4%)
incidence of late stent thrombosis that occurs up
to 3 years after the procedure. Future generations
of DES will include the development of more
deliverable stents using unique metallic alloys,
likely elimination of the durable polymer coating
in favor of bioresorbable ones, and the develop-
ment of antithrombotic agents that will reduce the
risk of late thrombosis. Ongoing studies will evalu-
ate the benefit of DES over coronary bypass
surgery in patients who have multi-vessel disease,
including those who have diabetes mellitus.
References
[1] Altmann DB, Racz M, Battleman DS, et al. Reduc-
tion in angioplasty complications after the
 DRUG-ELUTING STENTS
introduction of coronary stents: results from a con-
secutive series of 2242 patients. Am Heart J 1996;
132(3):503–7.
[2] Serruys PW, de Jaegere P, Kiemeneij F, et al. A
comparison of balloon-expandable stent implanta-
tion with balloon angioplasty in patients with
coronary artery disease. Benestent Study Group.
N Engl J Med 1994;331(8):489–95.
[3] Fischman DL, Leon MB, Baim DS, et al. A ran-
domized comparison of coronary stent placement
and balloon angioplasty in the treatment of coro-
nary artery disease. Stent Restenosis Study Investi-
gators. N Engl J Med 1994;331(8):496–501.
[4] Baim D, Cutlip D, Midei M, et al. Final results of
a randomized trial comparing the MULTI-LINK
stent with the Palmaz-Schatz stent for narrowings
in native coronary arteries. Am J Cardiol 2001;
87(2):157–62.
[5] Heuser R, Lopez A, Kuntz R, et al. SMART: The
microstent’s ability to limit restenosis trial. Cathe-
ter Cardiovasc Interv 2001;52(3):269–77 [discus-
sion 278].
[6] Baim D, Cutlip D, O’Shaughnessy C, et al. Final re-
sults of a randomized trial comparing the NIR stent
to the Palmaz-Schatz stent for narrowings in native
coronary arteries. Am J Cardiol 2001;87(2):152–6.
[7] Cutlip D, Chauhan M, Baim D, et al. Clinical reste-
nosis after coronary stenting: perspectives from
multicenter clinical trials. J Am Coll Cardiol
2002;40(12):2082–9.
[8] Moses J, Leon M, Popma J, et al. Sirolimus-eluting
stents versus standard stents in patients with steno-
sis in a native coronary artery. N Engl J Med 2003;
349(14):1315–23.
[9] Sousa J, Costa M, Abizaid A, et al. Sustained
suppression of neointimal proliferation by siroli-
mus-eluting stents: one-year angiographic and in-
travascular ultrasound follow-up. Circulation
2001;104(17):2007–11.
[10] Degertekin M, Serruys P, Tanabe K, et al. Long-
term follow-up of incomplete stent apposition
in patients who received sirolimus-eluting stent
for de novo coronary lesions: an intravascular
ultrasound analysis. Circulation 2003;108(22):
2747–50.
[11] Sousa J, Costa M, Abizaid A, et al. Four-year
angiographic and intravascular ultrasound fol-
low-up of patients treated with sirolimus-eluting
stents. Circulation 2005;111(18):2326–9.
[12] Sousa JE, Costa MA, Abizaid A, et al. Lack of
neointimal proliferation after implantation of siroli-
mus-coated stents in human coronary arteries: a quan-
titative coronary angiography and three-dimensional
intravascular ultrasound study. Circulation 2001;
103(2):192–5.
[13] Sousa J, Costa M, Sousa A, et al. Two-year angio-
graphic and intravascular ultrasound follow-up af-
ter implantation of sirolimus-eluting stents in human
coronary arteries. Circulation 2003;107(3):381–3.
227
[14] Morice M, Serruys P, Sousa J, et al. A randomized
comparison of a sirolimus-eluting stent with a stan-
dard stent for coronary revascularization. N Engl
J Med 2002;346(23):1773–80.
[15] Regar E, Serruys P, Bode C, et al. Angiographic
findings of the multi-center Randomized Study
With the Sirolimus-Eluting Bx Velocity Balloon-
Expandable Stent (RAVEL): sirolimus-eluting
stents inhibit restenosis irrespective of the vessel
size. Circulation 2002;106(15):1949–56.
[16] Tanabe K, Serruys P, Degertekin M, et al. Fate of
side branches after coronary arterial sirolimus-eluting
stent implantation. Am J Cardiol 2002;90(9):937–41.
[17] Serruys P, Degertekin M, Tanabe K, et al. Intravas-
cular ultrasound findings in the multi-center, ran-
domized, double-blind RAVEL (RAndomized
study with the sirolimus-eluting VElocity balloon-
expandable stent in the treatment of patients with
de novo native coronary artery Lesions) trial.
Circulation 2002;106(7):798–803.
[18] Fajadet J, Morice M, Bode C, et al. Maintenance of
long-term clinical benefit with sirolimus-eluting
coronary stents: three-year results of the RAVEL
trial. Circulation 2005;111(8):1040–4.
[19] Holmes D, Leon M, Moses J, et al. Analysis of
1-year clinical outcomes in the SIRIUS trial: a ran-
domized trial of a sirolimus-eluting stent versus
a standard stent in patients at high risk for coro-
nary restenosis. Circulation 2004;109(5):634–40.
[20] Popma J, Leon M, Moses J, et al. Quantitative
assessment of angiographic restenosis after siroli-
mus-eluting stent implantation in native coronary
arteries. Circulation 2004;110(25):3773–80.
[21] Schofer J, Schluter M, Gershlick A, et al. Sirolimus-
eluting stents for treatment of patients with long
atherosclerotic lesions in small coronary arteries:
double-blind, randomised controlled trial (E-
SIRIUS). Lancet 2003;362(9390):1093–9.
[22] Schampaert E, Cohen E, Schluter M, et al. The Ca-
nadian study of the sirolimus-eluting stent in the
treatment of patients with long de novo lesions in
small native coronary arteries (C-SIRIUS). J Am
Coll Cardiol 2004;43(6):1110–5.
[23] Sousa J, Costa M, Abizaid A, et al. Sirolimus-
eluting stent for the treatment of in-stent restenosis:
a quantitative coronary angiography and three-
dimensional intravascular ultrasound study. Circu-
lation 2003;107(1):24–7.
[24] Airoldi F, Rogacka R, Briguori C, et al. Compari-
son of clinical and angiographic outcome of siroli-
mus-eluting stent implantation versus cutting
balloon angioplasty for coronary in-stent resteno-
sis. Am J Cardiol 2004;94(10):1297–300.
[25] Saia F, Lemos P, Hoye A, et al. Clinical outcomes
for sirolimus-eluting stent implantation and vascu-
lar brachytherapy for the treatment of in-stent reste-
nosis. Catheter Cardiovasc Interv 2004;62(3):283–8.
[26] Degertekin M, Regar E, Tanabe K, et al. Sirolimus-
eluting stent for treatment of complex in-stent
228 POPMA & TULLI
restenosis. The first clinical experience. J Am Coll
Cardiol 2003;41(2):184–9.
[27] Holmes D, Popma J, Kuntz R, et al. A multi-center,
randomized study of the sirolimus-eluting Bx veloc-
ity stent versus intravascular brachytherapy in the
treatment of patients with in-stent restenosis in na-
tive coronary arteries. Late-breaking clinical trials.
Dallas (TX): American Heart Association; 2005.
[28] Lemos P, Lee C, Degertekin M, et al. Early out-
come after sirolimus-eluting stent implantation in
patients with acute coronary syndromes: insights
from the Rapamycin-Eluting Stent Evaluated At
Rotterdam Cardiology Hospital (RESEARCH)
registry. J Am Coll Cardiol 2003;41(11):2093–9.
[29] Lemos P, Saia F, Hofma S, et al. Short- and long-
term clinical benefit of sirolimus-eluting stents com-
pared to conventional bare stents for patients with
acute myocardial infarction. J Am Coll Cardiol
2004;43(4):704–8.
[30] Moussa I, Leon M, Baim D, et al. Impact of siroli-
mus-eluting stents on outcome in diabetic patients:
a SIRIUS (SIRolImUS-coated Bx Velocity balloon-
expandable stent in the treatment of patients with de
novo coronary artery lesions) substudy. Circulation
2004;109(19):2273–8.
[31] Ardissino D, Cavallini C, Bramucci E, et al. Siroli-
mus-eluting vs uncoated stents for prevention of
restenosis in small coronary arteries: a randomized
trial. JAMA 2004;292(22):2727–34.
[32] Hoye A, Tanabe K, Lemos P, et al. Significant
reduction in restenosis after the use of sirolimus-
eluting stents in the treatment of chronic total oc-
clusions. J Am Coll Cardiol 2004;43(11):1954–8.
[33] Hoye A, Lemos P, Arampatzis C, et al. Effective-
ness of the sirolimus-eluting stent in the treatment
of patients with a prior history of coronary artery by-
pass graft surgery. Coron Artery Dis 2004;15(3):171–5.
[34] Price M, Sawhney N, Kao J, et al. Clinical out-
comes after sirolimus-eluting stent implantation
for de novo saphenous vein graft lesions. Catheter
Cardiovasc Interv 2005;65(2):208–11.
[35] Valgimigli M, van MC, Ong A, et al. Short- and
long-term clinical outcome after drug-eluting stent
implantation for the percutaneous treatment of left
main coronary artery disease: insights from the
Rapamycin-Eluting and Taxus Stent Evaluated
At Rotterdam Cardiology Hospital registries
(RESEARCH and T-SEARCH). Circulation 2005;
111(11):1383–9.
[36] Park S, Kim Y, Lee B, et al. Sirolimus-eluting stent
implantation for unprotected left main coronary
artery stenosis: comparison with bare metal stent
implantation. J Am Coll Cardiol 2005;45(3):351–6.
[37] Aoki J, Hoye A, Staferov A, et al. Sirolimus-eluting
stent implantation for chronic total occlusion of the
left main coronary artery. J Interv Cardiol 2005;
18(1):65–9.
[38] Arampatzis C, Lemos P, Tanabe K, et al. Effective-
ness of sirolimus-eluting stent for treatment of left
main coronary artery disease. Am J Cardiol 2003;
92(3):327–9.
[39] Daemen J, Lemos P, Serruys P. Multi-lesion culotte
and crush bifurcation stenting with sirolimus-
eluting stents: long-term angiographic outcome.
J Invasive Cardiol 2003;15(11):653–6.
[40] Colombo A, Moses J, Morice M, et al. Random-
ized study to evaluate sirolimus-eluting stents
implanted at coronary bifurcation lesions. Circula-
tion 2004;109(10):1244–9.
[41] Rizik D, Dowler D, Villegas B. Balloon alignment
T-stenting for bifurcation coronary artery disease
using the sirolimus-eluting stent. J Invasive Cardiol
2005;17(8):437–9.
[42] Iakovou I, Sangiorgi G, Stankovic G, et al. Results
and follow-up after implantation of four or more
sirolimus-eluting stents in the same patient. Cathe-
ter Cardiovasc Interv 2005;64(4):436–9 [discussion
440–1].
[43] Schluter M, Schofer J, Gershlick A, et al. Direct
stenting of native de novo coronary artery lesions
with the sirolimus-eluting stent: a post hoc subanal-
ysis of the pooled E- and C-SIRIUS trials. J Am
Coll Cardiol 2005;45(1):10–3.
[44] Stone G, Ellis S, Cox D, et al. A polymer-based,
paclitaxel-eluting stent in patients with coronary
artery disease. N Engl J Med 2004;350(3):221–31.
[45] Grube E, Silber S, Hauptmann K, et al. TAXUS I:
six- and twelve-month results from a randomized,
double-blind trial on a slow-release paclitaxel-
eluting stent for de novo coronary lesions. Circula-
tion 2003;107(1):38–42.
[46] Grube E, Silber S, Hauptmann K, et al. Two-year-
plus follow-up of a paclitaxel-eluting stent in de
novo coronary narrowings (TAXUS I). Am J
Cardiol 2005;96(1):79–82.
[47] Colombo A, Drzewiecki J, Banning A, et al.
Randomized study to assess the effectiveness of
slow- and moderate-release polymer-based Pacli-
taxel-eluting stents for coronary artery lesions. Cir-
culation 2003;108(7):788–94.
[48] Serruys P, Degertekin M, Tanabe K, et al. Vascular
responses at proximal and distal edges of pacli-
taxel-eluting stents: serial intravascular ultrasound
analysis from the TAXUS II trial. Circulation
2004;109(5):627–33.
[49] Tanabe K, Serruys P, Grube E, et al. TAXUS III
trial: in-stent restenosis treated with stent-based de-
livery of paclitaxel incorporated in a slow-release
polymer formulation. Circulation 2003;107(4):
559–64.
[50] Stone G, Ellis S, Cox D, et al. One-year clinical
results with the slow-release, polymer-based, pacli-
taxel-eluting TAXUS stent: the TAXUS-IV trial.
Circulation 2004;109(16):1942–7.
[51] Cheneau E, Pichard A, Satler L, et al. Intravascular
ultrasound stent area of sirolimus-eluting stents
and its impact on late outcome. Am J Cardiol
2005;95(10):1240–2.
 DRUG-ELUTING STENTS
[52] Stone G, Ellis S, Cannon L, et al. Comparison of
a polymer-based paclitaxel-eluting stent with
a bare metal stent in patients with complex coro-
nary artery disease: a randomized controlled trial.
JAMA 2005;294(10):1215–23.
[53] Dawkins K, Grube E, Guagliumi G, et al. Clinical
efficacy of polymer-based paclitaxel-eluting stents
in the treatment of complex, long coronary artery
lesions from a multi-center, randomized trial: support
for the use of drug-eluting stents in contemporary clin-
ical practice. Circulation 2005;112(21):3306–13.
[54] Hermiller J, Raizner A, Cannon L, et al. Outcomes
with the polymer-based paclitaxel-eluting TAXUS
stent in patients with diabetes mellitus: the TAXUS-
IV trial. J Am Coll Cardiol 2005;45(8):1172–9.
[55] Lansky A, Costa R, Mooney M, et al. Gender-
based outcomes after paclitaxel-eluting stent im-
plantation in patients with coronary artery disease.
J Am Coll Cardiol 2005;45(8):1180–5.
[56] Moses J, Mehran R, Nikolsky E, et al. Outcomes
with the paclitaxel-eluting stent in patients with
acute coronary syndromes: analysis from the
TAXUS-IV trial. J Am Coll Cardiol 2005;45(8):
1165–71.
[57] Valgimigli M, Percoco G, Cicchitelli G, et al. High-
dose bolus tirofiban and sirolimus eluting stent
versus abiciximab and bare metal stent in acute
myocardial infarction (STRATEGY) study–protocol
design and demography of the first 100 patients.
Cardiovasc Drugs Ther 2004;18(3):225–30.
[58] Valgimigli M, Percoco G, Malagutti P, et al.
Tirofiban and sirolimus-eluting stent vs abciximab
and bare-metal stent for acute myocardial infarc-
tion: a randomized trial. JAMA 2005;293(17):
2109–17.
[59] Tsuchida K, Ong A, Aoki J, et al. Immediate and
one-year outcome of percutaneous intervention of
saphenous vein graft disease with paclitaxel-eluting
stents. Am J Cardiol 2005;96(3):395–8.
[60] Dangas G, Ellis S, Shlofmitz R, et al. Outcomes of
paclitaxel-eluting stent implantation in patients
with stenosis of the left anterior descending
coronary artery. J Am Coll Cardiol 2005;45(8):
1186–92.
[61] Applegate R, Draughn T, Davis B. Treatment of
complex LAD-diagonal bifurcation disease using
paclitaxel drug-eluting stents. J Invasive Cardiol
2005;17(7):390–2.
[62] Buellesfeld L, Gerckens U, Mueller R, et al. Poly-
mer-based paclitaxel-eluting stent for treatment of
chronic total occlusions of native coronaries: Re-
sults of a Taxus CTO registry. Catheter Cardiovasc
Interv 2005;66(2):173–7.
[63] Silber S, Hamburger J, Grube E, et al. Direct stent-
ing with TAXUS stents seems to be as safe and ef-
fective as with predilatation. A post hoc analysis of
TAXUS II. Herz 2004;29(2):171–80.
[64] Morice M, Colombo A, Meier B, et al. Sirolimus-
vs paclitaxel-eluting stents in de novo coronary
229
artery lesions: the REALITY trial: a randomized
controlled trial. JAMA 2006;295(8):895–904.
[65] Ellis S, Popma J, Lasala J, et al. Relationship
between angiographic late loss and target lesion
revascularization after coronary stent implanta-
tion: analysis from the TAXUS-IV trial. J Am
Coll Cardiol 2005;45(8):1193–200.
[66] Windecker S, Remondino A, Eberli F, et al. Siroli-
mus-eluting and paclitaxel-eluting stents for coro-
nary revascularization. N Engl J Med 2005;
353(7):653–62.
[67] Goy J, Stauffer J, Siegenthaler M, et al. A prospec-
tive randomized comparison between paclitaxel
and sirolimus stents in the real world of interven-
tional cardiology: the TAXi trial. J Am Coll Cardi-
ol 2005;45(2):308–11.
[68] Kastrati A, Mehilli J, von Beckerath N, et al. Siro-
limus-eluting stent or paclitaxel-eluting stent vs
balloon angioplasty for prevention of recurrences in
patients with coronary in-stent restenosis: a random-
ized controlled trial. JAMA 2005;293(2):165–71.
[69] Kastrati A, Dibra A, Eberle S, et al. Sirolimus-
eluting stents vs paclitaxel-eluting stents in patients
with coronary artery disease: meta-analysis of ran-
domized trials. JAMA 2005;294(7):819–25.
[70] Meredith I, Ormiston J, Whitbourn R, et al. First-
in-human study of the Endeavor ABT-578-eluting
phosphorylcholine-encapsulated stent system in
de novo native coronary artery lesions: Endeavor
I trial. Eurointervention 2005;1:157–64.
[71] Collingwood R, Gibson L, Sedlik S, et al. Stent-
based delivery of ABT-578 via a phosphorylcholine
surface coating reduces neointimal formation in the
porcine coronary model. Catheter Cardiovasc
Interv 2005;65(2):227–32.
[72] American College of Cardiology, late breaking
clinical trials, March 2005.
[73] Serruys P, Sianos G, Abizaid A, et al. The effect of
variable dose and release kinetics on neointimal hy-
perplasia using a novel paclitaxel-eluting stent plat-
form: the Paclitaxel In-Stent Controlled Elution
Study (PISCES). J Am Coll Cardiol 2005;46(2):
253–60.
[74] Tsuchida K, Piek J, Neumann FJ, et al. One-year
results of a durable polymer everolimus-eluting
stents in de novo coronary narrowings (The
SPIRIT FIRST Trial). Eurointervention 2006, in
press.
[75] Costa M, Angiolillo D, Teirstein P, et al. Sirolimus-
eluting stents for treatment of complex bypass graft
disease: insights from the SECURE registry. J In-
vasive Cardiol 2005;17(8):396–8.
[76] Serruys P, Ormiston J, Sianos G, et al. Actinomy-
cin-eluting stent for coronary revascularization:
a randomized feasibility and safety study: the
ACTION trial. J Am Coll Cardiol 2004;44(7):
1363–7.
[77] Hong M, Mintz G, Lee C, et al. Paclitaxel coating
reduces in-stent intimal hyperplasia in human
230 POPMA & TULLI
coronary arteries: a serial volumetric intravascular
ultrasound analysis from the Asian Paclitaxel-Elut-
ing Stent Clinical Trial (ASPECT). Circulation
2003;107(4):517–20.
[78] Park S, Shim W, Ho D, et al. A paclitaxel-eluting
stent for the prevention of coronary restenosis.
N Engl J Med 2003;348(16):1537–45.
[79] Gershlick A, De Scheerder I, Chevalier B, et al.
Inhibition of restenosis with a paclitaxel-eluting,
polymer-free coronary stent: the European evaLU-
ation of pacliTaxel Eluting Stent (ELUTES) trial.
Circulation 2004;109(4):487–93.
[80] Lansky A, Costa R, Mintz G, et al. Nonpolymer-
based paclitaxel-coated coronary stents for the
treatment of patients with de novo coronary le-
sions: angiographic follow-up of the DELIVER
clinical trial. Circulation 2004;109(16):1948–54.
[81] de la Fuente L, Miano J, Mrad J, et al. Initial results
of the Quanam drug-eluting stent (QuaDS-QP-2)
Registry (BARDDS) in human subjects. Catheter
Cardiovasc Interv 2001;53(4):480–8.
[82] Grube E, Lansky A, Hauptmann K, et al. High-
dose 7-hexanoyltaxol-eluting stent with polymer
sleeves for coronary revascularization: one-year re-
sults from the SCORE randomized trial. J Am Coll
Cardiol 2004;44(7):1368–72.
[83] Liistro F, Stankovic G, Di MC, et al. First clinical
experience with a paclitaxel derivate-eluting poly-
mer stent system implantation for in-stent resteno-
sis: immediate and long-term clinical and angiographic
outcome. Circulation 2002;105(16):1883–6.
[84] Virmani R, Liistro F, Stankovic G, et al. Mecha-
nism of late in-stent restenosis after implantation
of a paclitaxel derivate-eluting polymer stent sys-
tem in humans. Circulation 2002;106(21):2649–51.
[85] Bavry A, Kumbhani D, Helton T, et al. What is the
risk of stent thrombosis associated with the use of
paclitaxel-eluting stents for percutaneous coronary
intervention?: a meta-analysis. J Am Coll Cardiol
2005;45(6):941–6.
[86] Ong A, Hoye A, Aoki J, et al. Thirty-day incidence
and six-month clinical outcome of thrombotic stent
occlusion after bare-metal, sirolimus, or paclitaxel
stent implantation. J Am Coll Cardiol 2005;45(6):
947–53.
[87] Katritsis D, Karvouni E, Ioannidis J. Meta-analysis
comparing drug-eluting stents with bare metal
stents. Am J Cardiol 2005;95(5):640–3.
[88] Jeremias A, Sylvia B, Bridges J, et al. Stent throm-
bosis after successful sirolimus-eluting stent im-
plantation. Circulation 2004;109(16):1930–2.
[89] Kerner A, Gruberg L, Kapeliovich M, et al. Late
stent thrombosis after implantation of a sirolimus-
eluting stent. Catheter Cardiovasc Interv 2003;
60(4):505–8.
[90] Iakovou I, Schmidt T, Bonizzoni E, et al. Inci-
dence, predictors, and outcome of thrombosis after
successful implantation of drug-eluting stents.
JAMA 2005;293(17):2126–30.
[91] Ong A, McFadden E, Regar E, et al. Late angio-
graphic stent thrombosis (LAST) events with
drug-eluting stents. J Am Coll Cardiol 2005;
45(12):2088–92.
[92] Kang W, Han S, Choi K, et al. Acute myocardial
infarction caused by late stent thrombosis after de-
ployment of a paclitaxel-eluting stent. J Invasive
Cardiol 2005;17(7):378–80.
[93] Lee C, Tan H, Ong H, et al. Late thrombotic occlu-
sion of paclitaxel eluting stent more than one year
after stent implantation. Heart 2004;90(12):1482.
[94] Drachman D, Edelman E, Seifert P, et al. Neointi-
mal thickening after stent delivery of paclitaxel:
change in composition and arrest of growth over
six months. J Am Coll Cardiol 2000;36(7):2325–32.
[95] Carter A, Aggarwal M, Kopia G, et al. Long-term
effects of polymer-based, slow-release, sirolimus-
eluting stents in a porcine coronary model. Cardio-
vasc Res 2004;63(4):617–24.
[96] Virmani R, Guagliumi G, Farb A, et al. Localized
hypersensitivity and late coronary thrombosis sec-
ondary to a sirolimus-eluting stent: should we be
cautious? Circulation 2004;109(6):701–5.
[97] Finn A, Kolodgie F, Harnek J, et al. Differential
response of delayed healing and persistent inflam-
mation at sites of overlapping sirolimus- or pacli-
taxel-eluting stents. Circulation 2005;112(2):270–8.
[98] Tanabe K, Serruys P, Degertekin M, et al. Incom-
plete stent apposition after implantation of pacli-
taxel-eluting stents or bare metal stents: insights
from the randomized TAXUS II trial. Circulation
2005;111(7):900–5.
[99] Fujii K, Mintz G, Kobayashi Y, et al. Contribution
of stent underexpansion to recurrence after siroli-
mus-eluting stent implantation for in-stent resteno-
sis. Circulation 2004;109(9):1085–8.
[100] Kereiakes D, Kuntz R, Mauri L, et al. Surrogates,
substudies, and real clinical end points in trials of
drug-eluting stents. J Am Coll Cardiol 2005;45(8):
1206–12.
[101] Takebayashi H, Kobayashi Y, Dangas G, et al.
Restenosis due to underexpansion of sirolimus-
eluting stent in a bifurcation lesion. Catheter Cardi-
ovasc Interv 2003;60(4):496–9.
[102] Singh H, Singh C, Aggarwal N, et al. Mycotic
aneurysm of left anterior descending artery after
sirolimus-eluting stent implantation: a case re-
port. Catheter Cardiovasc Interv 2005;65(2):
282–5.
[103] Stabile E, Escolar E, Weigold G, et al. Marked mal-
position and aneurysm formation after sirolimus-
eluting coronary stent implantation. Circulation
2004;110(5):e47–8.
[104] Vik-Mo H, Wiseth R, Hegbom K. Coronary
aneurysm after implantation of a paclitaxel-eluting
stent. Scand Cardiovasc J 2004;38(6):349–52.
[105] Weintraub W. Economics of sirolimus-eluting
stents: drug-eluting stents have really arrived. Cir-
culation 2004;110(5):472–4.
 DRUG-ELUTING STENTS 231

[106] van Hout B, Serruys P, Lemos P, et al. One-year
cost-effectiveness of sirolimus-eluting stents
compared with bare metal stents in the treatment
of single native de novo coronary lesions: an anal-
ysis from the RAVEL trial. Heart 2005;91(4):
507–12.
[107] Cohen D, Bakhai A, Shi C, et al. Cost-effectiveness of
sirolimus-eluting stents for treatment of complex cor-
onary stenoses: results from the Sirolimus-Eluting
Balloon Expandable Stent in the Treatment of Pa-
tients With De Novo Native Coronary Artery Le-
sions (SIRIUS) trial. Circulation 2004;110(5):508–14.
 Cardiol Clin 24 (2006) 233–246
Coronary Bifurcation Lesions
Samin K. Sharma, MD, FACC*,
Annapoorna S. Kini, MD, FACC, MRCP
Cardiac Catheterization Laboratory, Cardiovascular Institute, Mount Sinai Hospital, Box 1030,
One Gustave Levy Place, New York, NY 10029-6574, USA
Coronary bifurcations are prone to develop
atherosclerotic plaque because of turbulent blood
flow and high shear stress. These lesions amount
to 15% to 20% of the total number of coronary
interventions. The true bifurcation lesion consists
of O50% diameter obstruction of the main vessel
(MV) and side branch (SB) in an inverted Y
fashion.
Treatment of coronary bifurcation lesions rep-
resents a challenging area in interventional cardi-
ology, but recent advances in percutaneous
coronary interventions have led to the dramatic
increase in the number of patients successfully
treated percutaneously. When compared with
nonbifurcation interventions, bifurcation inter-
ventions have a lower rate of procedural success,
higher procedural costs, longer hospitalization,
and a higher rate of clinical and angiographic
restenosis [1]. The recent introduction of drug-
eluting stents (DES) has resulted in a lower event
rate and reduction of MV restenosis in compari-
son with historical controls. SB ostial residual ste-
nosis and long-term restenosis remain a problem,
however. Although stenting the MV with provi-
sional SB stenting seems to be the prevailing ap-
proach, in the era of DES various two-stent
techniques have emerged to allow stenting of the
large SB (Fig. 1).
Anatomic classification
Coronary bifurcations have been classified
according to the angulation between the MV
* Corresponding author.
E-mail address: samin.sharma@msnyuhealth.org
(S.K. Sharma).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.02.003
and the SB and according to the location of the
plaque burden. The bifurcations are classified
based on the SB angulation: (1) Y angulation:
the angulation is !70
and access to the SB is
usually easy, but plaque shifting is more pro-
nounced and precise stent placement in the ostium
is difficult. (2) T angulation: the angulation is
O70
and access to the SB is usually more
difficult, but plaque shifting is often minimal and
precise stent placement in ostium is easy. Re-
garding plaque distribution, numerous attempts
have been made to categorize bifurcations, with
two classification patterns commonly used: Duke
classification (Fig. 2) and Lefevre classification
(Fig. 3) [2]. Although these classification pat-
terns are used commonly, they suffer the limita-
tions of coronary angiography (different plaque
distribution and extent of disease when evalu-
ated by intravascular ultrasound), and they do
not take into account what happens to the SB
on dilatation of the MV. Lesions alone in SB
or MV may convert into true Y bifurcation
because of plaque shift or stent protrusion during
coronary intervention. Consequently, each lesion
must be approached therapeutically in the con-
text of its own anatomy and the operator’s
experience.
Isolated ostial lesions involving the main vessel or
the side branch
With isolated ostial lesions, it is important to
place a stent accurately to cover the lesion entirely
without protruding into the other branch. Some
operators use intravascular ultrasound to facili-
tate appropriate stent placement. The following
techniques are suggested based on the lesion
location.
cardiology.theclinics.com
234 SHARMA & KINI

Fig. 1. Various techniques for stenting bifurcation lesions.

Isolated ostial lesion of main vessel
There are two approaches for treating these
lesions: (1) placement of a stent at the ostium of
the MV with a balloon protecting the SB and with
inflation of the SB balloon and kissing balloon
only if plaque shift occurs and (2) placement of
a stent in the MV covering the origin of the SB
and then wiring the SB and performing kissing
balloon inflation in case the ostium of the SB
deteriorates.
Isolated ostial lesions of side branch
The most common approach in treating these
lesions is to place a stent at the ostium of the SB,

Fig. 2. Duke’s classification of bifurcation lesions.

 CORONARY BIFURCATION LESIONS 235

Fig. 3. Lefevre’s classification of bifurcation lesions.

frequently with a low-pressure balloon inflated in
the MV (stent pull-back technique) (Fig. 4) [3]. If
after stent placement there is deterioration of the
MV at the site of the bifurcation, the balloon in
the MV is inflated, which protects the stent by a
simultaneous inflation of the stent delivery balloon.
In cases of suboptimal angiographic results in
MV, a stent can be deployed with final kissing
balloon dilatation.
Impossible side branch access
There are some circumstances in which
the location of the plaque in the MV or the

Fig. 4. Stent pull-back technique.

236 SHARMA & KINI
angulation of the SB prevents the wire from being
advanced at the SB. Although rare, after attempt-
ing different types of wires with all types of curves
and techniques, it still may be impossible to
advance a wire in the SB. At that point few
options are available: (1) Stop the procedure
because the risk of losing the SB is too high,
considering also the size and distribution of the
branch (typically an angulated circumflex artery
when stenting the distal bifurcation of an un-
protected left main). (2) Perform directional
coronary atherectomy on the MV with the intent
of removing the plaque that prevents entry toward
the SB. (3) Dilate the MV with a balloon after
advancing the intended SB wire into the MV
distally with the rationale that the plaque modi-
fication to a favorable plaque shift facilitates
access toward the SB. After balloon dilatations,
SB wire is withdrawn gently from the MV and
steered gently into the SB ostium.
Selection of the guiding catheter
The selection of the size (6, 7, or 8 F) of the
guiding catheter occurs after deciding whether to
stent the SB. Treatment of bifurcations frequently
requires simultaneous insertion of two balloons or
two stents; therefore, an appropriate guiding
catheter should be selected. With the currently
available low-profile balloons, it is possible to
insert two balloons inside a large-lumen 6-F
guiding catheter with an internal lumen diameter
of O0.070 in (1.75 mm). If two stents are needed,
they can be inserted only one after the other, not
simultaneously, in a large-lumen 6-F guiding
catheter. The crush or simultaneous kissing stents
technique requires a guiding catheter of minimum
7 F with an internal lumen diameter of 0.081 in
(2.06 mm) or 8 F with an internal lumen diameter
of 0.091 in (2.2 mm).
How many stents for the bifurcation lesion: one
or two?
The strategies of using one stent (in MV) or
two stents (one in MV and one in SB) for the
treatment of bifurcation lesions long have been
debated [4–6]. The most important initial question
is whether the SB is large enough (O2.25–2.50
mm) with a sufficient territory of distribution to
justify stent implantation or even balloon dilata-
tion regardless of the bifurcation pattern. If SB
is small (!1.5 mm) and supplies a small area
of myocardium, it should be ignored during
percutaneous coronary intervention and a stent
can be placed in the MV across the SB. There
are also rare circumstances in which the SB is im-
portant and large but cannot be wired, even by ex-
pert operators using single core wires or
hydrophilic wires. In these situations an operator
must consider alternative options, such as coro-
nary artery bypass surgery, if the bifurcation in
question is the left main or the left anterior de-
scending vessel and a large diagonal vessel.
The decision to use one or two stents should be
made as early as possible. An appropriate and
timely decision affects the results, saves time,
lowers costs, and lowers the risk of complications.
If the decision is made to use one stent (in the
MV), the possibility almost always exists of
placing a second stent on the SB in case of
suboptimal results. This strategy is defined as
provisional SB stenting. A recent randomized trial
compared one DES (Cypher) versus two DES in
the treatment of coronary bifurcation lesions [7].
This study revealed that although both techniques
resulted in low angiographic restenosis of MV
(approximately 5%), routine stenting of SB (two
stents) was associated with a trend toward higher
restenosis in the SB (23% versus 14%; P ¼ 0.22) and
insignificantly higher overall target lesion revascu-
larization (TLR) (9.5% versus 4.5%; P ¼ 0.42)
(Fig. 5).
One stent by intention to treat: conventional
(provisional) side branch stenting technique
The most common approach in the treatment
of bifurcations is stenting only the MV and
provisional stenting of the SB if needed for
suboptimal angiographic results before or after
stent deployment in the MV. The following steps
are taken in this technique:
1. Wire the MV and the SB.
2. Decide the predilation device for the MV or
the SB. Atherotomy devices (cutting balloon
or FXminiRAIL) can be used to dilate the
SB ostium with controlled focal cuts without
significant dissection. Rotational atherec-
tomy may be required in heavily calcified
lesions in the MV or SB using a single bur
with a bur:artery ratio of 0.4–0.5.
3. Place a stent in the MV. The stent should be
deployed at a pressure of 12 to 18 atm while
leaving the SB wire to prevent plaque shift,
closure, or dissections in the ostium. Rarely
postdilatation with a high-pressure balloon
may be needed at the area of maximal plaque
 CORONARY BIFURCATION LESIONS
Fig. 5. The bifurcation study with Cypher Sirolimus-Eluting Stent.
burden for full stent expansion. If angio-
graphic results in the MV and SB are satis-
factory, the procedure is completed, and
trapped guidewire in the SB behind the stent
struts can be removed gently.
4. Place a wire into the SB to perform kissing
balloon dilatations after dilatation of the
SB ostium if SB ostium remained narrow or
dissected. This procedure can be performed
with the wire trapped behind the stent to
serve as a marker. For re-entering the SB,
a floppy wire, such as the Balance Universal
(Guidant, Temecula, California) or Luge
wire (Boston Scientific, Natick, Massachu-
setts), is recommended. In rare cases, a hydro-
philic wire, such as the Pilot or Whisper
(Guidant), cross-it (Guidant), or Asahi wire
(Abbott Laboratories, Abbott Park, Illinois)
is recommended. In this case, dilatation of
the SB and kissing balloon inflation (usually
at 8–12 atm) between the main and the SB
is performed. If the result is acceptable after
kissing balloon inflation, the procedure is
considered complete.
5. Decide whether to stop if the result at the SB
remains unsatisfactory because of difficulty
in positioning a stent, size, distal runoff, or
complexity of the procedure. If the decision
is made to improve the result at the level of
the SB, then stenting is performed according
to the reverse T approach, advancing the
stent via the MV stent struts with final kiss-
ing balloon dilatation.
Two stents by intention to treat
Several two-stent techniques are available,
with various levels of complexity and indications:
237
the V, the simultaneous kissing stents, crush and
its variations (reverse and step), T and its varia-
tion (modified), culottes, and Y.
The V and the simultaneous kissing stents
techniques
The V technique consists of the delivery and
implantation of two stents together. One stent is
advanced in the SB, the other in the MV, and the
two stents touch each other to form a small
proximal carina (!2 mm) (Fig. 6). When the
carina extends a considerable length (usually 3 mm
or more) into the MV, this technique is called
simultaneous kissing stents (SKS) (Fig. 7), with its
modified alternative (trouser SKS) for the long
lesions proximally (to avoid new long carina)
(Fig. 8). The types of lesion most suitable for
this technique are proximal lesions, such as distal
left main bifurcation and other bifurcations with
moderate to large side branch (R2.5 mm) and ves-
sel portion proximal to bifurcation is free of
disease.
Simultaneous kissing stents technique
This technique (see Fig. 7A) involves using two
appropriately sized stents (1:1 stent-to-artery
ratio), one for the MV and one for the SB, with
an overlap of the two stents in the proximal seg-
ment of the MV (stent sized 1:1 to the MV after
the bifurcation). The proximal part of the MV
should be able to accommodate the two stents,
and its size should be approximately two thirds
of the aggregate diameter of the two stents (eg,
for two 3-mm stents in left anterior descending ar-
tery and the diagonal branch, the proximal MV
size should be approximately 4 mm). Stent lengths
are selected visually to cover the entire length
from distal end of the SB and MV lesions to
238 SHARMA & KINI
Fig. 6. The V stenting technique.
proximal end in the MV. A 7- or 8-F guide cath-
eter (internal diameter O0.78 in) is used. Debulk-
ing of MV or SB using cutting balloon or
rotablator with or without balloon angioplasty is
performed as clinically indicated. MV and SB
are wired, and lesions with O80% stenosis are di-
lated by appropriate sized balloons. Two stents
are advanced one by one, initially to the SB fol-
lowed by one to the MV. After this step, both
stents are pulled simultaneously back to the bifur-
cation to make a ‘‘V’’ and then into the proximal
part of the MV to configure a ‘‘Y.’’ The stem of
the ‘‘Y’’ in the MV completely covers the proxi-
mal end of the lesion, with one arm of the Y in
distal MV (covering the distal end of the MV le-
sion) and another arm in the SB (covering the dis-
tal end of SB lesions). The proximal overlapping
parts of the stents are kept as short as possible
but are long enough to cover the proximal end
of the MV lesion.
Once the position of the stents is confirmed
and proximal stent markers are overlapped, stents
are deployed with simultaneous inflation at 10 to
12 atm for 10 to 20 seconds and then deflated.
This procedure is followed by a second dilatation
of the MV stent at 16 to 20 atm for 10 to 20
seconds to expand fully the MV stent struts while
the other SB stent balloon remains deflated in the
SB stent. A third dilatation of the SB stent at 14 to
20 atm for 10 to 20 seconds is performed to
expand fully the SB stent struts while the other
MV stent balloon remains deflated in the MV
stent. This procedure is followed by the fourth
and final simultaneous inflation and deflation at
10 to 12 atm for 10 to 20 seconds to form the
uniform carina of the fully expanded kissing
stents. Deflated stent balloons are withdrawn
simultaneously. In cases of stent underexpansion,
two high-pressure balloons of similar length (they
may be different size) are advanced for the
simultaneous kissing balloon dilatations. In case
of distal dissection, prolonged balloon dilatation
is performed to avoid the need for stenting. In
cases of proximal dissection, two-balloon (one in
each stent) dilatation or a perfusion balloon
dilatation at low pressure in the MV is performed.
Modified simultaneous kissing stents technique
In cases with a long lesion in the proximal part
of MV, before bifurcation, a large stent first is
deployed proximally over the guidewire in the MV
(Fig. 8). It is followed by wiring the side branch
via proximal stent and advancing the two stents
through the MV stent to distal MV and the SB.
It is deployed as described in a previous section
(in a ‘‘trouser-and-seat’’ pattern).
During our initial experience, we compared
100 cases of SKS technique with 100 matched
cases of conventional stent technique and ob-
served lower major adverse cardica events and
TLR rates in SKS versus conventional stent tech-
nique (Fig 9) [8]. Later, we analyzed our first 200
consecutive patients (202 lesions) who underwent
SKS technique for true bifurcation lesions using
sirolimus-eluting stents, with a minimum follow-
up of 6 months (Fig. 10). Procedural success was
100% for MV and 99% for SB using SKS tech-
nique, with clinical success rate of 97%. In-hospi-
tal and 30-day major adverse cardiac events were
3% and 5%, respectively. At mean follow-up of
9 G 2 months, the incidence of target lesion re-
vascularization was 5% in the entire group [9].
The main advantage of these techniques is that
the access to either of the two branches is never
lost. When a final kissing inflation is performed,
there is no need to re-cross any stent. These
techniques also provide a definite SB coverage,
regardless of the angulation. It is intuitive how
 CORONARY BIFURCATION LESIONS 239

Fig. 7. (A) The SKS technique. (B) SKS technique for bifurcation calcified unprotected left main coronary artery lesion.
(C) SKS technique of distal left main bifurcation with subtotal ostial left anterior descending.
240 SHARMA & KINI

Fig. 8. The modified SKS (trouser SKS) technique (for long lesions).

problematic the need may be to position a stent
proximal to the double barrel. There is an in-
evitable bias toward one of the two branches and
the high likelihood of leaving a gap. If there is
a need to place a stent at the proximal segment of
a vessel treated with SKS stenting, two options are
available: (1) a stent is placed proximally, which
leaves a small gap between the kissing stents and
the proximal stent and (2) the kissing stent
technique is converted into a crush technique,
with the stent in the MV compressing the other
stent (one arm of the V) in the SB. A wire crosses
the struts into the SB, and a balloon is inflated
toward the SB. After wire removal from the SB,
the proximal stent is advanced toward the MV.
The crush technique
The crush technique was introduced at the time
of DES introduction and is described schemati-
cally in Fig. 11. Two stents are placed in the MV
and the SB, with the former more proximal than
the latter. The stent of the SB is deployed, and
its balloon and wire are removed. The stent subse-
quently deployed in the MV flattens the protrud-
ing cells of the SB stentdhence the name
crushing or crush technique. Wire re-crossing
and dilatation of the SB with a balloon of a diam-
eter at least equal to that of the stent and then
final kissing balloon inflation are recommended.
The implementation of final kissing balloon infla-
tion is performed to allow better strut contact
against the ostium of the SB and better drug deliv-
ery. Follow-up studies have shown that if resteno-
sis occurs, this narrowing is focal (!5 mm in
length) and most of the time is not associated
with symptoms or ischemia [10].
Ge and colleagues [11] published a study to
evaluate the long-term outcomes after implanta-
tion of DES in bifurcation lesions with the crush
technique. Although the long-term outcome of
crush stenting technique has yet to be determined,
results of this study showed that compared with

Fig. 9. SKS versus conventional stent technique: follow-up results.

 CORONARY BIFURCATION LESIONS 241

Fig. 10. SKS-DES technique for bifurcation lesions.

the absence of kissing balloon after dilation, the
crush stenting with kissing balloon after dilation
seems to be associated with more favorable long-
term outcomes (Fig. 12). When using the crush
stenting technique, kissing balloon after dilation
is mandatory to reduce the restenosis rate of SB
and the need for TLR.
The main advantage of the crush technique is
that the immediate patency of both branches is
assured. This technique also provides excellent

Fig. 11. The crush technique.

242
Fig. 12. Long-term outcome of crush stenting technique.
coverage of the ostium of the SB. The main
disadvantage is that the performance of the final
kissing balloon inflation makes the procedure
more laborious because of the need to re-cross
multiple struts with a wire and balloon.
The reverse crush
The main indication for performing the reverse
crush is to allow an opportunity for provisional
SB stenting. A stent is deployed in the MV, and
balloon dilatation with final kissing inflation to-
ward the SB is performed. It is assumed that the
result of the SB is suboptimal and stent placement
will be needed. A second stent is advanced into the
SB and left in position without being deployed. A
balloon sized according to the diameter of the MV
is positioned at the level of the bifurcation, and
the surgeon ensures that it stays inside the pre-
viously deployed MV stent. The stent in the SB is
retracted approximately 2 to 3 mm into the MV
and deployed, the deploying balloon is removed,
and an angiogram is obtained to verify that
a good result is present at the SB (no further
distal stent in the SB is needed). If this is the case,
the wire from the SB is removed and the balloon
in the MV is inflated at high pressure, with final
steps involving re-crossing into the SB, perform-
ing SB dilatation, and final kissing balloon
inflation.
The main advantages of the reverse crush
technique are that the immediate patency of
both branches is assured and the technique can
be performed using a 6-F guiding catheter. This
technique has the same disadvantages as the
standard crush but is more laborious.
T technique
The classic T technique consists of positioning
a stent first at the ostium of the SB while being
SHARMA & KINI
careful to avoid stent protrusion into the MV
(Fig. 13). Some operators leave a balloon in the
MV to help to further locate the MV. After de-
ployment of the stent and removal of the balloon
and the wire from the SB, a second stent is ad-
vanced in the MV. A wire is then re-advanced
into the SB, and final kissing balloon inflation is
performed. Modified T stenting is a variation per-
formed by simultaneously positioning stents at the
SB and the MV [12]. The SB stent is deployed
first, and after wire and balloon removal from
the SB, the MV stent is deployed. The reverse T
stenting technique is used when SB ostium deteri-
orates after stent deployment in the MV, which re-
quires re-crossing the SB, dilating, and then
advancing the stent in the SB. A final kissing bal-
loon dilatation is recommended.
This technique is simple and technically less
demanding. It can be used for the coverage of
lesions located proximal to the bifurcation. In
almost all cases, this technique leads to incomplete
coverage of the ostium of the SB.
The culottes technique
The culottes technique uses two stents and
leads to full coverage of the bifurcation at the
expense of an excess of metal covering of the
proximal end (Fig. 14). Both branches are predi-
lated. First a stent is deployed across the most an-
gulated branch, usually the SB. The nonstented
branch is then rewired through the struts of the
stent and dilated. A second stent is advanced
and expanded into the nonstented branch, usually
the MV. Finally, kissing balloon inflation is per-
formed [13].
This technique is suitable for all angles of
bifurcations and provides near-perfect coverage of
the SB ostium. This technique also leads to a high
 CORONARY BIFURCATION LESIONS
Fig. 13. The T stenting technique.
concentration of metal with a double-stent layer
at the carina and in the proximal part of the
bifurcation. The main disadvantage of the tech-
nique is that rewiring both branches through the
stent struts can be difficult and time consuming.
The Y technique
This technique involves an initial predilatation
followed by stent deployment in each branch
(Fig. 15). If the results are not adequate, a third
stent may be deployed in the MV. Currently this
technique is not commonly used.
This technique is a last resort for treating
demanding bifurcations in which there is a need
to maintain wire access to both branches. The
major limitations of this approach are the need to
modify the delivery system of the proximal stent
and manually crimp the stent on two balloons,
which could be problematic with DES.
Lesion preparation before proceeding
with the intervention
Plaque removal before stent implantation us-
ing directional atherectomy in noncalcified lesions
243
and rotational atherectomy in calcified lesions has
been attractive. Some encouraging results of many
single-center experiences have not been repro-
duced in the randomized trials, however. Rota-
tional atherectomy in the heavily calcified lesions
is the only procedure to permit adequate lesion
dilatation and subsequent stent delivery. Early
reports stated an advantage in facilitating stent
delivery and expansion, with a suggestion for
clinical benefit when used in calcified lesions.
The Stenting Post Rotational Atherectomy
(SPORT) randomized study, which used rota-
tional atherectomy and stenting, failed to support
any advantage of this technology over standard
bare metal stenting. Most of the time, rotablation
is performed only on the MV, but occasionally
also or only on the SB. We think that especially
with the use of DES, lesion preparation with
compliance change for a calcified lesion can
facilitate stent delivery and symmetrical stent
expansion substantially with more homogeneous
drug delivery. Several single-center studies re-
ported the beneficial combination of stenting
preceded by cutting balloon dilatation. In bi-
furcation lesions, in which a large fibrotic plaque
244 SHARMA & KINI
Fig. 14. The culottes stenting technique.
is present at the ostium of the SB, the use of the
cutting balloon as a predilatation strategy before
stenting seems reasonable.
Currently, we suggest the use of the cutting
balloon in moderately calcific and fibrotic lesions,
especially ones that involve the origin of the SB.
In heavily calcified lesions, instead of using
a larger bur, the cutting balloon could be used
after small-bur rotablation with the goal of
minimizing any distal embolization. Symmetric
stent expansion, avoidance of SB recoil, and stent
compression are all attractive hypotheses that
need proper evaluation.
Adjunct pharmacotherapy
Preprocedural heparin administration in doses
of 100 U/kg without and 70 U/kg with glycopro-
tein IIb/IIIa inhibitors to keep activated clotting
time between 250 and 300 sec is recommended.
Use of glycoprotein IIb/IIIa inhibitors is encour-
aged, especially in patients who have thrombus-
containing lesions or acute coronary syndromes,
with use of rotational atherectomy, and for use
with multiple long stents. These agents are some-
times administered when the final result at the
SB seems suboptimal and the operator wishes to
avoid implanting another stent. Periprocedural
preparation with thienopyridines with a 600-mg
loading dose of clopidogrel is routinely used.
The duration of combined thienopyridine and
aspirin treatment after stent implantation should
be emphasized strictly for a minimum of 1 year,
with extended duration in severely complex cases
using multiple stents to avoid delayed stent
thrombosis [14].
Future directions and summary
Based on the size of the SB, an algorithm can
be created in the treatment of bifurcation lesions.
As a conclusion, Fig. 16 shows a suggested algo-
rithm for treatment of bifurcation coronary le-
sions. A wire should be placed in the SB,
especially if there is disease at the ostium or there
is a problematic takeoff. The general consensus is
to try to keep the procedure safe and simple.
When the SB is not severely diseased, implanta-
tion of a stent in the MV and provisional stenting
in the SB is the preferred strategy. Implantation of
two stents as the initial approach is appropriate
when both branches are significantly diseased
and SB is O2.5 mm. Final kissing balloon infla-
tion should be performed in these cases.
 CORONARY BIFURCATION LESIONS 245

Fig. 15. The Y stenting technique.

The ongoing randomized trials entitled ‘‘Cor-
onary bifurcations: Application of the Crushing
Technique Using Sirolimus-eluting stents’’ (CAC-
TUS), which compare a provisional SB strategy
with the crush technique using Cypher stents, and
a pilot trial for treatment of true bifurcation
lesions with simultaneous kissing stents (Precise-
SKS trial) may help to answer better the approach
of one versus two stents in true bifurcation
lesions. Although dedicated stents are being de-
veloped (ie, petal stent [Boston Scientific Corp.],
which has a side hole covered by a 2-mm metal

Fig. 16. Interventional algorithm for bifurcation lesions. Fig. 17. Petal stent.
246 SHARMA & KINI
protrusion [Fig. 17]), their clinical use in the
format of DES is still limited. These devices po-
tentially may have important applications in
proximal large bifurcations and in the left main
trunk.
In summary, major achievements in the stent-
ing of bifurcation lesions since the introduction of
DES are single-digit restenosis rates on the MV
and focal restenosis at the SB, which is frequently
clinically silent [15,16].
References
[1] Dauerman H, Higgins P, Sparano A, et al. Mechan-
ical debulking versus balloon angioplasty for the
treatment of true bifurcation lesions. J Am Coll Car-
diol 1998;32:1845–52.
[2] Lefevre T, Louvard Y, Morice MC, et al. Stenting of
bifurcation lesions: classification, treatments, and
results. Catheter Cardiovasc Interv 2000;49:274–83.
[3] Kini A, Moreno P, Steinheimer A, et al. Effective-
ness of the stent pull-back technique for non-aorto
ostial coronary narrowings. Am J Cardiol 2005;96:
1123–8.
[4] Al Suwaidi J, Berger P, Rihal C, et al. Immediate and
long-term outcome of intracoronary stent implanta-
tion for true bifurcation lesions. J Am Coll Cardiol
2000;35:929–36.
[5] Yamashita T, Nishida T, Adamian M, et al. Bifurca-
tion lesions: two stents versus one stent. Immediate
and follow-up results. J Am Coll Cardiol 2000;35:
1145–51.
[6] Pan M, de Lezo SJ, Medina A, et al. Simple and
complex stent strategies for bifurcated coronary ar-
terial stenosis involving the side-branch origin. Am
J Cardiol 1999;83:1320–5.
[7] Colombo A, Moses J, Morice M, et al. Randomized
study to evaluate sirolimus-eluting stents implanted
at coronary bifurcation lesions. Circulation 2004;
109:1244–9.
[8] Sharma S, Ahsan C, Lee J, et al. Simultaneous kiss-
ing stents (SKS) technique for treating bifurcation
lesions in medium-to-large size coronary arteries.
Am J Cardiol 2004;94:913–7.
[9] Sharma SK. Simultaneous kissing drug-eluting stent
technique for percutaneous treatment of bifurcation
lesions in large-size vessels. Catheter Cardiovasc
Interv 2005;65:10–6.
[10] Colombo A, Stankovic G, Orlic D, et al. Modified
T-stenting technique with crushing for bifurcation
lesions: immediate results and 30-day outcome.
Catheter Cardiovasc Interv 2003;60:145–51.
[11] Ge L, Airoldi F, Iakovou I, et al. Clinical and angio-
graphic outcome after implantation of drug-eluting
stents in bifurcation lesions with the crush stent tech-
nique: importance of final kissing balloon post-
dilation. J Am Coll Cardiol 2005;46:613–20.
[12] Kobayashi Y, Colombo A, Akiyama T, et al. Mod-
ified ‘‘T’’ stenting: a technique for kissing stents in
bifurcational coronary lesions. Cathet Cardiovasc
Diagn 1998;43:323–6.
[13] Chevalier B, Glatt B, Royer T, et al. Placement of
coronary stents in bifurcation lesions by the ‘‘cu-
lotte’’ technique. Am J Cardiol 1998;82:943–9.
[14] Iakovou I, Schmidt T, Bonizzoni E, et al. Incidence,
prediction and outcome of thrombosis after success-
ful implantation of drug-eluting stents. JAMA 2005;
293:2126–30.
[15] Pan M, de Lezo SJ, Medina A, et al. Rapamycin-
eluting stents for the treatment of bifurcated coro-
nary lesions; a randomized comparison of a simple
versus complex strategy. Am Heart J 2004;148:
857–64.
[16] Toutouzas K, Stankovic G, Takagi T, et al. A new
dedicated stent and delivery system for the treatment
of bifurcation lesions: preliminary experience. Cath-
eter Cardiovasc Interv 2003;58:34–42.
 Cardiol Clin 24 (2006) 247–254
Chronic Total Coronary Occlusions
Gregory A. Braden, MD
Cardiology Specialists of North Carolina, 3866 Cedarfield Place Court, Winston-Salem, NC 27106, USA
Chronic total coronary occlusions (CTOs)
occur in up to one third of patients undergoing
coronary angiography and have become the final
frontier of interventional cardiology. Anatomi-
cally, CTOs typically consist of a hard fibro-
calcific proximal cap, a distal cap with generally
less fibrotic material, and a central area of
organized thrombus. Indications for opening
CTOs include relief of angina, improved left
ventricular function, and improved long-term
survival. The fibro-calcific nature of these occlu-
sions is responsible for the somewhat lower
success rates in opening these lesions, predomi-
nantly by increasing the difficulty in passing the
occlusion with a guidewire. Newer technology,
wire-based and non–wire-based, has improved the
ability to cross these previously uncrossable le-
sions, thereby improving the acute success rates of
opening these lesions. Stenting improved long-
term patency rates for these lesions, and now
drug-eluting stents have made the late restenosis
rates similar to those seen for nonoccluded
arteries. Therefore, the clinical imperative for
opening these arteries has increased.
CTOs are defined as occlusions in the coronary
arteries with Thrombolysis in Myocardial Infarc-
tion (TIMI) 0 flow or functional occlusions with
TIMI 1 flow (penetration of contrast without
filling of the distal vessel) of at least 1-month
duration. Age criteria in various studies have
ranged from 2 weeks to 3 months but are difficult
to assess unless serial angiograms are available.
Thus age often is difficult to define and is de-
pendent on clinical history [1–4]. A prior history
of myocardial infarction (MI) was present in
42% to 68% of patients who had angiographically
documented CTOs [5–9]. The prevalence of CTOs
E-mail address: GBraden@triad.rr.com
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.011
is high, occurring in approximately one third of
patients undergoing coronary angiography [10].
Not surprisingly, the incidence of CTOs seems
to increase with patient age, especially in the left
anterior descending coronary artery (LAD) distri-
bution [11]. The presence of one or more CTOs
was an angiographic exclusion for randomization
in 43% of patients ineligible for the German An-
gioplasty Bypass Investigation [12] and for 35%
of patients who had angiographic exclusions for
the Balloon Angioplasty Revascularization Inves-
tigation (BARI) trials [13]. Accordingly, although
CTOs represented 30% to 40% of patients listed
in the National Cardiovascular Registry of the
American Collage of Cardiology, CTO angio-
plasty accounted for only 12% of procedures be-
tween January 1998 and September 2000 in 139
United States hospitals [14].
Clinical indications for treating chronic total
coronary occlusions
The rationale for treating CTOs follows several
lines of evidence; the strongest include improved
survival with successful procedures. The Mid-
America Heart Institute published the results of
a 10-year retrospective analysis of 2007 patients
who had CTOs in whom percutaneous coronary
intervention (PCI) was attempted and matched
those patients with 2007 patients undergoing PCI
for nonocclusive disease between 1980 and 1999.
There was a 74.4% success rate in the CTO group.
Better in-hospital outcomes were associated with
a successful procedure (major adverse coro-
nary event [MACE] rate of 3.2% versus 5.4%;
P ¼ .02). Similarly, there was an improved 10-year
survival advantage associated with a successful
procedure (successful 73.5% versus unsuccessful
65.0%; P ¼ .001). The long-term outcome of
cardiology.theclinics.com
248
CTOs successfully recanalized was similar to suc-
cessful procedures in the matched cohort of pa-
tients undergoing PCI for nonocclusive disease
(73.5% versus 71.9%; P ¼ not significant) [15].
Several other registries have shown similar results.
There was a 56% (P ! .001) reduction in relative
risk for mortality over 7 years’ follow-up in the
British Columbia Cardiac Registry in which
1458 patients who had CTO were treated [16].
The Total Occlusion Angioplasty Study-Societa
Italiana di Cardiologia Invasiva (TOAST-GISE)
showed a similar result in a smaller cohort of
patients (369 patients) over a shorter follow-up
(1 year), with a reduced incidence of cardiac death
or MI with successful procedures (1.1% versus
6.2%; P ¼ .005) [15]. In the only study incorporat-
ing stent use to a significant degree, the Thorax
Center reported 5-year follow-up of 885 consecu-
tive patients who had CTO treated from 1992
through 2002. There was a 65.1% success rate in
these patients. Successful procedures were again
associated with improved 5-year survival (93.5%
versus 88.0%; P ¼ .02) [17].
Besides a survival benefit seen from of treating
CTOs, improvements in clinical symptoms, im-
provements in left ventricular function, and a re-
duced need for late coronary bypass surgery
(CABG) have been associated with successful
opening of CTOs. There was a greater freedom
from angina in TOAST-GISE for successful proce-
dures (88.7% versus 75.0%; P ¼ .008) [5]. Left
ventricular function improved in a series of 95
patients studied at baseline and at 6.7
1.4 months.
Left ventricular ejection fraction increased from
625
13% to 675
11% (P ! .001) with opening
these occluded arteries [18,19]. Chung and col-
leagues [20] showed a similar effect in a population
of 75 patients who did not have known MI, dem-
onstrating an improvement in ejection fraction
from 59.5% to 67.3% (P ! .001), but not in a sim-
ilar group with a previously documented MI
(48.9% to 50.5%; P ¼ not significant). Finally,
TOAST-GISE showed a decreased need for late
CABG in patients who had successful PCI of
CTOs (2.5% versus 15.7%; P ! .001) [5]. The
presence of a CTO was the major angiographic ex-
clusion for both the BARI [20] and Emory Angio-
plasty versus Surgery Trials [5].
Histopathology of chronic total coronary
occlusions
Insight into the difficulty of opening CTOs
and their propensity for complications and
BRADEN
restenosis can be understood better by examining
the histopathology of CTOs. CTOs generally
occur after thrombotic occlusion of the coronary
artery as a result of a ruptured plaque. The
thrombus organizes as it ages, with various
amounts of calcification, fibrosis, and inflamma-
tion. Generally, there is more fibro-calcific plaque
at the proximal cap and at the distal cap. Often
the middle section remains soft, with organized
thrombus present. There is a variable amount of
neovascularization present. Neovascularization
seems to be an early event. Occasionally, the
channels can become large and influence guide-
wire passage. The amount of calcification is
somewhat related to the age of the CTO, although
there can be extensive calcification even in youn-
ger occlusions (!3 months). The amount of
calcification may predict the ability to cross the
total occlusion with a guidewire and may cause
the guidewire to deflect behind the plaque into the
subintimal space during an attempt to cross
[21,22]. CTO vessels may undergo a significant
amount of negative remodeling, probably second-
ary to a chronic decrease in perfusion pressure
or to adventitial responses to neovascularization
[23].
Patient selection
Certain angiographic and clinical features
have been associated with higher success rates
for opening totally occluded coronary arteries.
Angiographic features include the presence of
a tapered stump at the occlusion site and the
presence of microchannels (subtotally occluded
arteries). The features that have been associated
with a lower likelihood of success include the
presence of a blunt or flush occlusion, the
presence of the occlusion at a side branch, small
vessel size, marked tortuosity, and heavy calcifi-
cation. The presence of bridging collaterals also
negatively impacts outcome with CTOs. Lesion
length also predictably affects outcomes, with
longer occlusions having worse results and
short occlusions better chance of success. Non-
visualization of the distal bed remains a strong
contraindication to attempting CTOs using per-
cutaneous techniques. Clinically, the duration of
occlusion has been associated with success, so
that long-duration occlusions are more difficult
to cross. The development of newer CTO spe-
cialty guidewires and newer nonguidewire tech-
nology has made some of the negative predictors
of success obsolete.
 CHRONIC TOTAL CORONARY OCCLUSIONS 249

Procedural techniques for chronic total coronary of dissecting the proximal RCA or the aorta.
occlusions Although there has been a migration toward the
use of smaller guide catheters, the CTO angio-
Like many complex catheterization laboratory
plasty generally is preformed better with larger-
procedures, CTO interventional outcomes are
lumen guides such as 7 or 8 Fr. The larger-caliber
highly technique dependant. Patient selection,
guides give better backup support and allow the
equipment selection, and procedure conduct all
use of multiple guidewires and balloon or support
affect outcomes.
catheters.
Equipment selection for CTOs involves the
Guidewire technology has evolved rather rap-
selection of guide catheters, support catheters,
idly. The ideal guidewire would provide nearly
guidewires, and new-generation technology for
one-to-one torque response with the ability to
crossing these lesions as well as the final dilatation
shape the tip in infinite configurations and to
strategies once across (Figs. 1 and 2).
retain the tip shape with use. Additionally, the tip
Guide catheters should be chosen to provide
should be atraumatic but should have enough
both good back-up support and good coaxial
stiffness to penetrate the hard fibro-calcific caps.
alignment with the coronary ostium. For the left
The frictional losses in feel along the guidewire
system, curved shapes such as extra backup or
should be such that the feel of the tip is main-
extra support are best, but occasionally, an
tained despite proximal tortuosity and calcifica-
Amplatz shape for the circumflex provides the
tion. Unfortunately, the ideal guidewire does not
best support. In the right system, backup support
exist. Previously, soft-tipped or intermediate
is dependent on the orientation of the right
guidewires were the initial choice for attempting
coronary artery (RCA) along with its course after
to open CTOs. In the mid 1990s specialty CTO
the take-off. In general, curves such as a hockey-
guidewires with tapered tips were developed to
stick shape provide moderate support, and a right
address issues with standard guidewires. This
or left Amplatz provides more aggressive support,
development paralleled the development of hy-
but these shapes have a tendency to deep seat the
drophilic guidewires for coronary use and led to
coronary artery and therefore have more chance
a debate concerning the use of hydrophilic versus

Fig. 1. Rota plus drug-eluting stent (Taxus) of CTO (12 years old) of RCA with 12-month follow-up. (A) Long total
occlusion of RCA with bridge collaterals (within white oval). (B,C) 1.25 mm Rota burr and 2.5/20 mm Maverick II bal-
loon (Boston Scientific, Maplegrove, MN). (D) 3.5/32.0 mm Taxus stent in RCS. (E) 0–10% residual obstruction (ar-
row). (F) No angiographic or clinical restenosis is seen at 1-year follow-up (arrow).
250 BRADEN

Fig. 2. CTO of the LAD: Safe-Cross RF wire followed by Rota plus drug-eluting stent (Cypher, Cordis Corp., Miami
Lakes, FL). (A) 100% long calcified lesion of mid-LAD with bridge collaterals (within white circle). (B) 0.010 inch Cross
It 300 wire (Guidant) in false lumen as shown in the contralateral RCA injection. (C) Safe-Cross RF wire advance in the
true lumen after seen bursts of RF energy. (D) After Rota plus percutaneous transluminal coronary angioplasty using
3.0/20 mm Maverick balloon at 6 atmospheres (1.25 mm Rota burr ablation for 55 seconds at 142,000 rpm). (E) 3.5/33
mm Cypher stent in proximal LAD. (F) After drug-eluting stent: 0–10% residual lesion in proximal mid-LAD.

hydrophobic guidewires for crossing CTOs. Hy-
drophilic guidewires generally are used for cross-
ing lesions that are subtotally occluded, with the
possibility that the hydrophilic wire might find
microchannels and traverse the CTO more easily.
Additionally, if there is marked tortuosity and
calcification proximally, there may be a benefit in
using hydrophilic wires to decrease the friction
and improve the manipulation of the wire. Other-
wise, most operators prefer hydrophobic anode
wires to optimize the feel of the lesion. Recently
unibody, solid-core wires with excellent tip-shap-
ing properties have been developed in Japan by
Asahi Intec and marketed in the United States
by Abbott Vascular (Redwood City, California).
These devices include both the Miraclebros and
Confienza lines. These two wire lines are similar in
construction, except that the Confienza is tapered
to 0.01 inches at the tip. The Confienza is also
available as a hybrid wire (Confienza Pro), which
has a hydrophilic coating except for the distal
5 mm of the tip. The Miraclebros series come in
various tip stiffness from 3 to 12 g force, whereas
the Confienza has a 9-g tip. These wires have
improved the crossability of CTOs. Medtronic
(Santa Rosa, California) has recently introduced
a similar line of CTO wires called the Predator
series. Guidant (Tamacula, California) has in-
troduced a CTO subspecialty wire of 0.010 inches
of various stiffness (100–400x) with a nonhydro-
philic stiff tip to facilitate the penetration of
fibrous cap but minimize the subintimal passage
of the wire. In general, one would start with
a softer-tipped, less traumatic guidewire and
advance in stiffness to penetrate the CTO while
minimizing the risk of dissection/perforation.
Support catheters are important in treating
CTOs for various reasons. Support catheters
provide the ability to exchange guidewires easily
or to change the shape of the wire tip during
use. Additionally, the use of a support catheter
decreases the friction associated with wire place-
ment and manipulation and provides additional
backup for the wire to penetrate tough lesions.
Various types of support catheters are available,
including low-profile, over-the-wire balloons, gen-
erally 1.5- or 2.0-mm short balloons. Selective
infusion catheters similar to the Transit Catheter
(Cordis, Miami, Florida), the Spectronetics (West-
bury, New York) Quick-Cross support catheter,
the Intraluminal (Carlsbad, California) Thera-
peutics angled support catheter, and St Jude’s
 CHRONIC TOTAL CORONARY OCCLUSIONS
(St. Paul, Minnesota) Venture catheter, which al-
lows different tip deflections in vivo, thus providing
different angles to the catheter. The Tornus catheter
(Abbott Vascular) is both a support catheter and
penetration catheter for uncrossable lesions.
The planning and conduct of the approach for
opening CTOs is important. To determine the
path of the vessel in the CTO segment, it is
necessary to take angiograms in multiple pro-
jections. The use of biplanar angiography saves
time and provides an easier method for obtaining
multiple angiographic projections. It is important
to define the distal vascular bed beyond the CTO.
Sometimes there are adequate homo-collaterals
for the distal vessel to be seen well. More often,
however, it is necessary to inject the contralateral
vessel to define the vessel distal, using collaterals
to see beyond the CTO well. The ability to see this
distal vessel is one of the most important steps to
wiring the CTO successfully.
During guidewire manipulation the guidewire
may end up in a blind pouch behind the plaque
associated with the CTO. If this occurs, it is very
difficult to recross the plaque and re-enter the true
lumen. In this situation, if the current guidewire is
left in place and a second guidewire is introduced,
the first guidewire serves as a landmark for the
false channel created and may prevent (block) the
second guidewire form entering this false channel.
This technique has been termed the ‘‘see-saw’’ or
‘‘parallel-wire’’ technique.
Another technique described by Antonio Co-
lombo and colleagues [24] uses a subintimal path
predominantly in the RCA. A 0.014-inch hydro-
philic guide with a J configuration is purposely ad-
vanced in the subintimal space beyond the CTO,
and re-entry into the true lumen is attempted in
the distal vessel using a guidewire with a sharply
bent tip. This technique has been termed the ‘‘sub-
intimal tracking and re-entry’’ (STAR) technique.
This technique should be attempted only in the
RCA, where the risk of occluding major side
branches is minimized.
Nonguidewire novel devices for chronic total
coronary occlusions
Newer, nonguidewire approaches for the treat-
ment of CTOs have been approved. These include
the Intraluminal Therapeutics (ILT) Safe-Cross
and the LuMend (Red Wood City, California)
Frontrunner catheters. The Safe-Cross system
uses a 0.014-inch guidewire sleeve, which incorpo-
rates a fiberoptic wire and transmission wire to
251
image and to deliver radio-frequency (RF) energy
for the ablation of plaque. The imaging is forward
looking with a resolution of 100 m, providing a sig-
nal to determine the proximity of the tip of the
guidewire to the vessel wall. It then can deliver
RF energy to the catheter tip to ablate plaque, in-
cluding calcified plaque. Using the imaging array
to avoid the outer vessel, the wire can be redirected
through the true lumen if progress is not being
made in crossing the CTO. Using RF energy
when necessary, the wire can be navigated across
the CTO while incorporating the imaging. The
value of the Safe-Cross system was seen in the
Guided Radio Frequency Energy Ablation of
Total Occlusions registry where Safe-Cross system
was used after an attempt with a conventional
guidewire for at least 10 minutes of flouro-time
failed to cross the lesion. In the trial, there was
a 54.3% success rate in these conventional guide-
wire failures. These device has a significant learn-
ing curve, because there was a marked difference
in the success rate between the first half and the
second half of the study (41.6% versus 67%; P !
.01) [25]. There are also 0.018-inch and 0.035-inch
Safe-Cross guidewires for the use in the periphery.
The other new device receiving Food and Drug
Administration approval for use in CTOs is the
Lumend Frontrunner catheter. This catheter uses
blunt dissection to negotiate through CTOs. Jaws
at the tip of the catheter open and close to push
plaque aside to go through or around the plaque
and make a channel through the CTO. Once
across, the frontrunner catheter is replaced with
a guidewire beyond the CTO, and the lesion is
treated with balloon angioplasty or stenting.
A registry trial was preformed for approval in
guidewire-refractory lesions using 10 minutes of
flouro-time. The Frontrunner was successful in
reaching the distal lumen in 56.1% of these
guidewire-refractory cases [26]. Severe acute com-
plications with both of these devices were low and
occurred in less than 1% of cases.
There are other devices under study in an
attempt to increase success rates for CTOs or to
decrease the time required with its antecedent use
of contrast and radiography to treat CTOs.
Closest to market is the FlowCardia (FlowCardia,
Inc., Sunnyvale, California). Crosser system using
vibrational energy transmitted down a catheter
to jackhammer open the CTO [27]. The Crosser
uses a nitinol wire, which has low energy losses, as
the transmission wire. Initial results demonstrate
a 64.2% success rate for guidewire-refractory le-
sions with an excellent safety profile. Other
252
devices are in the design and early feasibility
stages of development [28].
Outcomes of chronic total coronary occlusions
The short- and long-term outcomes of CTO
intervention have been variable over time and
with operator experience. CTO intervention rep-
resents between 6% and 10% of the total PCIs
preformed in the United States, thus representing
nearly 100,000 procedures annually. Therefore,
the outcome of this patient population contributes
in a significant way to total PCI outcome in the
United States.
The acute success rate for opening CTOs has
increased in contemporary series because of im-
proved technology, a better understanding of the
pathobiology of CTOs, and operator experience
with improved techniques. Generally, published
series have an acute success rate of 50% to 80%, but
the reported success rate may be an overestimation
resulting from a selection bias in the data submitted
for publication. In any circumstance, this patient
population clearly has the lowest acute success rate
and probably represents the last frontier to impact
angioplasty results dramatically. The most com-
mon mode of failure in unsuccessful cases included
failure to cross with a guidewire (in approximately
80% to 90% of cases), failure to cross with a balloon
(in approximately 10% to 15% of cases), and
failure to dilate a resistant lesion (in the remaining
approximately 2% to 5% of cases). Strategies
incorporating stiffer guidewires, more backup sup-
port from guide catheters and support catheters,
and newer non–wire-based technologies are ad-
dressing these issues. The inability to cross with
a balloon is addressed typically by changing to
a Rotablator wire (Boston Scientific, Maple Grove,
Minnesota) and treating the lesion with the excimer
laser (Spectronetics) over the wire that crossed the
CTO. Recently, the Tornus catheter (Abbott Vas-
cular) has been used as both a wire-support catheter
and as a drilling-type catheter to cross some re-
sistant lesions.
Long-term success rates traditionally have
been less than ideal, with high restenosis rates
and high reocclusion rates compared with PCI of
nonoccluded vessels. Various randomized trials
comparing percutaneous transluminal coronary
angioplasty with stenting with bare metal stents
have provided insight into these long-term out-
comes. Restenosis rates for balloon angioplasty of
CTOs have ranged from 33% to 74%. Addition-
ally, the reocclusion rates have been high, ranging
BRADEN
from 7% to 34% in various studies. The use of
stents has improved these outcomes markedly.
With bare metal stents, the restenosis rates have
declined by approximately 40%, ranging from
22% to 55% [5,9]. Likewise, late reocclusion rates
have declined with stenting by more than 50%,
ranging from 2% to 16%. With the advent of
drug-eluting stents, these restenosis rates have im-
proved. Although no randomized trials have been
completed, insight into the effect of drug-eluting
stents on restenosis can be gleaned from several
registries. In the Rapamycin-Eluting Stent evalu-
ated at Rotterdam Cardiology Hospital registry
from the Thorax Center in Rotterdam, The Neth-
erlands, the 1-year event-free survival for CTOs
treated with sirolimus-eluting stents was 96.4%,
compared with a 82.8% event-free survival rate
in a matched, consecutive series of patients treated
with bare metal stents [29]. In a series of 88 patients
from five Asian centers treated with sirolimus-
coated stents, Nakamura and colleagues [30] re-
ported a 6-month MACE rate of 4.5%, which
were all target vessel revascularizations. The angio-
graphic restenosis rate was only 3.4%. In the Siro-
limus Eluting Stent in Chronic Total Occlusion
study, 25 lesions were treated with sirolimus-coated
stents. At 6-month follow-up only two patients
(8%) had target vessel revascularization, and no
other MACE was seen. Two studies of the use of
paclitaxel-eluting stents have been reported. Werner
and colleagues [31] presented the results of 48
patients who had CTO treated with the Taxus (Bos-
ton Scientific Corp., Natick, Massachusetts) stent
and compared these results with historical controls
using bare metal stents. There was an 84% reduc-
tion in the rate of restenosis (8.3% versus 51.1%;
P ! .001) and a 91% reduction in target vessel re-
occlusion (2.1% versus 23.4%; P ! .001). The
1-year MACE rate was reduced by 74% (12.5%
versus 47.9%; P ! .001). The Wisdom registry re-
ported the results in 65 patients who had CTO
treated with TAXUS stents. At 12 months 6.7%
of patients had any MACE, and these were all tar-
get vessel revascularizations, with one late stent
thrombosis. Although the numbers are still small,
and no controlled clinical trials have been con-
ducted, there seems to be a striking reduction in cu-
mulative late events to less than 10% when CTOs
are treated with drug-eluting stents.
Summary
CTOs are prevalent in the coronary artery
disease population but account for only 6% to
 CHRONIC TOTAL CORONARY OCCLUSIONS
10% of coronary interventions. Clinical reasons
to open CTOs include improved survival in those
in whom the PCI is successful as well as im-
provement in angina, improved left ventricular
function, and decreased need for CABG. Despite
the high prevalence and clinical imperative to
open CTOs, these patients continue to be the most
problematic in interventional cardiology. Because
of the nature of the pathobiology, these lesions are
difficult to cross with a guidewire and sometimes
with a balloon catheter. New guidewire technol-
ogy has improved the success in crossing these
lesions as well as the time required to cross. Some
newer technology has improved the crossability of
guidewire-refractory lesions. CTOs, however, still
remain time consuming and difficult to open with
low but sometimes significant acute complica-
tions. The restenosis rates and late occlusion rates
are high. With the advent of stenting, there was
a marked improvement in both late restenosis and
late patency rates. Although the numbers are
small, the use of drug-eluting stents has incremen-
tally improved the long-term results with reste-
nosis rates approaching 10% and late reocclusion
rates in the low single-digit range.
References
[1] Stone G, Kandzari D, Mehran R, et al. Percutane-
ous recanalization of chronically occluded coronary
arteries: a consensus document: part I. Circulation
2005;112:2364–72.
[2] Werner G, Emig U, Mutschke O, et al. Regression if
collateral function after recanalization of chronic
total coronary occlusions: a serial assessment by
intracoronary pressure and Doppler recordings.
Circulation 2003;108:2877–82.
[3] Tamai H, Berger P, Tsuchikane E, et al, for the
Magic investigators. Frequency and time course of
reocclusion and restenosis in coronary artery occlu-
sions after balloon angioplasty versus Wiktor stent
implantation. Am Heart J 2004;147:E9.
[4] Zidar F, Kaplan B, O’Neill W, et al. Prospective,
randomized trial of prolonged intracoronary uroki-
nase infusion for chronic total occlusions on native
coronary arteries. J Am Coll Cardiol 1996;27:
1406–12.
[5] Olivari Z, Rubartelli P, Piscione F, et al, for the
TOAST-GISE investigators. Immediate and one-
year clinical outcome after percutaneous coronary
interventions in chronic total occlusions: data from
a multicenter, prospective, observational study
(TOAST-GISE). J Am Coll Cardiol 2003;41:1672–8.
[6] King S, Lembo N, Weintraub W, et al. A random-
ized trial comparing coronary angioplasty with
253
coronary bypass surgery. N Engl J Med 1994;331:
1044–50.
[7] Hoher M, Wohrle J, Grebe O, et al. A randomized
trial of elective stenting after balloon recanalization
of chronic total occlusions. J Am Coll Cardiol 1999;
34:722–9.
[8] Rubartelli P, Verna E, Niccoli L, et al, for the
Gruppo Italiano de Studio sullo Stent nelle Occlu-
sioni Coronariche investigators. Coronary stent
implantation is superior to balloon angioplasty
for chronic coronary occlusions: six-year clinical
follow-up of the GISSOC trial. J Am Coll Cardiol
2003;41:1488–92.
[9] Buller C, Dzavik V, Carere R, et al. Primary stenting
versus balloon angioplasty in occluded coronary
arteries: the Total Occlusion Study of Canada
(TOSCA). Circulation 1999;100:236–42.
[10] Kahn J. Angiographic suitability for catheter revas-
cularization of total coronary occlusions in patients
from a community hospital setting. Am Heart J
1993;126:561–4.
[11] Cohen H, Williams D, Holmes D, et al. Impact of
age on procedural and 1-year outcome in percutane-
ous transluminal coronary angioplasty: the NHLBI
Dynamic Registry. Am Heart J 2003;146:512–9.
[12] Hamm C, Reimers J, Ischinger T, et al, for the Ger-
man Angioplasty Bypass Surgery Investigation.
A randomized study of coronary angioplasty com-
pared with bypass surgery in patients with symptom-
atic multivessel coronary disease. N Engl J Med
1994;331:1037–43.
[13] Bourassa M, Roubin G, Detre K, et al. Bypass An-
gioplasty Revascularization Investigation: patient
screening, selection, and recruitment. Am J Cardiol
1995;75:3C–8C.
[14] Anderson H, Shaw R, Brindis R, et al. A contempo-
rary overview of percutaneous coronary interventions:
the American College of Cardiology-National Car-
diovascular Data Registry (ACC-NCDR). J Am
Coll Cardiol 2002;39:1096–103.
[15] Suero J, Marso S, Jones P, et al. Procedural
outcomes and long-term survival among patients
undergoing percutaneous coronary intervention of
a chronic total occlusion in native coronary arteries:
a 20-year experience. J Am Coll Cardiol 2001;38:
409–14.
[16] Ranmanathan K, Gao M, Nogareda G, et al. Suc-
cessful percutaneous recanalization of a non-acute
colluded coronary artery predicts clinical outcome
and survival. Circulation 2001;104:II–415a.
[17] Hoye A, van Domburg R, Sonnenschein K, et al.
Percutaneous coronary intervention for chronic to-
tal occlusions: the Thorax Center experience 1992–
2002. Eur Heart J 2005;26:2630–6.
[18] Dzavik V, Caere R, Mancini G, et al, for the Total
Occlusion Study of Canada Investigators. Predictors
of improvement in left ventricular function after per-
cutaneous revascularization of occluded coronary
arteries. Am Heart J 2001;142:301–8.
254 BRADEN
[19] Simes P, Myreng Y, Malsted P, et al. Improvement
of left ventricular ejection fraction and wall motion
after successful recanalization of chronic coronary
occlusions. Eur Heart J 1988;2:273–81.
[20] Chung C. Effect of recanalization of chronic total
occlusions on global and regional left ventricular
function in patients with and without previous myo-
cardial infarction. Cath Cardiovasc Interv 2003;
60(3):368–74.
[21] Srivatsa S, Edwards W, Boos C, et al. Histologic cor-
relates of angiographic chronic total coronary occlu-
sions influence of occlusion duration an neovascular
channel patterns an intimal plaque composition.
J Am Coll Cardiol 1997;29:955–63.
[22] Katsuragawa M, Fujiware H, Miyamae M, et al.
Histologic studies in percutaneous transluminal cor-
onary angioplasty for chronic total occlusion: com-
parison of tapering and abrupt types of occlusion
and short and long occluded segments. J Am Coll
Cardiol 1993;21:604–11.
[23] Burke A, Kolodgie F, Farb A, et al. Morphological
predictors of arterial remodeling in coronary athero-
sclerosis. Circulation 2002;105:297–303.
[24] Colombo A, Mikhail G, Michev I, et al. Treating
chronic total occlusion using subintimal tracking
and reentry: the STAR technique. Catheter Cardio-
vasc Interv 2005;64:407–11.
[25] Baim D, Braden G, Heuser R, et al. Utility of the
Safe-Cross-guided radio frequency total occlusion
crossing system in chronic coronary total occlusions.
Am J Cardiol 2004;94:853–8.
[26] Whitlow P, Selmon M, O’Neill W, et al. Treatment
of uncrossable chronic total coronary occlusions
with the Frontrunner: multicenter experience. J Am
Coll Cardiol 2002;90:168H.
[27] Michalis L, Rees M, Davis J, et al. Use of vibrational
angioplasty for the treatment of chronic total coro-
nary occlusions. Catheter Cardiovasc Interv 1999;
46:98–104.
[28] Serruys P, Hamburger J, Koolen J, et al. Total occlu-
sion trial with angioplasty by using laser guidewire.
Eur Heart J 2000;21:1797–805.
[29] Hoye A, Tanabe K, Lemos P, et al. Significant
reduction in restenosis after the use of sirolimus-
eluting stents in the treatment of chronic total
occlusions. J Am Coll Cardiol 2004;43:1954–8.
[30] Nakamura S, Selvan TS, Bae JH, et al. Impact of
sirolimus-eluting stents on the outcome of patients
with chronic total occlusions: multicenter registry
in Asia. J Am Coll Cardiol 2004;43:35A.
[31] Werner GS, Krack A, Schwarz G, et al. Prevention
of lesion recurrence in chronic total coronary occlu-
sions by paclitaxel-eluting stents. J Am Coll Cardiol
2004;44:2301–6.
 Cardiol Clin 24 (2006) 255–263

Percutaneous Coronary Interventions: Guidelines,

Short- and Long-Term Results, and Comparison
with Coronary Artery Bypass Grafting
Joaquin E. Cigarroa, MD, L. David Hillis, MD*
Department of Internal Medicine, Cardiovascular Division, University of Texas Southwestern Medical Center,
5323 Harry Hines Boulevard, Dallas, TX 75390-9030, USA

In the United States and throughout the
western world, the prevalence of ischemic heart
disease is high due to a relatively high incidence
among the population of advanced age, obesity,
and certain risk factors for atherosclerosis (most
notably hypertension and diabetes mellitus). The
widespread use of noninvasive testing for identi-
fying those who are likely to have coronary artery
disease (ie, exercise testing with or without echo-
cardiographic or radionuclide imaging and the
more recently developed MRI and CT angiogra-
phy) has resulted in a substantial increase in the
number of patients undergoing diagnostic coro-
nary angiography, which in turn has resulted in an
increased number of individuals who are referred
for percutaneous or surgical coronary revascular-
ization. In the United States in 2002, according to
the American Heart Association, approximately
1.2 million people had a percutaneous coronary
intervention (PCI) and 0.5 million underwent
coronary artery bypass grafting (CABG). In in-
dividuals who have coronary artery disease,
coronary revascularization may be performed
(1) to reduce morbidity by eliminating or reducing
anginal frequency and severity or, on occasion, by
alleviating symptoms of heart failure (dyspnea or
fatigue); or (2) to reduce mortality in selected
patient subsets. The decision to proceed with
percutaneous or surgical revascularization should
be based on a thorough and complete
* Corresponding author.
E-mail address: dhilli@parknet.pmh.org
(L.D. Hillis).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.01.002
understanding of the short- and long-term risks
and benefits of each procedure in conjunction
with the individual patient’s coronary arterial
anatomy and clinical risk profile.
Coronary artery bypass grafting
CABG was first reported in 1969 by Favaloro
[1]. Subsequent randomized comparisons in the
1970s of CABG and medical therapy in subjects
who had stable or unstable angina [2–6] estab-
lished the superiority of surgical revascularization
in relieving angina, improving exercise tolerance,
and reducing the need for antianginal medica-
tions. In addition, apart from symptom relief,
CABG was superior to medical therapy in impro-
ving survival in subjects who had (1) three-vessel
coronary artery disease and a left ventricular ejec-
tion fraction less than 0.50, (2) two- or three-vessel
coronary artery disease in which the proximal left
anterior descending coronary artery was signifi-
cantly narrowed [4], and (3) left main coronary
artery disease.
Percutaneous coronary intervention
The first PCI was reported by Gruntzig and
colleagues in 1979 [7]. Over the next several years,
substantial improvements in equipment, including
the design and manufacture of guiding catheters,
guidewires, and balloon catheters, led to the use
of this therapeutic modality in a growing number
of patients so that by the mid- to late-1980s, its
use was widespread. Small randomized trials dem-
onstrated that PCI was superior to medical
cardiology.theclinics.com
256 CIGARROA & HILLIS

therapy in the alleviation of angina [8]. At the angina or other evidence of myocardial ischemia,
same time, these studies failed to demonstrate (2) multivessel coronary artery disease, and
that PCI reduced the occurrence of subsequent is- (3) coronary arterial anatomic features that made
chemic events or death and established the rela- them eligible for either procedure were randomly
tively high incidence (35%–40%) of a required assigned to receive CABG (n ¼ 914) or PTCA
repeat revascularization procedure during the sub- (n ¼ 915) (Table 1).
sequent year in patients whose initial PCI had Of the 914 patients who were assigned to
been successful, a process known as restenosis. receive CABG, 98% received it. All intended
The procedural success and safety of PCI in pa- vessels were grafted in 91% of patients, and in
tients who had single-vessel coronary artery dis- 82%, the left internal mammary artery was used
ease led to its application in patients who had as one of the conduits. The median hospital stay
multivessel disease. Subsequently, the growing following CABG was 7 days. Of the 915 patients
use of PCI in this patient population led to the assigned to receive PTCA, 99% received it.
performance of randomized comparisons of PCI Balloon angioplasty of multiple coronary arterial
and CABG [9–17]. stenoses was attempted in 78% of patients and
resulted in an average of 1.9 of 3.5 (54%)
clinically important stenoses dilated successfully.
Coronary artery bypass grafting versus The median hospital stay after PTCA was 3 days.
percutaneous coronary intervention in patients With both treatment modalities, the in-hospital
who have stable or unstable angina mortality was less than 1.5% (Table 2). PTCA re-
sulted in a lower incidence of periprocedural Q
Coronary artery bypass grafting versus
wave myocardial infarction (2.1% for PTCA,
percutaneous transluminal coronary angioplasty
4.5% for CABG, P!0.01) but was accompanied
From the mid-1980s to the mid-1990s, nine by the need for more subsequent urgent revascu-
randomized comparisons of CABG and percuta- larization procedures (8.3% for PTCA, 0.1% for
neous transluminal coronary angioplasty (PTCA; CABG, P!0.001) and nonurgent revasculariza-
balloon angioplasty) in patients who had single- tion procedures (5.1% for PTCA, 0% for
or multivessel coronary artery disease were re-
porteddall of which produced remarkably similar
and consistent results [9–17]. First, the two Table 1
methods of revascularization were associated Characteristics of the patients assigned to undergo per-
with a similarly low periprocedural mortality cutaneous transluminal coronary angioplasty or coro-
(1.0%–1.5%). Second, the periprocedural morbid- nary artery bypass grafting in the Bypass Angioplasty
ity that occurred with CABG was higher (the inci- Revascularization Investigation
dence of periprocedural Q wave myocardial CABG PTCA
infarction ranged from 4.6% to 10.3% with Characteristic (n ¼ 914) (n ¼ 915)
CABG and only 2.1% to 6.3% with PTCA) and Mean age (y) 61.1 61.8
the length of hospital stay was longer with Female sex (%) 26 27
CABG than with PTCA. Third, those who under- Previous myocardial 55 54
went CABG experienced more effective relief of infarction (%)
angina. As a result, they had less need for antian- Congestive heart failure (%) 9 9
ginal medications and were less likely to require Treated diabetes 25 24
a repeat revascularization procedure. Lastly, mellitus (%)
Unstable angina (%) 65 63
long-term survival was similar for the two treat-
Three-vessel CAD (%) 41 41
ment strategies.
Proximal LAD 37 36
A subsequent analysis of the results of the involvement (%)
largest of these trials, the Bypass Angioplasty LV ejection 21 23
Revascularization Investigation (BARI) [14], fraction !0.50 (%)
strongly suggested that long-term survival was
Abbreviations: CAD, coronary artery disease; LAD,
better with CABG than with PTCA in patients left anterior descending; LV, left ventricular.
who had treated diabetes mellitus, whereas sur- Data from Comparison of coronary bypass surgery
vival with the two treatment strategies was similar with angioplasty in patients with multivessel disease.
in those who did not have treated diabetes melli- The Bypass Angioplasty Revascularization Investigation
tus. In BARI [14], 1829 patients who had (1) (BARI) investigators. N Engl J Med 1996;335:219.
 PCI VERSUS CABG
Table 2
In-hospital complications of coronary artery bypass
grafting and percutaneous transluminal coronary angio-
plasty in the Bypass Angioplasty Revascularization
Investigation
Complication CABG (n ¼ 914) PTCA (n ¼ 915)
Death 12 (1.3%) 10 (1.1%)
Q wave MI 41 (4.5%) 19 (2.1%)*
Death/Q wave MI 52 (5.7%) 27 (3.0%)*
Urgent CABG 1 (0.1%) 57 (6.2%)**
Urgent PTCA 0 19 (2.1%)**
Stroke 7 (0.8%) 2 (0.2%)
Nonurgent 0 47 (5.1%)**
CABG/PTCA
Abbreviation: MI, myocardial infarction.
* P!0.01; ** P!0.001 in comparison to CABG.
Data from Comparison of coronary bypass surgery
with angioplasty in patients with multivessel disease.
The Bypass Angioplasty Revascularization Investigation
(BARI) investigators. N Engl J Med 1996;335:220.
CABG, P!0.001). Procedural morbidity includ-
ing respiratory failure, reoperation for bleeding,
and wound infection was higher for CABG
(9.4% for CABG, 1.8% for PTCA, P!0.001).
In summary, CABG and PTCA were accompa-
nied by a low periprocedural mortality. In com-
parison to PTCA, CABG achieved more
complete revascularization, thereby resulting in
less need for antianginal medications and repeat
revascularization procedures, but was associated
with a higher incidence of periprocedural Q
wave myocardial infarction and a longer length
of hospital stay (see Table 2).
The 1829 subjects enrolled in BARI were
followed for an average of 5.4 years (Table 3).
In comparison to those who had CABG, those
who had PTCA were more likely to require a re-
peat revascularization procedure during this pe-
riod of follow-up (54% for PTCA, 8% for
CABG, P!0.001). The incidence of myocardial
infarction and death was similar for the two treat-
ment strategies (myocardial infarction: 21% for
CABG, 20% for PTCA, not significant [NS];
death: 14% for CABG, 11% for PTCA, NS). In
patients who were undergoing treatment for dia-
betes mellitus, a post hoc analysis revealed a sub-
stantial survival advantage with CABG (total
mortality 19.4% for CABG, 34.5% for PTCA,
P!0.003).
Subsequent analyses of ‘‘registry data’’ from
nonrandomized groups of patients who had multi-
vessel coronary artery disease and diabetes
257
Table 3
Results of the Bypass Angioplasty Revascularization
Investigation
CABG PTCA
Result (n ¼ 914) (n ¼ 915)
In-hospital mortality 1.3% 1.1%
Periprocedural Q wave MI 4.5% 2.1%*
Periprocedural stroke 0.8% 0.2%
5-y survival (all patients) 89.3% 86.3%
5-y survival free of MI 80.4% 78.7%
Repeat revascularization 8.0% 54.0%**
within 5 y
5-y survival 80.6% 65.5%***
(treated diabetics)
Abbreviation: MI, myocardial infarction.
* P!0.01; ** P!0.001; *** P!0.003 in compari-
son to CABG.
Data from Comparison of coronary bypass surgery
with angioplasty in patients with multivessel disease.
The Bypass Angioplasty Revascularization Investigation
(BARI) investigators. N Engl J Med 1996;335:220.
mellitus have provided mixed results in compar-
ison to the results of the BARI randomized trial.
Similar to BARI, some of these analyses demon-
strated a survival advantage of CABG over PTCA
in this patient population [18], whereas others
showed that this population’s long-term mortality
was similar with CABG or with PTCA [19]. Even
the nonrandomized registry data from the BARI
investigators showed no difference in long-term
mortality between the two treatment groups [20].
These disparate results are likely due to differences
in coronary anatomy between the patients en-
rolled in the registry and those in the randomized
trial.
As noted, previously published studies have
demonstrated that CABG is superior to medical
therapy in improving survival in patients who had
three-vessel coronary artery disease and a left
ventricular ejection fraction less than 0.50 [2,3]
and who had two- or three-vessel coronary artery
disease with normal left ventricular systolic func-
tion if the proximal left anterior descending coro-
nary artery is significantly narrowed [4]. Are the
long-term results of PTCA similar to those of
CABG in these specific patient populations? The
answer appears to be yes. Berger and colleagues
[21] examined the results of BARI in patients
who had these specific anatomic features and who
did not have diabetes mellitus and provided actu-
arial survival data over a period of observation of
7 years. For all patients who had three-vessel
258 CIGARROA & HILLIS

coronary artery disease, survival was similar at Coronary artery bypass grafting versus stenting
7 years for CABG (87%) and PTCA (85%)
By the mid-1990s, intracoronary stenting had
(Fig. 1). For those who had three-vessel coronary
largely supplanted PTCA in most patients un-
artery disease and a left ventricular ejection frac-
dergoing percutaneous coronary revasculariza-
tion less than 0.50, survival at 7 years for CABG
tion. In comparison to PTCA, stenting offers
or PTCA was similar (73% for CABG, 82% for
several advantages including (1) a reduced in-
PTCA, NS) (Fig. 2). For patients who had two-
cidence of symptomatic restenosis (35%–40%
vessel coronary artery disease and significant nar-
with PTCA, 20%–25% with stenting); (2) a re-
rowing of the proximal left anterior descending
duced need for urgent CABG; and (3) a decreased
coronary artery, survival at 7 years was similar
procedural time. The acceptance of intracoronary
for the two treatment groups (86% for CABG,
stenting as the preferred method of PCI led to
93% for PTCA, NS) (Fig. 3). Finally, for those
several randomized comparisons of stenting and
who had two-vessel coronary artery disease involv-
CABG in patients who had single- and multivessel
ing the proximal left anterior descending coronary
coronary artery disease [22–25].
artery and a left ventricular ejection fraction less
From April 1997 to June 1998, Serruys and
than 0.50, survival at 7 years was statistically sim-
colleagues [23] randomly assigned 1205 patients
ilar for those undergoing CABG and PTCA (67%
who had stable or unstable angina and multivessel
and 90%, respectively, NS) (Fig. 4).
coronary artery disease to CABG (n ¼ 605) or
In summary, the nine randomized comparisons
stenting (n ¼ 600). Two thirds of the patients
of CABG and PTCA [9–17] showed that (1) the
had two-vessel coronary artery disease and the re-
two procedures are associated with a similar peri-
mainder had three-vessel disease. Only 4% of sub-
procedural and long-term mortality, but the data
jects had an occluded (as opposed to narrowed)
from the randomized portion of BARI suggest
major epicardial coronary artery. Of the patients
strongly that long-term survival is better with
who were randomly assigned to stenting, an aver-
CABG in patients who have treated diabetes mel-
age of 2.8 coronary arterial stenoses were noted
litus; and (2) although CABG is accompanied
on angiography, and an average of 2.6 of these
by greater periprocedural morbidity, it affords
were stented (in 89%) or dilated with a balloon
greater angina relief and less of a need for subse-
(in 11%). Of the individuals randomly assigned
quent urgent and elective repeat revascularization
to undergo CABG, an average of 2.8 coronary
procedures.

Fig. 1. Actuarial survival, in years, expressed as a percentage, in nondiabetic patients who had three-vessel coronary
artery disease. At 7 years, the survival in those undergoing CABG and PTCA was similar. (From Berger PB, Velianou
JL, Aslanidou Vlachos H, et al. Survival following coronary angioplasty versus coronary artery bypass surgery in ana-
tomic subsets in which coronary artery bypass surgery improves survival compared with medical therapy. Results from
the Bypass Angioplasty Revascularization Investigation (BARI). J Am Coll Cardiol 2001;38:1445; with permission.)
 PCI VERSUS CABG 259

Fig. 2. Actuarial survival, in years, expressed as a percentage, in nondiabetic patients who had three-vessel coronary
artery disease and depressed left ventricular systolic function. At 7 years, the survival in those having CABG and
PTCA was similar. (From Berger PB, Velianou JL, Aslanidou Vlachos H, et al. Survival following coronary angioplasty
versus coronary artery bypass surgery in anatomic subsets in which coronary artery bypass surgery improves survival
compared with medical therapy. Results from the Bypass Angioplasty Revascularization Investigation (BARI). J Am
Coll Cardiol 2001;38:1445; with permission.)

arterial stenoses were noted, and these patients re-
ceived an average of 2.5 conduits; 93% of this
group received at least one arterial conduit.
In comparison to those undergoing stenting,
the patients treated with CABG had a higher
incidence of periprocedural myocardial infarction,
defined as a rise in serum MB fraction of creatine
kinase (CK-MB) greater than five times the upper
limit of normal (12.6% for CABG, 6.2% for
stenting, P!0.001). The incidence of subsequent
unplanned revascularization procedures was sta-
tistically similar (0.3% for CABG, 2.3% for

Fig. 3. Actuarial survival, in years, expressed as a percentage, in nondiabetic patients who had two-vessel coronary ar-
tery disease and significant narrowing of the proximal left anterior descending coronary artery. At 7 years, the survival in
those undergoing CABG and PTCA was similar. (From Berger PB, Velianou JL, Aslanidou Vlachos H, et al. Survival
following coronary angioplasty versus coronary artery bypass surgery in anatomic subsets in which coronary artery by-
pass surgery improves survival compared with medical therapy. Results from the Bypass Angioplasty Revascularization
Investigation (BARI). J Am Coll Cardiol 2001;38:1447; with permission.)
260 CIGARROA & HILLIS

Fig. 4. Actuarial survival, in years, expressed as a percentage, in nondiabetic patients who had two-vessel coronary ar-
tery disease, significant narrowing of the proximal left anterior descending coronary artery, and depressed left ventricular
systolic function. At 7 years, the survival in those having CABG and PTCA was similar. (From Berger PB, Velianou JL,
Aslanidou Vlachos H, et al. Survival following coronary angioplasty versus coronary artery bypass surgery in anatomic
subsets in which coronary artery bypass surgery improves survival compared with medical therapy. Results from the
Bypass Angioplasty Revascularization Investigation (BARI). J Am Coll Cardiol 2001;38:1447; with permission.)

stenting, NS). During the first month after stent- had CABG to experience an adverse outcome, de-
ing, subacute stent thrombosis occurred in 2.8%. fined as the occurrence of death, myocardial in-
As shown in Table 4, the incidence of death, farction, stroke, or a repeat revascularization
stroke, and myocardial infarction was similar in procedure. Specifically, in the 5 years following
the two treatment groups 1 year after randomiza- randomization, 78.2% of those who underwent
tion. During this first year of observation, how- CABG had none of these adverse outcomes,
ever, those who had stenting were more likely to whereas only 58.3% of those who had stenting
require a second revascularization procedure were free of them (P!0.001). This difference be-
(21.0% for stenting, 3.8% for CABG, P!0.01). tween the two treatment groups was largely driven
The patients who received CABG were more by the substantial difference in the need for a re-
likely to be free of angina (89.5% for CABG, peat revascularization procedure in those under-
78.9% for stenting, P!0.001). going stenting (41.7% for stenting, 21.8% for
As displayed in Table 5, the data acquired after
5 years of follow-up demonstrate that the subjects
who had stenting were more likely than those who Table 5
Arterial Revascularization Therapies I trial 5-y out-
comes: coronary artery bypass grafting versus stenting
Table 4
CABG Stenting
Arterial Revascularization Therapies I trial 1-y out-
Outcome (n ¼ 605) (n ¼ 600)
comes: coronary artery bypass grafting versus stenting
Freedom from death/MI/ 78.2% 58.3%*
CABG Stenting
CVA/revascularization
Outcome (n ¼ 605) (n ¼ 600)
Death 7.6% 8.0%
Death 2.8% 2.5% Repeat revascularization 21.8% 41.7%*
Cerebrovascular accident 2.1% 1.7% Angina 15.5% 21.2%
Myocardial infarction 4.8% 6.2%
Abbreviations: CVA, cerebrovascular accident; MI,
Repeat Revascularization 3.8% 21.0%*
myocardial infarction.
* P!0.01 in comparison to CABG. * P!0.001 in comparison to CABG.
Data from Serruys PW, Unger F, Sousa JE, et al. Data from Serruys PW, Unger F, Sousa JE, et al.
Comparison of coronary-artery bypass surgery and Comparison of coronary-artery bypass surgery and
stenting for the treatment of multivessel disease. N Engl J stenting for the treatment of multivessel disease. N Engl J
Med 2001;344:1119. Med 2001;344:1119.
 PCI VERSUS CABG
CABG, P!0.001). During the 5 years of observa-
tion, mortality was similar in the two groups
(7.6% for CABG, 8.0% for stenting, NS), as
was the likelihood of continued angina (21.2%
for stenting, 15.5% for CABG, NS) and the
need for antianginal medications.
Serruys and colleagues [23] studied the mone-
tary costs of the two treatment strategies. The
cost of the initial procedure averaged $4212 less
for stenting than for CABG ($6441 for stenting,
$10,653 for CABG). Although this advantage in
favor of stenting persisted at 1 year, eventually it
was reduced in magnitude (to $2973 per patient)
due to a greater need for repeat revascularization
procedures in the stented group. Similar short-
and long-term cost advantages of stenting over
CABG have been shown to be present in the Stent
or Surgery Trial [26].
In summary, the randomized comparisons of
CABG and coronary stenting provided the fol-
lowing results. First, the two procedures were
associated with a similar periprocedural and long-
term mortality. Second, although CABG was
accompanied by greater periprocedural morbid-
ity, it afforded better angina relief and less need
for subsequent urgent or elective repeat revascu-
larization procedures. Third, those who had
stenting had a shorter length of hospital stay
and a smaller hospital bill. In comparison to
PTCA, coronary stenting improves the complete-
ness of revascularization, reduces the need for
urgent CABG, and decreases the need for sub-
sequent revascularization by decreasing the in-
cidence of restenosis. In addition, stenting
provides the operator with the ability to address
complex coronary arterial stenoses (ie, those that
are particularly long or angulated and those
located at a coronary arterial bifurcation) with
less uncertainty in the outcome.
In all the previously cited comparisons of
CABG and PCI, the consistent (and only) disad-
vantage of the latter has been the substantial
percentage of patients who have recurrent symp-
toms in the weeks to months after the initially
successful procedure, thereby requiring repeat
coronary angiography and, in most individuals,
an additional revascularization procedure (CABG
or repeat PCI). The recurrence of symptoms is
usually caused by restenosis at the site of previous
balloon dilatation or stent deployment. As stated
previously, symptomatic restenosis occurs in about
35% of patients in the 12 months after successful
PTCA and in 20% to 25% of those who have had
successful stenting. If one could effectively prevent
261
restenosis, this one major disadvantage of PCI
would be alleviated, thereby making PCI more
attractive than CABG (provided that contraindi-
cations to PCI do not exist). Recently published
data [26] demonstrate that the use of sirolimus-
eluting stents dramatically reduces the incidence
of restenosis. Specifically, of 238 patients ran-
domly assigned to receive a standard (n ¼ 118)
or a sirolimus-eluting (n ¼ 120) stent, 27% of the
former and 0% of the latter group developed reste-
nosis within 6 months of the procedure (P!0.001).
These striking results await confirmation in studies
of patients who have multivessel coronary artery
disease. The results of the Arterial Revasculariza-
tion Therapies II trial, a nonrandomized open-label
registry that compared the results of drug-eluting
stent deployment in 607 subjects with the results
obtained in a group of individuals who underwent
CABG and whose outcomes were described pre-
viously in the literature (so-called ‘‘historical
controls’’), suggest that the two procedures are
accompanied by a similar incidence of myocardial
infarction, required repeat revascularization pro-
cedures, and death during 1 year of observation
(Table 6) [27].
Limitations of randomized trials comparing
coronary artery bypass grafting and percutaneous
coronary intervention
The multiple randomized trials comparing
CABG and PCI described previously have
Table 6
Arterial Revascularization Therapies II trial 1-y clinical
outcomes
Drug-eluting CABG
Outcome stents (n ¼ 607) (n ¼ 605)
Death 1.0% 2.7%
MI 1.2% 3.5%
Cerebrovascular 0.8% 1.8%
accident
CABG 2.0% 0.7%
PCI 5.4% 3.0%
Death/MI/CVA/ 10.4% 11.6%
revascularization
Abbreviations: CVA, cerebrovascular accident; MI,
myocardial infarction.
Data from Serruys PW. Preliminary results of the
second Arterial Revascularization Therapy Study trial
(ARTS II). Presented at the American College of Cardi-
ology Annual Scientific Sessions, March 6, 2005, Orlando,
Florida.
262 CIGARROA & HILLIS
provided valuable insights into the relative bene-
fits and limitations of each procedure. By elimi-
nating (or at least minimizing) physician and
patient ‘‘selection bias,’’ these trials help to ensure
that patient comorbidities and concurrent man-
agement strategies are distributed similarly be-
tween the two revascularization techniques.
Nonetheless, they have limitations. First, only
a small fraction (5%–12%) of patients who were
evaluated for possible enrollment in these ran-
domized trails were found to be eligible; therefore,
the results may not be easily generalized and
applicable to the much larger group of patients
requiring revascularization (most of whom would
have been ineligible for enrollment). Second, most
of these randomized trials were not of sufficient
statistical power to allow definitive conclusions
regarding survival. Data from large registries,
although not randomized, avoid the aforemen-
tioned potential limitations of randomized trials.
In this regard, the results of the New York
Registry, recently reported by Hannan and col-
leagues [28], are of considerable interest. In this
analysis of data from almost 60,000 patients
who had multivessel coronary artery disease un-
dergoing CABG (n ¼ 37,212) or PCI (n ¼
22,102) between January 1997 and December
2000, those with two- or three-vessel disease had
a better long-term survival with CABG than
with PCI. The survival advantage of CABG over
PCI was demonstrable in all patient subgroups
who had multivessel coronary artery disease, irre-
spective of (1) the location of the stenosis in the
left anterior descending coronary artery (proximal
or nonproximal); (2) the presence or absence of di-
abetes mellitus; and (3) left ventricular systolic
function.
Implications for patient management
The management of the patient who has
(1) single-vessel coronary artery disease (regardless
of left ventricular systolic function); (2) two-vessel
disease not involving the proximal left anterior
descending coronary artery (regardless of left
ventricular systolic function); and (3) three-vessel
disease and normal left ventricular systolic func-
tion can be individualized and ‘‘symptom-driven.’’
For some of these patients, medical therapy will
suffice, in that it will substantially or completely
alleviate symptoms. When nonmedical treatment
is chosen, the decision of whether to use percuta-
neous or surgical revascularization should be
reached ‘‘with the understanding that CABG is
associated with greater initial morbidity but results
in more effective relief of angina and freedom from
repeat procedures in the ensuing years. On the
other hand, PCI is associated with a lower rate of
initial morbidity but a greater likelihood of re-
current angina, need for antianginal medications,
and subsequent revascularization procedures’’
[29]. The use of drug-eluting stents is likely to re-
duce the need for subsequent revascularization
procedures.
The patient who has three-vessel coronary
artery disease and depressed left ventricular sys-
tolic function (ejection fraction !0.50) or two- or
three-vessel disease with narrowing of the proxi-
mal left anterior descending coronary artery (re-
gardless of left ventricular systolic function)
should be encouraged to undergo percutaneous
or surgical revascularization (irrespective of symp-
toms) to achieve an improved survival. If the
patient is diabetic, then CABG is probably pre-
ferred (based on the results of the randomized
portion of BARI). If the patient is not diabetic
and his or her coronary anatomy is amenable to
either therapeutic strategy, then the decision of
whether to use percutaneous or surgical therapy
can be based on the patient’s preferences, with
appropriate input from his or her physician.
References
[1] Favaloro RG. Saphenous vein graft in the surgical
treatment of coronary artery disease. Operative
technique. J Thorac Cardiovasc Surg 1969;58:
178–85.
[2] The Veterans administration Coronary artery
Bypass Surgery Cooperative Study Group. Eleven-
year survival in the Veterans Administration ran-
domized trial of coronary bypass surgery for stable
angina. The Veterans Administration Coronary
Artery Bypass Surgery Cooperative Study Group.
N Engl J Med 1984;311:1333–9.
[3] Alderman EL, Bourassa MG, Cohen LS, et al. Ten-
year follow-up of survival and myocardial infarction
in the randomized Coronary Artery Surgery Study.
Circulation 1990;82:1629–46.
[4] Varnauskas E. Twelve-year follow-up of survival in
the randomized European Coronary Surgery Study.
N Engl J Med 1988;319:332–7.
[5] Booth DC, Deupree RH, Hultgren HN, et al.
Quality of life after bypass surgery for unstable
angina. 5-year follow-up results of a Veterans Affairs
Cooperative Study. Circulation 1991;83:87–95.
[6] Russell RO Jr, Moraski RE, Kouchoukos N, et al.
Unstable angina pectoris: National Cooperative
Study Group to Compare Surgical and Medical
Therapy. Am J Cardiol 1978;42:839–48.
 PCI VERSUS CABG
[7] Gruntzig AR, Senning A, Siegenthaler WE. Nonop-
erative dilatation of coronary-artery stenosis: percu-
taneous transluminal coronary angioplasty. N Engl
J Med 1979;301:61–8. [19]
[8] Parisi AF, Folland ED, Hartigan P. A comparison
of angioplasty with medical therapy in the treatment
of single-vessel coronary artery disease. Veterans Af-
fairs ACME Investigators. N Engl J Med 1992;326: [20]
10–6.
[9] RITA Trial Participants. Coronary angioplasty ver-
sus coronary artery bypass surgery: the Randomized
Intervention Treatment of Angina (RITA) trial.
Lancet 1993;341:573–80. [21]
[10] Rodriguez A, Boullon F, Perez-Balino N, et al. Ar-
gentine randomized trial of percutaneous translumi-
nal coronary angioplasty versus coronary artery
bypass surgery in multivessel disease (ERACI): in-
hospital results and 1-year follow-up. ERACI
Group. J Am Coll Cardiol 1993;22:1060–7.
[11] Hamm CW, Reimers J, Ischinger T, et al. A random- [22]
ized study of coronary angioplasty compared with
bypass surgery in patients with symptomatic multi-
vessel coronary disease. German Angioplasty By-
pass Surgery Investigation (GABI). N Engl J Med
1994;331:1037–43.
[12] King III SB, Lembo NJ, Weintraub WS, et al. A ran- [23]
domized trial comparing coronary angioplasty with
coronary bypass surgery. Emory Angioplasty versus
Surgery Trial (EAST). N Engl J Med 1994; 331:
1044–50. [24]
[13] CABRI Trial Participants. First-year results of CAB-
RI (Coronary Angioplasty versus Bypass Revascu-
larisation Investigation). CABRI Trial Participants.
Lancet 1995;346:1179–84.
[14] The Bypass angioplasty Revascularization Investi-
gation (BARI) Investigators. Comparison of coro- [25]
nary bypass surgery with angioplasty in patients
with multivessel disease. The Bypass Angioplasty
Revascularization Investigation (BARI) investiga-
tors. N Engl J Med 1996;335:217–25.
[15] Goy JJ, Eeckhout E, Burnand B, et al. Coronary an-
gioplasty versus left internal mammary artery graft- [26]
ing for isolated proximal left anterior descending
artery stenosis. Lancet 1994;343:1449–53.
[16] Carrie D, Elbaz M, Puel J, et al. Five-year outcome
after coronary angioplasty versus bypass surgery in [27]
multivessel coronary artery disease: results from
the French Monocentric Study. Circulation 1997;
96:II-1–6.
[17] Hueb WA, Bellotti G, de Oliveira SA, et al. The
Medicine, Angioplasty or Surgery Study (MASS):
a prospective, randomized trial of medical therapy,
balloon angioplasty or bypass surgery for single [28]
proximal left anterior descending artery stenoses.
J Am Coll Cardiol 1995;26:1600–5.
[18] Niles NW, McGrath PD, Malenka D, et al. Survival
of patients with diabetes and multivessel coronary [29]
artery disease after surgical or percutaneous coro-
nary revascularization: results of a large regional
263
prospective study. Northern New England Cardio-
vascular Disease Study Group. J Am Coll Cardiol
2001;37:1008–15.
Barsness GW, Peterson ED, Ohman EM, et al. Re-
lationship between diabetes mellitus and long-term
survival after coronary bypass and angioplasty. Cir-
culation 1997;96:2551–6.
Detre KM, Guo P, Holubkov R, et al. Coronary re-
vascularization in diabetic patients: a comparison of
the randomized and observational components of
the Bypass Angioplasty Revascularization Investi-
gation (BARI). Circulation 1999;99:633–40.
Berger PB, Velianou JL, Aslanidou Vlachos H, et al.
Survival following coronary angioplasty versus cor-
onary artery bypass surgery in anatomic subsets in
which coronary artery bypass surgery improves sur-
vival compared with medical therapy. Results from
the Bypass Angioplasty Revascularization Investi-
gation (BARI). J Am Coll Cardiol 2001;38:1440–9.
Hueb W, Soares PR, Gersh BJ, et al. The medicine,
angioplasty, or surgery study (MASS-II): a random-
ized, controlled clinical trial of three therapeutic
strategies for multivessel coronary artery disease:
one-year results. J Am Coll Cardiol 2004;43:
1743–51.
Serruys PW, Unger F, Sousa JE, et al. Comparison
of coronary-artery bypass surgery and stenting for
the treatment of multivessel disease. N Engl J Med
2001;344:1117–24.
The SOS Investigators. Coronary artery bypass sur-
gery versus percutaneous coronary intervention with
stent implantation in patients with multivessel coro-
nary artery disease (the Stent or Surgery trial):
a randomised controlled trial. Lancet 2002;360:
965–70.
Rodriguez A, Bernardi V, Navia J, et al. Argentine
randomized study: coronary angioplasty with stent-
ing versus coronary bypass surgery in patients with
multiple-vessel disease (ERACI II): 30-day and
one-year follow-up results. ERACI II Investigators.
J Am Coll Cardiol 2001;37:51–8.
Weintraub WS, Mahoney EM, Zhang Z, et al. One
year comparison of costs of coronary surgery versus
percutaneous coronary intervention in the stent or
surgery trial. Heart 2004;90:782–8.
Serruys PW, for the ARTS-II Investigators. ARTS-
II: Arterial Revascularization Therapies Study Part
II of the sirolimus-eluting stent in the treatment of
patients with multivessel de novo coronary artery le-
sions. Late Breaking Clinical Trial. Presented at the
American College of Cardiology 54th Annual Meet-
ing: Orlando, FL; March 2005.
Hannan EL, Racz MJ, Walford G, et al. Long-
term outcomes of coronary-artery bypass grafting
versus stent implantation. N Engl J Med 2005;
352:2174–83.
Hillis LD, Rutherford JD. Coronary angioplasty
compared with bypass grafting. N Engl J Med
1994;331:1086–7.
 Cardiol Clin 24 (2006) 265–275
Percutaneous Left Ventricular Support Devices
Michael S. Lee, MD, Raj R. Makkar, MD*
Cardiovascular Intervention Center, Cedars-Sinai Medical Center, University of California,
Los Angeles School of Medicine, 8631 West Third Street, Room 415E, Los Angeles, CA 90048, USA
Patients in cardiogenic shock or patients who
have left ventricular dysfunction or complex
coronary lesions such as multivessel coronary
disease, bypass graft disease, or left main disease
who undergo percutaneous coronary intervention
(PCI) are at increased risk of adverse outcomes
from detrimental hemodynamic effects due to
ischemia from balloon inflations, dissection,
abrupt vessel closure, malignant arrhythmias, or
‘‘no reflow’’ [1–3]. Left ventricular assist devices
(LVADs) have been used as therapeutic instru-
ments for cardiac insufficiency including left ven-
tricular failure, cardiogenic shock, and low
cardiac output syndrome and as a bridge to trans-
plantation. Percutaneous LVADs have long been
of major interest to operators to provide partial
or total hemodynamic support to patients in car-
diogenic shock and during high-risk PCI without
the need for surgical implantation.
Expedient initiation of percutaneous circula-
tory support may provide hemodynamic stability
during high-risk PCI. Percutaneous LVADs
through the femoral approach have been devel-
oped to provide circulatory support during high-
risk PCI by withdrawing oxygenated blood from
the left heart into systemic circulation [4–6]. Cur-
rent percutaneous LVADs include the Tandem-
Heart (CardiacAssist, Pittsburgh, Pennsylvania)
and the Impella Recover LP 2.5 System (Impella
CardioSystems, Aachen, Germany), which can
be inserted under fluoroscopy in the cardiac cath-
eterization laboratory without the need for extra-
corporeal oxygenation and surgical implantation
* Corresponding author.
E-mail address: raj.makkar@cshs.org
(R.R. Makkar).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.01.004
and provide temporary hemodynamic stabiliza-
tion during high-risk PCI. In this article, the au-
thors review the current percutaneous LVADs
available for circulatory support.
Cardiogenic shock, which occurs in 7% to
10% of cases after acute myocardial infarction, is
the most common cause of in-hospital death in
patients who have acute myocardial infarction [7].
The intra-aortic balloon pump (IABP) can aug-
ment coronary perfusion in cardiogenic shock
and high-risk PCI patients but provides minimal
circulatory support in the setting of complete he-
modynamic collapse and minimal benefit in mor-
tality [8,9]. The LVAD markedly reduced the
extent of myocardial necrosis in animal models
of acute myocardial infarction [10–12]. The rou-
tine use of the LVAD in acute myocardial infarc-
tion and cardiogenic shock is not practical
because the surgical implantation of the LVAD
requires a midline sternotomy [13]. Furthermore,
patients who had acute myocardial infarction
who required temporary circulatory support
with an LVAD had a very high mortality rate
[14].
With the improvement of device technology
and adjunctive pharmacotherapy, interventional
cardiologists are tackling more complex and high-
risk PCI cases. LVADs provide hemodynamic
support during high-risk PCI, especially during
balloon inflation, by unloading the left ventricle
and augmenting systemic circulation. One of the
earlier devices used to provide circulatory support
during elective PCI was the femorofemoral car-
diopulmonary support system, which required
a perfusionist to direct the system [15]. With ‘‘sup-
ported angioplasty,’’ the femorofemoral cardio-
pulmonary support system was associated with
complications at the cannula site, including
cardiology.theclinics.com
266 LEE & MAKKAR
vascular complications, femoral neuropathy, skin
infection, and skin necrosis. These complications
occurred in equal frequency in percutaneous and
cut-down cannulation. Thirty percent of patients
required blood transfusions, perhaps because the
activated clotting time required to maintain the
system was greater than 400 seconds [16]. Another
LVAD, the Hemopump Cardiac Assist System
(Medtronic, Minneapolis, Minnesota), is no lon-
ger used today because of increased morbidity
and mortality due to large arterial cannulae, he-
molysis, thromboembolism, and the need for anes-
thetists and perfusionists [17,18]. Since then, the
TandemHeart percutaneous LVAD and the Im-
pella Recover LP 2.5 System have been introduced
to provide circulatory support during high-risk
PCI (Table 1).
Intra-aortic balloon pump
The IABP was first used clinically for support-
ing patients who had cardiogenic shock after acute
myocardial infarction [19]. Since then, its use has
continued to increase in a variety of settings, in-
cluding providing mechanical assistance in pa-
tients who have hemodynamic instability during
PCI. The IABP, however, only modestly augments
cardiac output and coronary blood flow and is un-
able to provide total circulatory support when he-
modynamic collapse occurs [20]. Although the
IABP restored epicardial perfusion in animal
models, it had little effect on microvascular flow
in the setting of acute myocardial infarction [21].
Furthermore, the IABP requires a certain level of
left ventricular function. In addition, it has not
been independently associated with improved car-
diac function or able to reverse cardiogenic shock
and provide a survival benefit [22–24].
Hemodynamic effects of the intra-aortic balloon
pump
The effects of the IABP include decreases in
heart rate, left ventricular end-diastolic pressure,
Table 1
Percutaneous ventricular assist devices
Device Pump Speed (rpm) Duration
TandemHeart Centrifugal 3000–7500 Up to 14 d Up to 4 L/min
Impella Recover Axial Up to 50,000 Up to 5 d
LP 2.5 System
mean left atrial pressure, afterload, and myocar-
dial oxygen consumption by at least 20% to 30%
[25]. The IABP also modestly increases coronary
perfusion pressure and decreases the right atrial
pressure, pulmonary artery pressure, and pulmo-
nary vascular resistance. No significant changes
in the mean aortic pressure have been observed,
however, and it only modestly augments cardiac
output [26]. The IABP is also dependent on an in-
trinsic cardiac rhythm. Furthermore, it is ineffec-
tive in patients in cardiac arrest or in patients
who have severe left ventricular dysfunction.
TandemHeart
The TandemHeart percutaneous LVAD
(Fig. 1) is a left atrial-to-femoral bypass system
that can provide rapid short-term circulatory sup-
port with resolution of pulmonary edema and
deranged metabolism in patients who have cardio-
genic shock [27]. In an animal model of acute
myocardial infarction and cardiogenic shock, the
left atrial-to-femoral artery bypass assist device
restored endocardial (microvascular) and epicar-
dial blood flow to baseline [26] and resulted in
a substantial reduction in the infarct size [28].
The TandemHeart has advantages over earlier
LVADs because it can be implanted percutane-
ously and provides rapid hemodynamic support
regardless of the native heart rhythm.
The TandemHeart has been used in a variety
of situations, including in high-risk PCI patients,
in acute myocardial infarction patients in cardio-
genic shock, and in decompensated heart failure
with myocarditis. The TandemHeart has also
been used in cardiac surgery patients to preoper-
atively unload the left ventricle and provide
mechanical circulatory support during the perio-
perative and postoperative period until cardiac
function sufficiently recovers or until an LVAD
can be surgically implanted for long-term hemo-
dynamic support.
There are three subsystems that make up the
TandemHeart system. The first subsystem is a
Cardiac support Anticoagulation Motor
Required Rotor powered
by electromagnetic
coupling
Up to 2.5 L/min Required Integrated electric
motor
 PERCUTANEOUS LEFT VENTRICULAR SUPPORT DEVICES 267

Fig. 2. A 21 F venous transseptal inflow cannula made

Fig. 1. The TandemHeart provides circulatory support
by continuously withdrawing oxygenated blood from
the left atrium by way of a transseptal cannula placed
in the femoral vein. The pump then returns blood by
way of the femoral artery. The hemodynamic effects of
the TandemHeart include an increase in cardiac output
and blood pressure and a decrease in afterload and pre-
load, thus decreasing myocardial oxygen demand.
(Courtesy of CardiacAssist, Pittsburgh, Pennsylvania.)
of polyurethane contains a large end hole and 14 side
holes to allow for aspiration of the oxygenated blood
from the left atrium. A two-stage (14/21 F) dilator is
used to dilate the transseptal puncture after a 0.035-
inch pigtail guidewire is inserted into the left atrium.
The obturator is tapered at its tip to allow for easy inser-
tion into the left atrium. (Courtesy of CardiacAssist,
Pittsburgh, Pennsylvania.)

21 F venous transseptal inflow cannula that is

made of polyurethane (Fig. 2). This cannula has
a curved design at its end to facilitate ideal tip
placement in the left atrium and contains a large
end hole at its distal tip and 14 side holes to aspi-
rate oxygenated blood from the left atrium. The
obturator is tapered at its tip to allow for easy in-
sertion into the left atrium. The cannula is at-
tached to a continuous-flow centrifugal blood
pump. This compact pump contains a hydrody-
namic fluid bearing to support a six-bladed rotat-
ing impeller that provides up to 4 L/min of pump
flow (Fig. 3). Power is supplied by a direct current
brushless electromagnetic motor that operates at
a range of 3000 to 7500 rpm. There are generous Fig. 3. The continuous-flow centrifugal blood pump con-
gaps between the impella and the housing which tains a hydrodynamic fluid bearing to support a six-
permits blood to flow freely with low friction, bladed rotating impeller that provides up to 4 L/min
thereby limiting the generation of heat, hemolysis, of pump flow. (Courtesy of CardiacAssist, Pittsburgh,
and thromboembolism [29]. Blood is then delivered Pennsylvania.)
268 LEE & MAKKAR
from the pump to the femoral artery with an arterial
perfusion catheter. This catheter ranges from 15 to
17 F and pumps blood from the left atrium to the
right femoral artery. Alternatively, two 12 F arte-
rial perfusion catheters pump blood into the right
and left femoral arteries.
The pump is driven by the external micropro-
cessor-based controller. A constant 10 mL/h flow
of saline with unfractionated heparin is infused
into the pump to provide (1) fluid bearing for
the rotor, (2) local anticoagulation for the pump
chamber to prevent the formation of thrombus,
(3) systemic lubrication, and (4) cooling of the
unit (Fig. 4). A pressure transducer monitors the
infusion pressure and identifies any disruption in
the infusion line. An in-line air bubble detector
monitors for the presence of air in the infusion
line.
Contraindications
Because adequate left atrial pressures are re-
quired for sufficient pumping, the TandemHeart is
contraindicated in patients who have predomi-
nant right ventricular failure [27]. Patients who
have a ventricular septal defect are not good can-
didates for the TandemHeart because of the risk
of hypoxemia due to right-to-left shunting.
Fig. 4. Cutaway section of TandemHeart. The Tandem-
Heart has a dual-chamber pump. The upper housing
provides a conduit for inflow and outflow of blood.
The lower housing provides communication with the
controller, the ability to rotate the impeller, and a fluid
path for a constant 10 mL/h flow of saline with unfrac-
tionated heparin, which is infused into the pump to
provide (1) fluid bearing for the rotor, (2) local anticoa-
gulation for the pump chamber to prevent the formation
of thrombus, (3) systemic lubrication, and (4) cooling of
the unit. (Courtesy of CardiacAssist, Pittsburgh,
Pennsylvania.)
Because the left ventricle can become distended
in the setting of severe left ventricular dysfunction
and impedes subendocardial perfusion, aortic in-
sufficiency is another contraindication to the Tan-
demHeart. The TandemHeart can induce critical
limb ischemia in patients who have severe periph-
eral vascular disease.
Implantation
The TandemHeart is implanted in the cardiac
catheterization laboratory. Under fluoroscopic
guidance, a transseptal puncture with the Brock-
enbrough needle is inserted through a Mullins
sheath into the superior vena cava to obtain access
from the right atrium into the left atrium (Fig. 5).
A two-stage (14/21 F) dilator is used to dilate the
transseptal puncture after a 0.035-inch pigtail
guidewire is inserted into the left atrium. Subse-
quently, the transseptal cannula is inserted into
the right femoral vein, advanced over the guide-
wire toward the right atrium, and placed into
the left atrium. All of the side holes of the trans-
septal cannula are required to be completely
Fig. 5. Access to the left atrium is obtained by way of
standard transseptal techniques. Catheter exchanges
are made with the valvuloplasty wire. The septum is di-
lated with a two-stage dilator. Next, the cannula is
placed into the left atrium while ensuring all drainage
holes are in the left atrium (last 2.5 cm of the cannula).
(Courtesy of CardiacAssist, Pittsburgh, Pennsylvania.)
 PERCUTANEOUS LEFT VENTRICULAR SUPPORT DEVICES
positioned in the left atrium to avoid right-to-left
shunting. After the position of the cannula is con-
firmed by angiography and echocardiography, the
guidewire and obturator are removed. The can-
nula is then sutured to the skin over the thigh
and cross-clamped with a vascular clamp.
Because of the risk of critical limb ischemia in
patients who have severe peripheral vascular
disease, an iliofemoral angiogram before implan-
tation of the arterial perfusion catheter is crucial.
The tip of the arterial outflow cannula is advanced
to the level of the aortic bifurcation over
a guidewire after access in the right femoral artery
is obtained. After the air is purged from the
extracorporeal system, the transseptal cannula is
attached to the inflow port of the centrifugal blood
pump in the standard wet-to-wet fashion with
3/8 inch. Tygon tubing. The power supply
for the TandemHeart is subsequently connected
to the microprocessor-based controller (Fig. 6).
With skilled and experienced operators, the entire
insertion, assembly, and mechanical circulatory
support can be accomplished in 30 minutes or
less. Because of the risk of thromboembolic com-
plications, systemic anticoagulation with unfrac-
tionated heparin to maintain an activated
clotting time of 400 seconds during insertion and
200 seconds during support is mandatory. The
TandemHeart has been used for up to 14 days [5].
Explantation
When the TandemHeart is ready to be discon-
tinued, the percutaneous arterial and venous
cannulas can easily be removed percutaneously
after the heparin drip and pump are turned off.
Fig. 6. TandemHeart microprocessor-based controller has a primary and backup control unit to reduce the need for
back-up equipment. (Courtesy of CardiacAssist, Pittsburgh, Pennsylvania.)
269
When the activated clotting time is low enough to
allow for safe removal of the arterial cannula,
hemostasis is obtained with manual compression
of the puncture site. A small iatrogenic atrial
septal defect is left after the explantation of the
transseptal cannula, which resolves after 4 to 6
weeks or has no clinically significant left-to-right
shunt [28]. A bubble study demonstrated no echo-
cardiographic evidence of a residual atrial septal
defect after removal of the transseptal cannula [5].
Hemodynamic effects
The TandemHeart provides circulatory sup-
port by diverting oxygenated blood from the left
atrium into systemic circulation, which increases
cardiac output and blood pressure and decreases
afterload and preload, thus decreasing myocardial
oxygen demand. The increase in mean arterial
pressure may optimize the supply and demand of
oxygen in the myocardium at risk and increase
tissue perfusion at the coronary and peripheral
level [27]. This increase in tissue perfusion leads to
the reversal of metabolic acidosis and a decrease
in serum lactate levels in patients who have car-
diogenic shock. In addition, Doppler flow studies
of the carotid vasculature demonstrate augmenta-
tion of carotid artery flow during diastole (Fig. 7).
Potential complications
The transseptal puncture required for the
TandemHeart is a potential source of complica-
tions (Box 1). Inadvertent puncture of the aortic
root, coronary sinus, or posterior free wall of the
right atrium can lead to disastrous complications
270 LEE & MAKKAR

Fig. 7. Carotid artery flow at varying TandemHeart rpm. Doppler flow studies of the carotid vasculature demonstrate
augmentation of carotid artery flow during diastole. (Courtesy of CardiacAssist, Pittsburgh, Pennsylvania.)

including death. Thromboembolism can be an-
other potential source of complication. Cerebral
thromboembolism occurred in a patient who
formed thrombus at the edge of a large ventricular
septal defect and at the site of the left atrial punc-
ture despite anticoagulation with unfractionated
heparin [27]. Because unfractionated heparin is
needed to achieve a high activated clotting time,
bleeding, especially from the groin, can occur
with the TandemHeart. Systemic hypothermia
can occur when contact of the system circuit with
room temperature leads to the cooling of the blood
flowing through the pump [30]. Accidental dis-
lodgement of the arterial cannula has led to acute
decompensation and death from cardiogenic
shock [27]. Local infections, bacteremia, and sepsis
are potential complications with any implantable
device.
valvuloplasty, percutaneous aortic valve replace-
ment, and functioning as a percutaneous right
ventricular assist device for right heart failure
(Fig. 8). A high-risk patient underwent PCI for os-
tial left main coronary artery stenosis (Fig. 9), os-
tial right coronary artery stenosis (Fig. 10), and
aortic valvuloplasty for critical aortic stenosis
(aortic valve area of 0.53 cm2) with the support
of the TandemHeart (Fig. 11) [30].
Impella CardioSystems
The Impella Recover LP 2.5 System (Fig. 12) is
one of four percutaneous or minimally invasive
ventricular unloading catheters manufactured by

Potential applications
In addition to providing hemodynamic sup-
port in patients who have cardiogenic shock and
in high-risk PCI patients, potential applications
for the TandemHeart include high-risk aortic

Box 1. Potential complications of the

TandemHeart
Fig. 8. In addition to providing hemodynamic support
Puncture of aortic root, coronary sinus, in patients who have cardiogenic shock and in high-
risk PCI patients, potential applications for the Tandem-
or posterior free wall of right atrium
Heart include functioning as a percutaneous right ven-
Thromboembolism tricular assist device for right heart failure. The 21 F
Systemic hypothermia venous transseptal inflow cannula aspirates oxygenated
Cannula dislodgement blood from the right atrium. Blood is then delivered
Bleeding from the pump to the pulmonary artery with an arterial
Infection perfusion catheter. (Courtesy of CardiacAssist, Pitts-
burgh, Pennsylvania.)
 PERCUTANEOUS LEFT VENTRICULAR SUPPORT DEVICES 271

Fig. 9. The ostial left main coronary artery showing severe stenosis (A). An excellent angiographic result was obtained
after stenting with a 3.0  8 mm Cypher (Cordis, Johnson and Johnson Corp.; Miami, Florida) drug-eluting stent (B).

Impella CardioSystems. These devices have been
used in Europe to provide rapid and short-term
hemodynamic support in patients who have acute
heart failure, acute myocardial infarction, PCI,
and during and after cardiac surgery (especially
for postcardiotomy low cardiac output syndrome)
by aspirating blood from the left ventricle and ex-
pelling it to the ascending aorta. The Impella Re-
cover LP 2.5 System implements a miniaturized
rotary blood pump that provides circulatory assist
in acute myocardial infarction, cardiogenic shock
or low-output states, or during high-risk PCI for
up to 5 days.
The intracardiac axial flow pump contains
a rotor that is driven by an electrical motor and
has an inflow cannula. The left ventricular pump
can provide up to 2.5 L/min of cardiac output,
depending on the speed of the rotor (which can
operate at a maximum of 50,000 rpm), and the
difference between aortic blood pressure and left
ventricular pressure. Located in the front of the
rotor is a differential pressure sensor that moni-
tors this difference in pressure (Fig. 13). The ap-
propriate position of the flow pump can be
confirmed by the pressure difference between the
aorta and left ventricle. The system has a catheter
coming from the patient that connects to a mobile
console that permits the management of the rota-
tional speed of the pump and displays of the dif-
ferences of pressure between the inflow and
outflow (Fig. 14). The Impella pump is continu-
ously purged with a glucose solution (10%) that
is drawn into a 50-mL syringe with heparin
(2500 IU) (Fig. 15). The purge flow rates normally
range from 2 to 6 mL/h to continuously rinse and
prevent thrombus formation in the pump.

Fig. 10. The ostial right coronary artery showing severe stenosis (A). An excellent angiographic result was obtained after
stenting with a 3.0  13 mm Cypher (Cordis, Johnson and Johnson Corp.; Miami, Florida) drug-eluting stent (B).
272 LEE & MAKKAR
Fig. 11. The aortic valve gradient was 75 mm Hg and the
aortic valve area was 0.53 cm2. After crossing the aortic
valve with a 6 F multipurpose catheter and standard
glidewire (Boston Scientific, Natick, Massachusetts) and
exchanging the glidewire with a stiff Amplatz (Cook,
Bloomington, Indiana) 260-cm wire, a 5.0  20 mm
Zmed-II valvuloplasty balloon (Numed, Hopkinton,
New York) was inflated twice. The final gradient across
the aortic valve was 16 to 20 mm Hg and the aortic valve
area was 0.8 cm2.
Contraindications
In addition to severe peripheral vascular dis-
ease, the Impella Recover LP 2.5 System is
contraindicated in patients who have mechanical
aortic valves or heavily calcified aortic valves.
Implantation
The Impella Recover LP 2.5 System, which is
4 mm in diameter (12 F), is mounted on a 9 F
pigtail catheter and inserted percutaneously in the
Fig. 12. The Impella Recover LP 2.5 System is an
LVAD that provides hemodynamic support in patients
that can be inserted percutaneously. The pigtail tip is in-
serted into the left ventricle to withdraw blood into the
aorta. (Courtesy of Impella CardioSystems, Aachen,
Germany.)
Fig. 13. The intracardiac axial flow pump contains a ro-
tor that is driven by an electrical motor and has an in-
flow cannula. Blood is drawn in at the tip of the
cannula and is expelled at the sides upstream from the
motor into the aorta. A differential pressure sensor mea-
sures the difference in pressure between the inflow and
outflow of the pump. (Courtesy of Impella CardioSys-
tems, Aachen, Germany.)
cardiac catheterization laboratory into the femo-
ral artery over a guidewire through a peel-away
introducer with a removable hemostatic valve
(Fig. 16). An advantage of the Impella over the
TandemHeart is that there is no need for a trans-
septal puncture and no extracorporeal blood. The
inflow cannula miniaturized pump is inserted
through the aortic valve under fluoroscopic guid-
ance into the left ventricle to pump blood from the
left ventricle into the ascending aorta (Fig. 17)
[31]. If the patient also has an IABP on the
Fig. 14. The mobile console permits the management of
the rotational speed of the pump and displays of the
differences of pressure between the inflow and outflow.
It provides 60 minutes of battery time and nine flow
settings. (Courtesy of Impella CardioSystems, Aachen,
Germany.)
 PERCUTANEOUS LEFT VENTRICULAR SUPPORT DEVICES
Fig. 15. The Impella motor is continuously purged with
a glucose solution (10%) that is drawn into a 50-mL
syringe with heparin (2500 IU) The purge flow rates
normally range from 2 to 6 mL/h to continuously rinse
and prevent thrombus formation in the pump. (Courtesy
of Impella CardioSystems, Aachen, Germany.)
contralateral groin, it can be put on standby as the
Impella pump is being positioned. Continuous
anticoagulation with heparin to maintain the acti-
vated clotting time above 160 seconds is required
to minimize thromboembolism while the pump is
in use.
Hemodynamics effects
The Impella pump rapidly unloads the left
ventricle by delivering blood into the ascending
Fig. 16. The Impella pump, which is 4 mm in diameter
(12 F), is mounted on a 9 F pigtail catheter and inserted
percutaneously in the cardiac catheterization laboratory
into the femoral artery over a guidewire through a peel-
away introducer with a removable hemostatic valve.
(Courtesy of Impella CardioSystems, Aachen, Germany.)
273
Fig. 17. The Impella is placed across the aortic valve al-
lowing the pump to aspirate blood from the left ventricle
and expel it to the ascending aorta. The Impella imple-
ments a miniaturized rotary blood pump that provides
circulatory assist in acute myocardial infarction, cardio-
genic shock, low-output states, or during high-risk PCI
for up to 5 days. (Courtesy of Impella CardioSystems,
Aachen, Germany.)
aorta. In an animal model, the Impella pump
reduced myocardial oxygen consumption during
ischemia and reperfusion, leading to a reduced
infarct size [32]. Patients who have cardiogenic
shock treated with the Impella pump system had
improvements in cardiac output and mean blood
pressure and a reduction in pulmonary capillary
wedge pressure [31]. Compared with the IABP,
the Impella pump improved cardiac and systemic
hemodynamics during acute mitral regurgitation
in an animal model [33]. No significant increase
in aortic regurgitation was observed with intracar-
diac echocardiography when the catheter was
through the aortic valve [34].
The average pressure-volume loops demon-
strate a reduction in the left ventricular end-
diastolic pressure (18–11 mm Hg), end-diastolic
volume (345–321 mL), and stroke volume (94–76
mL) [34].
Potential complications
Meyns and colleagues [31] reported the loss of
the sensor signal in three patients, which did not af-
fect pump function. Similar to the TandemHeart,
an iliofemoral angiogram before sheath insertion
is indicated because of the risk of critical limb
ischemia in patients who have severe peripheral
274 LEE & MAKKAR
vascular disease. No patients have required treat-
ment for hemolysis when treated with the Impella
Recover LP 2.5 System. Displacement of the
pump back into the aorta can also occur, but the
addition of the pigtail catheter tip minimizes dis-
placement potential. Similar to the TandemHeart,
the percutaneous blood cannulae pose the inherent
risk of infectious complications in the form of local
infections, bacteremia, and sepsis.
Summary
Patients undergoing PCI who have severely
compromised left ventricular systolic function and
complex coronary lesions including multivessel
disease, left main disease, or bypass graft disease
are at higher risk of adverse outcomes from
hemodynamic collapse. The TandemHeart percu-
taneous LVAD and the Impella Recover LP 2.5
System can provide short-term hemodynamic
stability in patients who require hemodynamic
support in a variety of settings. Both devices
provide more ventricular support than the IABP
and can be implanted rapidly and percutaneously
in a prophylactic or emergency setting. Identifica-
tion of those who are at high risk for severe
hemodynamic compromise and most likely to
benefit from mechanical circulatory support is
crucial to derive the greatest benefit from this
modality. Multicenter randomized clinical trials
are needed to clearly define the role of these two
devices in providing circulatory support in a vari-
ety of clinical settings.
References
[1] Black A, Cortina R, Bossi I, et al. Unprotected left
main coronary artery stenting: correlates of midterm
survival and impact of patient selection. J Am Coll
Cardiol 2001;37:832–8.
[2] Ellis SG, Tamai H, Nobuyoshi M, et al. Contempo-
rary percutaneous treatment of unprotected left
main coronary stenoses: initial results from a multi-
center registry analysis 1994–1996. Circulation 1997;
96:3867–72.
[3] Kosuga K, Tamai H, Ueda K, et al. Initial and long-
term results of angioplasty in unprotected left main
coronary artery. Am J Cardiol 1999;83:32–7.
[4] Kar B, Butkevich A, Civitello AB, et al. Hemody-
namic support with a percutaneous left ventricular
assist device. Tex Heart Inst J 2004;31:84–6.
[5] Vranckx P, Foley DP, de Feijter PJ, et al. Effective
use of the TandemHeart during high-risk percutane-
ous coronary intervention. Int J Cardiovasc Inter-
vent 2003;5:35–9.
[6] Aragon J, Lee MS, Kar S, et al. Percutaneous left
ventricular assist device: ‘‘TandemHeart’’ for high-
risk coronary intervention. Catheter Cardiovasc
Interv 2005;65:346–52.
[7] Becker RC, Gore JM, Lambrew C, et al. A compos-
ite view of cardiac rupture in the United States
National Registry of Myocardial Infarction. J Am
Coll Cardiol 1996;27:1321–6.
[8] Scheidt S, Wilner G, Mueller H, et al. Intra-aortic
balloon counterpulsation in cardiogenic shock: re-
port of a co-operative clinical trial. N Engl J Med
1973;288:979–84.
[9] DeWood MA, Notske RN, Hensley GR, et al. Intra-
aortic balloon counterpulsation with and without
reperfusion for myocardial infarction shock. Circu-
lation 1980;61:1105–12.
[10] Catinella FP, Cuningham JN, Glassman E, et al.
Left atrium-to-femoral artery bypass: effectiveness
in reduction of acute experimental myocardial in-
farction. J Thorac Cardiovasc Surg 1983;86:887–96.
[11] Grossi EA, Krieger KH, Cunningham JN, et al.
Time course of effective interventional left heart as-
sist for limitation of evolving myocardial infarction.
J Thorac Cardiovasc Surg 1986;91:624–9.
[12] Laschinger JC, Grossi EA, Cunningham JN Jr, et al.
Adjunctive left ventricular unloading during myo-
cardial reperfusion plays a major role in minimizing
myocardial infarct size. J Thorac Cardiovasc Surg
1985;90:80–5.
[13] Frazier OH, Rose EA, Macmanus Q, et al. Multicen-
ter clinical evaluation of the HeartMate 1000 IP left
ventricular assist device. Ann Thorac Surg 1992;53:
1080–90.
[14] Chen JM, DeRose JJ, Slater JP. Improved survival
rates support LVAD implantation early after myo-
cardial infarction. J Am Coll Cardiol 1999;33:
1903–8.
[15] Vogel RA, Shawl F, Tommaso C, et al. Initial report
of the national registry of elective cardiopulmonary
bypass supported coronary angioplasty. J Am Coll
Cardiol 1990;15:23–9.
[16] Shawl F, Domanski MJ, Wish MH, et al. Percutane-
ous cardiopulmonary bypass support in the cathe-
terization laboratory: technique and complications.
Am Heart J 1990;120:195–203.
[17] Kaul U, Sahay S, Bahl VK, et al. Coronary angio-
plasty in high-risk patients: comparison of elective
intraaortic balloon pump and percutaneous cardio-
pulmonary bypass support-a randomized study.
J Interv Cardiol 1995;8:199–205.
[18] Scholz KH, Dubois-Rande JL, Urban P, et al. Clin-
ical experience with the percutaneous hemopump
during high-risk coronary angioplasty. Am J Cardi-
ol 1998;82:1107–10.
[19] Kantrowitz A, Tjonneland S, Freed PS, et al. Initial
clinical experience with intraaortic balloon pumping
in cardiogenic shock. JAMA 1968;203:113–8.
[20] Nanas JN, Moulopoulos SD. Counterpulsation:
historical background, technical improvements, he-
 PERCUTANEOUS LEFT VENTRICULAR SUPPORT DEVICES
modynamic and metabolic effects. Cardiology 1994;
84:156–67.
[21] Hata M, Shiono M, Orime Y, et al. Coronary micro-
circulation during left heart bypass with a centrifugal
pump. Artif Organs 1996;20:678–80.
[22] Stone GW, Marsalese D, Brodie BR, et al. A pro-
spective, randomized evaluation of prophylactic
intraaortic balloon counterpulsation in high risk pa-
tients with acute myocardial infarction treated with
primary angioplasty. Second Primary Angioplasty
in Myocardial Infarction (PAMI-II) Trial Investiga-
tors. J Am Coll Cardiol 1997;29:1459–67.
[23] Berger PB, Holmes DR Jr, Stebbins AL, et al. Im-
pact of an aggressive invasive catheterization and
revascularization strategy on mortality in patients
with cardiogenic shock in the Global Utilization of
Streptokinase and Tissue Plasminogen Activator
for Occluded Coronary Arteries (GUSTO-1) trial.
Circulation 1997;96:122–7.
[24] Pae W, Pierce W. Temporary left ventricular assis-
tance in acute myocardial infarction and cardiogenic
shock. Chest 1981;79:692–5.
[25] Mahaffey KW, Kruse KR, Ohman EM. Perspectives
on the use of intra-aortic balloon counterpulsation
in the 1990s. In: Topol EJ, editor. Textbook of inter-
ventional cardiology. (Update 21). Philadelphia:
WB Saunders; 1996. p. 303–21.
[26] Mulukutla SR, Pacella JJ, Cohen HA. Contempo-
rary cardiology: cardiogenic shock: diagnosis and
treatment. Totowa (NJ): Humana Press; 2002. p.
303–24.
[27] Thiele H, Lauer B, Hambrecht R, et al. Reversal of
cardiogenic shock by percutaneous left atrial-to-
275
femoral atrial bypass assistance. Circulation 2001;
104:2917–22.
[28] Fonger JD, Zhou Y, Matsuura H, et al. Enhanced
preservation of acutely ischemic myocardium with
transseptal left ventricular assist. Ann Thorac Surg
1994;57:570–5.
[29] Lemos PA, Cummins P, Lee CH, et al. Usefulness of
percutaneous left ventricular assistance to support
high-risk percutaneous coronary interventions. Am
J Cardiol 2003;91:479–81.
[30] Makkar RR, Aragon J, Soleimani T, et al. Ostial
left main stenosis, ostial right coronary stenosis
and aortic valvuloplasty performed with the support
of the TandemHeart percutaneous left ventricular
assist device. Available at: http://www.tctmd.com.
Accessed February 14, 2005.
[31] Meyns B, Dens J, Sergeant P, et al. Initial experi-
ences with the Impella-device in patients with cardio-
genic shock. Thorac Cardio Surg 2003;51:1–6.
[32] Meyns B, Stolinski J, Leunens V, et al. Left ventric-
ular support by catheter-mounted axial flow pump
reduces infarct size. J Am Coll Cardiol 2003;41:
1087–95.
[33] Reesink K, Dekker A, van der Nagel T, et al. New
Impella intracardiac minipump supports the acutely
failing left heart significantly more effective than in-
tra aortic balloon pumping [abstract]. J Am Coll
Cardiol 2003;41:215A.
[34] Valgimigli M, Steendijk P, Sianos G, et al. Left ventric-
ular unloading and concomitant total cardiac output
increase by the use of percutaneous Impella Recover
LP 2.5 assist device during high-risk coronary inter-
vention. Cathet Cardiovasc Interv 2005;65:263–7.
 Cardiol Clin 24 (2006) 277–286
Vascular Closure Devices
Michael C. Kim, MD
Cardiac Catheterization Laboratory, The Mount Sinai School of Medicine, 5 East 98th Street,
3rd Floor, New York, NY 10029, USA
As the number of femoral artery catheteriza-
tions continues to increase over the next decade to
more than 9 million, the importance of quick,
efficient, and effective vascular closure techniques
cannot be overemphasized. Vascular access com-
plications remain the leading source of morbidity,
cost, and legal ramifications after a cardiac cath-
eterization, with or without percutaneous coro-
nary intervention (PCI) [1]. Any experienced
interventional cardiologist knows the importance
of vascular access and hemostasis; many patients
remember only the pain of entering and closing
the artery. Patients who return for a repeat cathe-
terization often do not worry about the actual
procedure, but are concerned about the length
of time that they are required to remain in bed
afterward.
Vascular complicationsdreported to be as
high as 14%dwith PCI lead to increased hospital
length of stay and hospital costs [2]. Vascular clo-
sure devices (VCDs) were introduced in 1995 to
decrease complications and the time to hemostasis
and ambulation. Although the frequency of com-
plications with VCDs is debatable, the VCD mar-
ket has become a greater than $500 million
industry because of advances in patient comfort
and physician efficiency [3].
In 1953, Dr. Sven-Ivar Seldinger [4], a radiolo-
gist from Sweden, published his landmark paper
that detailed the Seldinger percutaneous technique
for obtaining femoral artery access. In that paper,
he described the technique of manual compression
(MC), which has been the gold standard for vas-
cular access hemostasis for more than 50 years.
The access site should be held with direct pressure
for 20 to 30 minutes followed by overnight bed
E-mail address: michael.kim@msnyuhealth.org
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.03.006
rest. Only over the last decade has this technique
been challenged with the introduction of VCDs
as the numbers of PCIs increased and the antipla-
telet and anticoagulation regimens became more
complicated.
This article summarizes the VCD technologies
that are available to physicians who perform
percutaneous catheter-based procedures.
Patches
Vascular closure patches still require MC. It is
a possible means of decreasing time to hemostasis
and time to ambulation. This is helpful to improve
throughput in the laboratory at a cost that is less
than a VCD; however, no randomized clinical
trial has comparing the efficacy of these patches
with MC or VCDs. The major advantage of
patches over VCDs is a lower purchasing cost
and less hardware in the patient.
Syvek Patch
The Syvek Patch (Marine Polymer Technolo-
gies, Inc., Danvers, Massachusetts) is based on the
ability of poly-N-acetylglucosamine (pGIcNAc)
to promote vasoconstriction, red blood cell ag-
glutination, and platelet activation for fibrin clot
formation. Studies have shown that pGIcNAc
helps to promote vasoconstriction by way of
endothelin-1 activation. Animal models also dem-
onstrated red blood cell agglutination when
a patch laced with pGIcNAc was placed on an
open wound. Its mechanism of action is unknown.
Furthermore, studies demonstrated the activation
of platelets when in contact with pGIcNAc, which
leads to rapid fibrinogen cross-linking and platelet
aggregation.
Several small trials showed effective hemostasis
using the Syvek Patch when applied for 10 minutes
cardiology.theclinics.com
278
or less. One study showed safety in discharging
patients after 2 hours when using the patch [5]. The
largest study, by Nader and colleagues [6], looked
at 1000 consecutive patients who used the Syvek
Patch. The 636 diagnostic procedures required 10
minutes of compression and the 364 PCI proce-
dures required 20 minutes of patch compression.
This observational study showed a major complica-
tion rate of 0.1% (pseudoaneurysm in the PCI
group), and a minor complication of 1.3% (mostly
small hematomas and nuisance bleeds).
Scion Clo-Sur P.A.D.
The Clo-Sur P.A.D. (Scion Cardiovascular,
Miami, Florida) is a soft, nonwoven hydrophilic
wound dressing that contains a naturally occur-
ring biopolymer (polyprolate acetate). This bio-
polymer is cationically charged, and it is theorized
that the positive ionic charges interact with the
negatively charged red blood cells and platelets,
and result in the acceleration of a naturally
forming fibrin and platelet clot. It has been
approved by the Food and Drug Administration
(FDA) for management in bleeding wounds. A
single center looked at 122 patients (40 PCIs) who
used the Clo-Sur P.A.D. for at least 10 minutes;
most ambulated within 2 hours [7]. There were no
major complications in the PCI group and 2
patients in the diagnostic catheterization group
developed a pseudoaneurysm that was diagnosed
4 days after the procedure.
Chito-Seal pad
The Chito-Seal pad (Abbott Vascular Devices,
Redwood City, California) is similar to the Clo-
Sur P.A.D. This pad is approved by the FDA
for the management of bleeding wounds, includ-
ing vascular access sites. It is coated with Abbott’s
proprietary chitosan gel, a powerful hemostatic
agent. Chitosan gel is twice as chemically active
as chitin, the biopolymer from which it is derived.
When placed over a bleeding wound, it becomes
a cell-binding agent that consists of positively
charged chitosan molecules that attract negatively
charged red blood cells and platelets; it is designed
to accelerate natural clot formation.
Only one study has compared different
patches. Fisher and colleagues [8] used a coagulo-
pathic swine spleen-bleeding model to compare
the Syvek Patch with the Clo-Sur P.A.D. and
the Chito-Seal pad. The Syvek Patch produced
in vitro red blood cell aggregation and reduced
the time to in vitro fibrin clot formation using
KIM
platelet-rich plasma samples in 100% of cases
compared with 50% of cases when using a dry
gauze. The Clo-Sur P.A.D. achieved hemostasis
in no cases, whereas the Chito-Seal pad achieved
hemostasis in only 25% of cases.
D-Stat products
D-Stat products (Vascular Solutions, Minne-
apolis, Minnesota) are a family of topical hemo-
stats that are coated with thrombin, and thereby,
stimulate the production of fibrinogen, which
activates fibrin hemostasis. Products include
a foam pad, gel, bandage, radial pressure device,
and a clamp accessory device. No randomized
clinical trials have been published to confirm their
efficacy over standard compression devices or MC.
Suture-mediated closure devices
Suture-mediated closure devices have distinct
advantages and disadvantages. Devices that can
deliver a U-stitch at the arteriotomy site have
a long history of effective hemostasis; vascular
surgeons have been using the technique for de-
cades in the operating room. The original Sones’
technique of brachial artery access for diagnostic
cardiac catheterization used a cut-down to access
the artery and manual sutures to close the brachial
artery access site. Additionally, immediate reac-
cess at the femoral artery site can be performed
using this technique, because there are no con-
cerns of placing a new sheath over the old site or
suture. Compare that with devices that use
collagen/thrombin plugs at the site where a new
sheath could embolize the plug into the femoral
artery and produce thrombosis of the leg or
disrupt the plug in the tissue tract, which leads
to bleeding at the previous arteriotomy site.
Vascular surgeons have reported a diminished
inflammatory response to suture-mediated closure
techniques at the tissue tract site [9]; the clinical
relevance of this is unknown.
Perclose
The Perclose Prostar (Abbott Vascular De-
vices) device was introduced in 1994 as the first
suture-mediated device to be approved by the
FDA. Using a braided polyester suture, the
Prostar can close 6.5F to 8F access sites. It is
designed to provide complete tissue apposition
that results in primary healing. Initial studies
 VASCULAR CLOSURE DEVICES
showed a reduced time to hemostasis, ambulation,
and discharge in diagnostic femoral artery cathe-
terization compared with MC.
Each case requires a femoral angiogram to
ensure proper sheath placement in the common
femoral artery and to rule out significant periph-
eral vascular disease (PVD). The Perclose device
requires five successful steps for hemostasis. First,
one must position the device into the femoral
artery. When the device is placed ideally in the
femoral artery, blood will be seen through the
marker lumen. The foot pedals are deployed
inside the arteries and the device is pulled until
pressure is felt. Second, the needles are deployed.
Third, the plunger must be removed, which allows
the suture to be retrieved. Next, the suture is
harvested and must be knotted. The fifth and final
step is knot advancement of the suture until it is
placed at the femoral arteriotomy site. One tests
for hemostasis by asking the patient to cough, and
the suture is cut above the artery within the tissue
tract.
The Prostar was never able to capture a large
market share because of the complexity of deploy-
ing the device. The needle and suture was placed
into the artery by the device but it had to be
retrieved within the tissue tract manually and then
knotted using a separate manual device that was
included in the Prostar package. The end result
was a complicated and time-consuming device
that led to many points at which error could be
introduced. The Prostar XL is still available if one
needs to close a 8.5F to 10F access site.
In 1997, the Perclose Closer was introduced as
a more user-friendly suture-mediated closure de-
vice. Additionally, studies showed that it was an
effective alternative to MC in anticoagulated
patients who were undergoing PCI. This device
was a smaller 6F device where the needle de-
ployment occurs from outside the femoral artery;
this allows the device to capture the suture
automatically by the device. The operator then
knots the suture using the manual knot-tying
device that is provided. Although this advance
allowed for a quicker deployment of the device, it
was still cumbersome because of the time needed
to produce a knot and then deploy it to the
arteriotomy site.
In 2002, The Perclose A-T (Auto-Tie) was
introduced. This marked a major step in suture-
mediated closure because time to closure was
reduced significantly. The Perclose A-T produces
a pretied knot that obviates the need for the
manual knot-tying device. This step was
279
considered the most technically demanding and
the most time consuming. The auto-tie feature
decreased the procedure time and led to a decrease
in unsuccessful deployments. Additionally, the
auto-tie function allows for true single-operator
deployment.
In 2004, the most recent Perclose model, called
the Proglide, was introduced. The Proglide uses
a polypropylene monofilament suture in place of
the polyester braided suture that was used in
previous models. This monofilament suture re-
portedly increases knotted tensile strength, pro-
vides easier knot delivery because the material is
more slippery, and is designed to provide less
inflammatory response compared with a braided
suture. The Proglide maintains the automated
knot tying with a pretied, heat-set knot and adds
a quick-cut mechanism that eliminates the need
for extra sharps or a sterile scissor (Fig. 1).
The Perclose system represents approximately
35% to 40% of the VCD market. It is quick,
reliable, and effective in producing hemostasis in
diagnostic interventional cases that use the femo-
ral artery. It can be deployed by a single operator
and allows for immediate reaccess. Because the
catheter remains in the femoral artery until it is re-
moved, it allows for reintroduction of a wire and
sheath into the artery if the device fails to deploy
the needles or suture appropriately. Because the
risk for a vascular complication is greatest when
a VCD fails, this advantage cannot be overempha-
sized. Complications include infection, laceration
of the artery by the foot pedals, and posterior
wall partial dissection by the foot pedal that
may lead to abrupt or subacute femoral artery clo-
sure or thrombosis that is due to inadvertent knot-
ting of the posterior dissection plane to the
anterior femoral arteriotomy site by way of the su-
tures. Early trials with the Prostar version showed
a 90% successful deployment rate with a minor
vascular complication rate of 5% and a major
complication rate of 3% to 4% [10].
X-Site suture closure device
The X-Site (Datascope Corp., Montvale, New
Jersey) suture-mediated closure device is an alter-
native to Perclose. Using a braided polyester
suture, the X-Site device is able to deploy the
needles and suture through the femoral artery
without the use of foot pedals, which decreases the
possibility of any intra-arterial trauma. Because
there is no auto-tie mechanism in the device,
however, it will not be available until 2007.
280
Fig. 1. Perclose Proglide deployment. Five-step deployment process for suture-mediated closure. (Courtesy of Abbott
Vascular, Redwood City, CA.)
The randomized trial that led to FDA ap-
proval of the device showed a striking decrease in
vascular complications compared with MC. The
Rapid Ambulation to Closure trial randomized
250 patients who underwent diagnostic or PCI
procedures to the X-Site device or MC. There
were no major vascular complications with the
KIM
X-Site device, whereas the frequency of major
vascular complications with MC use in diagnostic
procedures and PCI were 1.1% and 3.4%, re-
spectively [11]. This represents the only prelimi-
nary randomized trial of any device that showed
a decrease in vascular complications when com-
pared with MC.
 VASCULAR CLOSURE DEVICES
Sealant closure devices
Sealant devices have the longest history of
effective hemostasis. The concept is to enhance
the body’s natural method of achieving hemosta-
sis by delivering a procoagulant, such as collagen
or thrombin, extravascularly to the surface of the
femoral artery. Collagen and thrombin attract
and activate platelets and form a quick coagulum
at the surface of the artery and in the tissue tract
that results in a seal. Unlike the suture-mediated
closure devices, the sealant devices do not allow
for immediate reaccess at the same arterial punc-
ture site because of potential disruption of the
sealant itself or potential embolization of the
sealant into the femoral artery.
Angio-Seal closure device
Because of its incredibly simple design and
deployment, the Angio-Seal (St. Jude Medical,
Minneapolis, Minneapolis) VCD has become the
most popular VCD in the United States, and
represents more than 60% of the VCD market
share. It works by a combination of mechanical
forces and placement of a collagen plug that stim-
ulates thrombus formation and platelet aggrega-
tion. Mechanically, it uses a ‘‘sandwich
technique’’ at the arteriotomy site with an anchor
on the intravascular side and a collagen plug on
the extravascular side.
All cases require a predevice femoral angio-
gram to rule out PVD and to confirm femoral
artery sheath placement. The original Angio-Seal
device was cumbersome. It consisted of four
components within a single carrier in an 8F
package: the anchor, the connecting suture, the
collagen plug, and a postplacement spring. De-
ploying the device is simple. First the artery is
located when blood rushes out of the sheath and
signifies that the sheath distal end is in the femoral
arterial lumen. The dilator is removed and the
carrier device is introduced into the sheath; this
sets the anchor and introduces a 15-mg collagen
sponge. A tamper is pushed downward to com-
press the collagen against the outer femoral
arterial wall, the postplacement spring is attached
between the tamper, and a metal tag is fixed to the
positioning suture while applying continuous
pressure on the tamper. All components are
bioabsorbable within 90 days.
Because there was difficulty placing the post-
collagen spring and there were reports of mala-
lignment of the anchor that led to cessation of
281
blood flow, embolization of the anchor itself, and
inadvertent entry of the collagen sponge into the
artery, the earlier Angio-Seal devices never proved
to be popular. A series of iterations occurred over
the last 5 years, including the introduction of the
Millenium Platform that incorporated a monofold
sheath design. This allowed for the use of a 6F
sheath in place of the original 8F design.
The next iteration was revolutionary. The STS
(self-tightening suture) Platform eliminated the
need for a postplacement spring, which makes
deployment incredibly simple and ensures an
almost 100% success rate. Furthermore, it pro-
duced a more secure anchor–collagen sandwich.
A single operator can deploy the device in less
than 1 minute when proficient. The newest gener-
ation, the STS Plus, uses a smoother transition
from sheath tip to dilator and has repositioned the
blood inlet holes so that locating the artery is
ensured once initial blood flow is seen. There is no
need to enter, exit, and reenter the artery as was
necessary with the original STS Platform.
The current Angio-Seal system is simple and
effective. Successful deployment is seen in almost
100% of cases. Time to hemostasis is in the range
of 2 to 4 minutes. The FDA allows for patient
ambulation 20 minutes after diagnostic proce-
dures. The author’s laboratory has discharged
patients home safely 1 hour after Angio-Seal
closure. Immediate reaccess 1 cm above the
original puncture site has been approved. Early
reports showed a minor complication rate of 5.9%
and major complication rates of 0.4% to 1.3% in
PCI patients [12]. The author’s laboratory has de-
ployed more than 3000 Angio-Seal devices in pa-
tients who required PCI with a major
complication rate of less than 0.4% (mostly retro-
peritoneal bleeds). The 6F device can close up to
8F sheaths and an 8F device is able to close up
to 10F sheaths (Fig. 2).
Vasoseal closure device
Vasoseal (Datascope Corp.) represents the
original sealant device. Originally introduced in
1995, this device consists of an entirely extravas-
cular bovine collagen plug that induces the
formation of a hemostatic cap above the puncture
site in the tissue tract. The collagen is biodegrad-
able and induces platelet aggregation, which re-
leases coagulation factors and stimulates fibrin
production and thrombus formation. The colla-
gen reabsorbs by way of macrophages and
282 KIM

Fig. 2. Angio-Seal deployment steps. Angio-Seal is considered to be the most user-friendly device to deploy. (Courtesy
of St. Jude Medical, Minneapolis, MN.)
 VASCULAR CLOSURE DEVICES
granulocytes over a 6-week period. Because the
device does not leave a foreign body inside the
artery, it can be used in patients who have
peripheral vascular disease. Additionally, no fem-
oral angiogram is needed, which is a unique
feature among VCDs.
The original Vasoseal VHD consisted of four
parts: an 11F dilator, 11.5F sheaths of seven
different lengths that were selected after a prepro-
cedure needle depth technique was performed,
and two 90-mg collagen cartridges. Placing this
device required two people. First a guidewire was
inserted into the femoral sheath and the sheath
was removed while maintaining MC. A blunt-
tipped 11F dilator was inserted over the wire to
the site of the arterial puncture. The predeter-
mined 11.5F sheath was advanced over the dilator
down to the arterial surface. The dilator and wire
were removed while the second operator held
manual pressure from above. Finally, the collagen
cartridge was deployed with a ‘‘push and pull’’
method through the sheath, and the sheath was
removed. Light MC is then applied for a few
minutes.
Recent iterations include the Vasoseal ES,
which eliminated the need to premeasure arterial
depth and the need for multiple 11.5F sheaths. It
is approved for closure of 5F to 8F arterial
sheaths. The Vasoseal Low Profile system was
introduced more recently. This was a new, smaller
device that was designed to be compatible with
smaller sheaths (including 4F systems). The col-
lagen plug was reduced by 40% and the delivery
system was reduced by 3F sizes. Most recently, the
Vasoseal Elite was launched with the promise of
more rapid and effective hemostasis because of
new collagen plug technology that allows for the
immediate radial expansion of the plug that
secures the collagen within the tissue tract within
seconds of deployment in the tract. Vasoseal Elite
reportedly no longer requires light MC after
deployment and is effective in anticoagulated
patients after PCI.
Clinical studies have demonstrated consistently
decreased time to hemostasis compared with MC,
usually within 5 to 8 minutes of deployment, in
patients who undergo diagnostic procedures or
PCI. The time to ambulation also is decreased
significantly compared with MC; patients are able
to ambulate safely at 1 hour after diagnostic
procedures. Complications, such as infection,
can occur from leaving a foreign body in the
tissue tract. Minor local complications have been
reported in approximately 8% of patients,
283
whereas major complications average approxi-
mately 5% in patients who undergo PCI [13]. Un-
fortunately, a large tissue tract is required to
deploy the device, and if it fails to deploy the op-
erator is left with a large tissue tract in an antico-
agulated patient. This can lead to a large increase
in vascular complications if the device fails to
achieve hemostasis immediately. Because of the
uncertainty in efficacy, especially when compared
with Perclose or Angio-Seal, the Vasoseal device
has not been able to capture a large market share,
despite its unique feature of not leaving intra-arte-
rial material at the end of the procedure.
Duett closure device
Duett (Vascular Solutions) is a newer sealant
device that uses a combination of collagen and
thrombin. It leaves no intra-arterial material and
works solely by generation of a thrombus at the
arterial puncture site. It has the advantage of
leaving perhaps the least amount of inflammation
after deployment but has the disadvantage of the
possibility of a catastrophic complicationdinject-
ing thrombin/collagen mixture into the femoral
artery or having the mixture leak into the femoral
artery. It is easy to recognize this complication
because it leads to an acutely ischemic leg almost
immediately.
A femoral arteriogram should be obtained to
rule out PVD and to ensure placement of the
sheath in the common femoral artery. The Duett
system uses a low-profile balloon-positioning
catheter that goes over a core wire after the sheath
is removed. When the device is inflated, the
balloon becomes elliptical and provides tempo-
rary sealing of the arterial puncture site from
inside. The balloon does allow for antegrade flow
of blood in the artery during inflation so ischemia
does not occur during balloon inflation. A mix-
ture of 250 mg of bovine microfibrillar collagen
and 10,000 units of bovine thrombin is made by
the operator, and is injected slowly above the
arterial puncture site through the side arm of the
catheter. Almost immediate hemostasis is
achieved and the patient usually feels a strong
burning sensation as the procoagulant mixture
is injected. The balloon is deflated and the
catheter is removed from the patient. Because of
the potency of the mixture, oozing is a rare
complication after Duett closure. Immediate re-
access is acceptable if the new puncture site is
above the just closed arterial site.
284 KIM

Initial trials showed a success rate of 98% to The wire is removed and the ‘‘three-step’’ in-
100%. Time to hemostasis and time to ambula- troducer stabilizes the anterior vessel wall. The
tion were decreased significantly compared with dilator is removed and the staple device is
MC. Trials in patients who underwent diagnostic advanced through the introducer until the staple
procedures or PCI showed a minor vascular reaches the level of the anterior wall. As the
complication rate of about 2% with a major trigger is activated, the staple is deployed into the
complication rate of 1% [14]. Unfortunately, media and adventitia of the femoral wall,
there is a 0.4% chance of inadvertent intra-arterial the stabilization filaments are retracted, and the
injection of the procoagulant mixture into the introducer is removed all at once (Fig. 3).
femoral artery that requires emergent therapy The Angiolink trial that led to FDA approval
(Angiojet thrombectomy, intra-arterial infusion showed an in-hospital major complication rate of
of thrombolytics, surgical thrombectomy). Be- 0.4% versus 1.7% for MC. It is one of the few
cause of the fear of this catastrophic complication, studies to show a decrease in complication rates
the Duett market share has been low. compared with MC [15]. Of note, 80% of these
A similar balloon occlusion system, called the patients received heparin and glycoprotein IIB/
Matrix, is to be released soon; it uses polyethylene IIIA inhibition.
glycol as the procoagulant. This has the advan- The Starclose (Abbott Vascular Devices) just
tage of being water soluble, and thus, dispersible, received FDA approval and has been launched
if injected into the arterial lumen. Trials are nationwide to much enthusiasm. It rivals the
underway to investigate its clinical efficacy. Angio-Seal device in simplicity and ease of use;
it can be used in patients who have PVD or
delivered in low sticks below the femoral artery
bifurcation because of its extraluminal design. It
Staple-mediated closure devices
delivers a nitinol clip between the media and
In the operating room, surgeons are now using adventitia, and thereby, leaves no material inside
staple technology at an increasing rate as a means the arterial lumen. Oozing can be an issue because
for hemostasis over sutures. Staples can create an
anatomic ‘‘purse-string suture’’ closure by using
pledgets to gather vessel adventitia and media at
the arteriotomy edges; this allows for tissue
approximation and hemostasis. This technique is
so effective that decannulation of large-bore
cannulas from the ascending aorta is followed
by a ‘‘purse-string suture’’ closure technique by
way of large staples. This extraluminal closure
does not impinge on intraluminal flow and does
not leave any intra-arterial materials. Hence, one
can use this device in patients who have PVD and
in patients in whom the arterial puncture is below
the femoral artery bifurcation.
The Angiolink vascular closure system (Med-
tronic, Santa Rosa, California) has received FDA
approval as the first staple-mediated closure de-
vice. It consists of three components: a low-profile,
3-mm titanium staple; a one-piece ‘‘three-step’’
introducer assembly that contains an introducer,
a dilator with a blood-marking lumen proximally,
and two small stabilization filaments that stabilize
the vessel wall during staple deployment; and
a trigger-activated staple-deployment device.
To deploy the device, a guidewire is introduced
into the arterial sheath before it is removed. The Fig. 3. Angiolink Staple Device. The first staple-
dilator and introducer are placed over the wire mediated closure device. (Courtesy of Medtronic,
until blood is seen in the blood-marking lumen. Santa Rosa, CA.)
 VASCULAR CLOSURE DEVICES
of the need to have a 7F tissue tract for the device
to deliver the clip. There have been no reported
cases of infection although real-world experience
is not available. The Clip Closure in Percutaneous
Procedures study randomized 208 patients who
underwent diagnostic catheterization to Starclose
or MC, and showed a mean time to hemostasis
of 17 seconds and 15 minutes, respectively, and
a mean time to ambulation of 4 hours and 6
hours, respectively [16]. If future studies show
equivalent efficacy for hemostasis and low compli-
cation rates for the staple-and-clip technology, the
use of suture-mediated and sealant closure devices
may become obsolete (Fig. 4).
Do vascular closure devices lead to more
complications?
Almost every trial has shown that VCDs
decrease the time to hemostasis and the time to
ambulation, and thereby, increase patient comfort
and improve catheterization laboratory efficiency;
however, because each device leaves behind for-
eign material (clip, sealant, suture), infection is
always an increased concern. Because of the cost
of VCDs (w$200–$250 for each device), overall
safety and cost-effectiveness must be confirmed
for most laboratories to invest in their usage.
The safety issue is controversial. Dangas and
colleagues [17] reported that in the early experi-
ence with the first generations of VCDs, vascular
complications were twice as common compared
with MC, especially the frequency of surgical re-
pair, hematoma formation, and the large hemato-
crit drop. This finding is probably due to the
difficulty in placing these devices and the unfamil-
iarity of the devices to most operators at that
Fig. 4. The recently approved Starclose nitonol clip de-
vice. The new Starclose device may replace the popular
Perclose suture-mediated closure device. (Courtesy of
Abbott Vascular, Redwood City, CA.)
285
time. Each device has its own learning curve; if
one becomes proficient in a certain device, vascu-
lar complications become less common.
Koreny and colleagues [18] published a large
meta-analysis of more than 30 randomized trials
(O4000 patients) that reported an increased risk
for hematoma and pseudoaneurysm with the use
of VCDs. This trial looked at a heterogenous pa-
tient population over a long period of time and
probably was flawed by the use of earlier versions
of the devices and the lack of operator comfort.
More recent data showed better results for
VCDs when more recent devices were used.
Meyerson and colleagues [19] reported no differ-
ence in the frequency of obtaining vascular surgi-
cal consults when VCDs were compared with MC
in 4800 patients. Resnic and colleagues [20] stud-
ied more than 3000 patients in their institutional
registry who underwent PCI; most patients were
closed with Angio-Seal. There was a decrease in
the rate of vascular complications in the device
arm (3% versus 5.5%) and a 1-day decrease in
hospital length of stay (2.8 versus 4.0 days). These
indicate the cost effectiveness of these devices if
they decrease hospital complications that lead to
an increased length of stay. Applegate and col-
leagues [21] similarly showed an overall decrease
in the rate of minor and major vascular complica-
tions (1.5% versus 2.5% with MC) when Perclose
was used in 80% of nearly 4800 patients who un-
derwent PCI. This again shows how the frequency
of vascular complications is low when operators
are familiar with a particular device.
The data from the Mt. Sinai School of Med-
icine, where more than 9000 patients have un-
dergone PCI (55% Perclose and 45% Angio-Seal),
show a significant decrease in pseudoaneurysms,
arteriovenous fistulas, and large hematomas com-
pared with MC [22]. This has led to a decrease in
hospital length of stay. Each patient who receives
a VCD is reprepped partially with new towels and
new gloves. Diabetics and obese patients receive
a dose of antibiotics. This has led to almost no
cases of device infection.
Summary
VCDs have become a mainstay in high-volume
cardiac catheterization laboratories because of
their ability to decrease time to hemostasis and
ambulation. It is obvious from patients who
remember the days of only MC that they are
satisfied with their more recent experiences with
286
VCDs. It cannot be overemphasized that only
a laboratory that becomes dedicated in the use of
VCDs and trains a small number of individuals to
place a large number of certain VCDs can succeed
in demonstrating a low rate of vascular compli-
cations that justifies their usage.
Although the efficacy and relative ease of
deployment has pushed Angio-Seal and Perclose
to the top of the field today, the newer staple-
and-clip technologies may become the wave of
the future for their simplicity and ability to close
almost any patient.
References
[1] Meyerson SL, Feldman T, Desai TR. Angiographic
access site complications in the era of arterial closure
devices. Vasc Endovasc Surg 2002;14:652–6.
[2] Carey D, Martin JR, Moore CA, et al. Complica-
tions of femoral artery closure devices. Catheter
Cardiovasc Interv 2001;52:3–7.
[3] Strategic growth opportunities in cardiovascular
interventional treatment drives cardiology sector,
American Health Consultants. BBI Newsletter 2001;
5:1–6.
[4] Seldinger SI. Catheter replacement of the needle in
percutaneous arteriography. Acta Radiol 1953;39:
366–76.
[5] Najjar SF, Healey N, Healey CM, et al. Evaluation
of poly-N-acetyl glucosamine as a hemostatic agent
in patients undergoing cardiac catheterization:
a double-blind, randomized study. J Trauma 2004;
57(1 Suppl):S38–41.
[6] Nader RG, Garcia JC, Drushal K, et al. Clinical
evaluation of SyvekPatch in patients undergoing
interventional, EPS, and diagnostic cardiac catheter-
ization procedures. J Invasive Cardiol 2002;14:
305–7.
[7] Alter BM. Noninvasive Hemostasis Pad. Endovas-
cular Today 2005;April:60.
[8] Fisher TH, Connolly R, Thatte HS, et al. Compari-
son of structural and hemostatic properties of the
poly-N-acetyl glucosamine Syvek Patch with prod-
ucts containing chitosan. Microsc Res Tech 2004;
63:168–74.
[9] Vetter J, Ribeiro E, Hinohara T, et al. Suture medi-
ated percutaneous closure of femoral artery access
sites in fully anticoagulated patients following coro-
nary interventions. Circulation 1994;90:I-621.
[10] Carere RG, Webb JG, Ahmed T, et al. Initial expe-
rience using Prostar: a new device for percutaneous
suture-mediated closure of arterial puncture sites.
Catheter Cardiovasc Diagn 1996;37:367–72.
[11] Sanborn TA, Ogilby JD, Ritter JM, et al. Reduced
vascular complications after percutaneous coronary
KIM
interventions with a nonmechanical suture device:
results from the randomized RACE study. Catheter
Cardiovasc Interv 2004;61:327–32.
[12] Cremonesi A, Castriota F, Tarantino F, et al.
Femoral arterial hemostasis using the Angio-Seal
system after coronary and vascular percutaneous
angioplasty and stenting. J Invas Cardiol 1998;10:
464–9.
[13] Sanborn TA, Gibbs HH, Brinker JA, et al. A
multicenter randomized trial comparing a percuta-
neous collagen hemostasis device with conven-
tional manual compression after diagnostic
angiography and angioplasty. J Am Coll Cardiol
1993;22:1273–9.
[14] Mooney MR, Ellis SG, Gershony G, et al. Imme-
diate sealing of arterial puncture sites after cardiac
catheterization and coronary interventions: initial
US feasibility trial using the Duett vascular clo-
sure device. Catheter Cardiovasc Interv 2000;50:
96–102.
[15] Caputo RP, Ebner A, Grant WG, et al. Percutane-
ous femoral arteriotomy repair: initial experience
with a novel staple closure device. J Invas Cardiol
2002;14:652–6.
[16] Hermiller J, Simonton C, Hinohara T, et al. Clinical
experience with a circumferential clip-based vascular
closure device in diagnostic catheterization. J Inva-
sive Cardiol 2005;17:504–10.
[17] Dangas G, Mehran R, Kokolis S, et al. Vascular
complications after percutaneous coronary interven-
tions following hemostasis with manual compres-
sion versus arteriotomy closure devices. J Am Coll
Cardiol 2001;38:638–41.
[18] Koreny M, Riedmuller E, Nikfardjam M, et al.
Arterial puncture closing devices compared with
standard manual compression after cardiac
catheterization: systematic review and meta-
analysis. JAMA 2004;291:350–7.
[19] Meyerson SL, Feldman T, Desai TR. Angiographic
access site complications in the era of arterial clo-
sure devices. Vasc Endovasc Surg 2002;36(2):
137–44.
[20] Resnic FS, Blake GJ, Ohno-Machado L, et al. Vas-
cular closure devices and the risk of vascular compli-
cations after percutaneous coronary intervention in
patients receiving glycoprotein IIb/IIIa inhibitors.
Am J Cardiol 2001;88:493–6.
[21] Applegate RJ, Grabarczyk MA, Little WC, et al.
Vascular closure devices in patients treated with
anticoagulation and IIb/IIIa receptor inhibitors
during percutaneous revascularization. J Am Coll
Cardiol 2002;40:78–83.
[22] Kim MC, Kini AS, Lee PC. Arterial closure devices
decrease vascular complications after percutaneous
coronary interventions. J Am Coll Cardiol 2004;43:
A53.
 Cardiol Clin 24 (2006) 287–297
Quality Management and Volume-Related Outcomes
in the Cardiac Catheterization Laboratory
Richard L. Snider, MD, Warren K. Laskey, MD*
Department of Medicine, Division of Cardiology, University of New Mexico School of Medicine,
MSC 10-5550, 1 University of New Mexico, Albuquerque, NM 87131, USA
Quality in the cardiac catheterization labora-
tory (CCL) environment can be viewed as the end
result of a dynamic interaction among numerous
factors encompassing clinical, procedural, techni-
cal, cognitive, and process-related elements. Al-
though procedural ‘‘outcomes’’ are the most
conspicuous and increasingly relied-on measure
of quality, it is essential to understand that such
outcomes are a complex and composite mix of
the previously noted general elements. Given the
dependence of quality (and outcomes) on these
elements, the assessment of quality may be viewed
in quantitative terms as the output of a model
whose independent and, it is hoped, scalable
covariates account for most of the variance in
the dependent variable (ie, quality). The assess-
ment of quality, its control, and the effort to
improve quality when the latter is perceived as
a continuous process can be viewed in much the
same way as their counterparts in industry [1,2]. It
is unfortunate that the variability in human dis-
ease and human performancedkey covariates in
any modeldcontinue to elude precise definition
and quantification.
The identification of quality as the principal
focus in the CCL environment [3,4] and the stan-
dards and processes necessary to achieve and im-
prove quality have been described in detail [5,6].
The purposes of the present discussion are (1) to
review the factors identified as necessary elements
of quality in the CCL; (2) to review current ap-
proaches to the assessment of quality (of
* Corresponding author.
E-mail address: wlaskey@salud.unm.edu
(W.K. Laskey).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.01.001
outcomes) in the CCL; (3) to discuss the applica-
tions, implications, and limitations of current
means of assessing quality of outcomes in the
CCL; and (4) to outline areas where there is
need for additional information to more precisely
quantify this elusive variable.
An understanding of the use of outcomes as
a potentially quantifiable measure of quality posits
a consensus regarding terminology and a compre-
hensive assessment of the factors contributing to
such outcomes. Regarding the former, it is un-
fortunate that there is substantial variation in
definitions of outcomes that reflects the heteroge-
neity of the evidence base (ie, observational cohort
studies, randomized controlled clinical trials, post
hoc retrospective analyses, and so forth). Outcome
data obtained from safety-oriented clinical trials
are more likely to include or emphasize traditional
‘‘hard’’ end points (eg, death, stroke, myocardial
infarction [MI]), whereas outcome data obtained
from efficacy-oriented clinical trials are more likely
to include ‘‘softer’’ end points (eg, repeat revascu-
larization, repeat hospitalization). Although there
is unanimity regarding the hierarchic categoriza-
tion of (adverse) clinical outcomesddeath fol-
lowed by MI followed by emergent coronary
bypass surgery, and so forthdthere is less agree-
ment regarding the proper place of more recently
recognized adverse sequelae of catheter-based pro-
cedures such as postprocedural cardiac biomarker
detection and contrast-associated nephropathy.
There is also substantial variation in agreement
regarding the time window of ascertainment of
outcomes (eg, in-hospital, postdischarge, 30 days).
Such variability in the ascertainment of the de-
pendent variable (outcome) presents its own obvi-
ous set of difficulties regarding an association (or
cardiology.theclinics.com
288 SNIDER & LASKEY
lack thereof) with important exposure variables
(see later discussion). Thus, in studies or ap-
proaches to the assessment of quality in the
CCL, a clear and consistent definition of the
outcome is essential. The overwhelming majority
of outcome studies from the CCL environment
focus on adverse outcomes that have a low fre-
quency, which often presents insurmountable
statistical issues [7]. Curiously, the reporting of
beneficial outcomes is seldom found in the quality
and outcome literature for the CCL.
Among the numerous factors contributing to
CCL outcomes, operator-specific elements (tech-
nical competency, clinical judgment, cognitive
base) and patient-specific elements (clinical pre-
sentation, lesion characteristics) have received the
most scrutiny [8]. Additional important (although
less well studied) contributions to outcomes arise
from the ‘‘facility’’ and the continuous quality im-
provement process (Table 1). The number and the
diversity of these factors underscore the critical
importance of estimating risk-adjusted outcomes
[9], notwithstanding the potential imprecision of
this estimate owing to lack of transportability of
models over time or geography [10–12]. Most
studies have focused on in-hospital mortality or
a composite major adverse clinical event rate as
the outcome variable of interest; however, low
event (mortality) rates remain a significant limita-
tion to meaningful interpretation and extrapola-
tion of these models to other settings [10,11]. In
addition, the use of standardized mortality rates
is subject to similar limitations owing to instability
of the denominator (which is derived from such
models).
It is unfortunate that there is a paucity of
literature analyzing the comprehensive relation-
ship between the elements outlined in Table 1 and
the composite end point of in-hospital adverse
Table 1
Factors contributing to clinical outcomes in the cardiac
catheterization laboratory
System/
Operator-specific Patient-specific process - specific
Competency/ Demographic Institutional volume
technical features
proficiency
Experience base Clinical features Clinical QA/CQI
process
Cognitive base Lesion features Technical QA/CQI
process
Abbreviations: CQI, continuous quality improvement;
QA, quality assessment.
outcomes (death/MI/coronary bypass surgery/
stroke). As these latter measures of outcome con-
tinue to improve [13], the earlier-noted limitations
of mortality-centered models will also extend to
these latter models. For this reason, several inves-
tigators have proposed using adverse outcome
rates at 30 days [10,14,15], although no models in-
corporating the variables from Table 1 have been
examined in this setting. Further complicating this
issue is the absence of scalable covariates that
capture the full dimensions of the ‘‘process’’ or
‘‘system’’ pieces.
Recently, the addition of procedural volume
(institutional and operator-specific) has been
added to the list of covariates in statistical models
of CCL procedural outcomes [16,17]. Although
attractive by virtue of its immediacy and simplic-
ity, rigorous study of the accuracy, validity, and
incremental value of this parameter to the classic
risk- adjusted models of procedural outcome is
lacking. Nevertheless, the currently accepted asso-
ciation between procedural volume (operator level
and institutional level) and outcome is increas-
ingly used as an index of (if not a surrogate for)
quality and therefore deserves careful analysis.
The volume–outcome relationship
Historically, the relationship between proce-
dural volume and clinical outcome has been
attributed to the idea that ‘‘practice makes per-
fect.’’ A wide variety of health care–related
treatment strategies and procedures have been
evaluated, including those related to cancer and
HIV, congenital and acquired cardiac disease, and
peripheral vascular disease. The data extracted
from these studies have become the subject of
much interest for health insurance purchasers,
public consumers of health care, hospital admin-
istrators, and state regulators.
The surgical experience
The relationship between surgical procedural
volume and clinical outcome has been extensively
evaluated. As early as 1979, Luft and colleagues
[18] examined clinical outcomes of 12 different
surgical procedures varying in complexity. Mor-
tality rates were reduced when open heart and
coronary artery bypass graft (CABG) surgery,
vascular surgery, or transurethral resection of
the prostate were performed in centers that had
an experience of greater than 200 procedures per
year compared with lower-volume institutions.
 QUALITY IN THE CARDIAC CATHETERIZATION LABORATORY
More recently, large population–based stu-
dies have demonstrated a reduced mortality rate
in higher-volume centers compared with lower-
volume centers [19–21]. The magnitude of differ-
ence and the threshold at which these differences
are seen are variable among different procedures.
For example, surgical mortality was studied in
the Medicare population using Medicare Provider
Analysis and Review files from the Center for
Medicare and Medicaid Services from 1994
through 1999 [21]. Analysis performed on 14 dif-
ferent surgical procedures showed improved
observed and adjusted mortality rates for higher-
volume compared with lower-volume centers.
The largest adjusted differences in mortality rates
were seen in patients undergoing pancreatic resec-
tion (16.3% versus 3.8%), esophagectomy (20.3%
versus 8.4%), and pneumonectomy (16.1% versus
10.7%) when very low volume and very high vol-
ume centers were compared. The smallest differ-
ence between low- and high-volume centers was
seen in the group undergoing carotid endarterec-
tomy (1.7% versus 1.5%).
The percutaneous transluminal coronary
angioplasty experience
The first studies evaluating the volume–out-
come relationship within the realm of percutane-
ous coronary intervention (PCI) included patients
undergoing percutaneous transluminal coronary
angioplasty (PTCA) (Table 2). In 1989, Ritchie
and colleagues [22] analyzed outcomes following
PTCA in over 24,000 patients in nonfederal hospi-
tals in California. Unadjusted in-hospital mortal-
ity was similar in admitted patients who had
acute MI (AMI) or who did not have AMI among
low-, intermediate-, and high-volume institutions.
The unadjusted same-stay CABG rate and the
composite rate of same-stay CABG or death (for
the AMI and non-AMI groups), however, were
significantly lower in the high-volume institutions.
Jollis and colleagues [23] analyzed a large data-
set of Medicare beneficiaries from 1987 to 1990
and found that in-hospital and 30-day mortality
rates and same-stay CABG rates declined when
highest to lowest deciles of hospital PTCA volume
were compared. These differences remained signif-
icant after adjustment for age, sex, race, and year
of procedure. Individual operator and hospital
volumes were analyzed when the Medicare popu-
lation was studied again in 1992 [24]. There was
no in-hospital or 30-day mortality difference as
annual physician volume increased, but the rates
289
of same-stay CABG were higher in the low-
volume group. Hospital volume remained a statis-
tically significant risk factor for mortality and
CABG, although the differences among volume
groups were smaller compared with what was
seen in the earlier report [23].
Examination of the Society for Cardiac Angi-
ography and Interventions registry allowed Kim-
mel and colleagues [25] to collect detailed clinical
patient information and thereby further adjust
for differences among the volume strata. Despite
the low overall rates for death and emergency
CABG, a volume-dependent variation in out-
comes was seen. A statistically significant decrease
in the composite end point of death, emergency
CABG, and postprocedural MI in laboratories
performing more than 400 procedures per year
was present compared with those performing
fewer than 400. When multivariable logistic re-
gression models were used to adjust for clinical
characteristics (eg, chronic renal insufficiency,
emergency PTCA, left main angioplasty, and
shock, which were all more common in lower-vol-
ume centers; and congestive heart failure, recent
MI, and complex lesion morphology, which were
all more common in higher-volume centers), a sta-
tistically significant difference in emergency
CABG, MI, and composite of major complica-
tions remained.
Hannan and colleagues [26] analyzed the rela-
tionships between hospital volume or individual
operator volume and clinical outcomes by extract-
ing data from the Coronary Angioplasty Report-
ing System (CARS). This system contains
clinical information that is more detailed than
what is found in administrative datasets. The
study, which included over 60,000 patients under-
going PTCA in New York State from 1991 to
1994, found statistically significant higher risk-ad-
justed rates of mortality and same-stay CABG in
the lowest hospital volume and lowest individual
volume groups compared with the state as a whole.
Further analysis indicated that high-volume oper-
ators working in high-volume centers tended
to have improved outcomes compared with low-
volume operators working in high-volume centers.
Experience in the current (stent) era
With advances in technology and pharmacol-
ogy, there has been a decline in rates of post-
procedural MI, CABG surgery, and in-hospital
mortality [27–30]. The effect of these new thera-
pies on the volume–outcome relationship has
 290
Table 2
Percutaneous transluminal coronary angioplasty volume–outcome studies
Effect size:
Definition Outcome lowest-minus
Year/population Unit of analysis of volume (overall highest-volume
Study [ref.] studied Data source (no. in group) (cases per year) event rate) groupa
Ritchie et al, 1989 multicenter Administrative Hospital (110) Low (!200) In-hospital 0.9% in AMI
1993 [22] High (O400) mortality (1.4%) 0.0% in non-AMI
Same-stay CABG (5%) 3.7% in AMI
1.9% in non-AMI
Jollis et al, 1987–1990 Administrative Hospital (1194) Deciles In-hospital 1.4%
1994 [23] multicenter Low (!47) mortality (2.9%)
High (O371) 30-day mortality (3.2%) 1.5%
In-hospital CABG (3.8%) 2.5%
Jollis et al, 1992 multicenter Administrative Hospital (984) and Physician In-hospital Physician: 0.1%

SNIDER & LASKEY

1997 [24] individual (6115) Low (!25) mortality (2.5%) Hospital: 0.6%
High (O50)
Hospital 30-day mortality (2.9%) Physician: 0.1%
Low (!100) Same-stay Hospital: 1.0%
High (O200) CABG (3.3%) Physician: 1.2%
Hospital: 0.9%
Kimmel et al, 1992–1993 Clinical Hospital (48) Low (!200) In-hospital 0.2%
1995 [25] multicenter High (R600) mortality (0.25%)
Emergency 0.7%
CABG (1.3%)
Hannan et al, 1991–1994 Clinical Hospital (31) and Quintiles In-hospital Physician: 0.06%b
1997 [26] multicenter individual Physician mortality (0.90%) Hospital: 0.17%b
(not available) Low (!75) Physician: 0.87%b
High (R250) Same-stay CABG (3.43%) Hospital: 0.95%b
Hospital
Low (!400)
High (R1000)
Abbreviation: AMI, acute myocardial infarction.
a
Rates presented are the unadjusted difference between the lowest and highest volume for the outcome shown unless otherwise specified.
b
Risk adjusted difference.
 QUALITY IN THE CARDIAC CATHETERIZATION LABORATORY
been addressed in single and multicenter studies
(Table 3).
Kastrati and colleagues [31] evaluated out-
comes in 3409 consecutive patients who under-
went coronary stent implantation performed by
10 different operators in a single center from
1992 to 1997. Using a composite end point of car-
diac death, MI, or CABG at 30 days post proce-
dure, the investigators found that compared with
the overall event rate, operators performing
greater than 483 procedures annually had better
outcomes, whereas operators performing less
than 90 procedures annually had worse outcomes.
Using a classification and regression tree analysis,
the investigators also found that minimum proce-
dural volume appeared to be a determinant of
outcome, even for patients who had less complex
lesion types.
Data from larger multicenter samples such as
the Medicare population and information from
the National In-Patient Sample (NIS) have also
been published. McGrath and colleagues [32] ex-
amined individual and hospital volume and the
relation to outcome for over 167,000 Medicare
patients who underwent PCI in 1997. The differ-
ence in crude rates of CABG between high- and
low-volume hospitals became nonsignificant
when adjustment was made for clinical risk fac-
tors; however, a difference in 30-day mortality
favoring high-volume centers persisted. When
physician volume was examined, differences in
crude 30-day mortality became statistically insig-
nificant when outcomes were adjusted for risk;
however, a difference in risk-adjusted rates of
CABG occurred in favor of high- versus low-vol-
ume strata. Investigators from an NIS-based
study [33] compared high- versus low-volume cen-
ters and demonstrated a statistically significant
decrease in in-hospital mortality of patients who
did and did not have AMI in 1997. Analysis of
the same database over the interval from 1998 to
2000 showed a similar trend in mortality benefit
in high-volume centers [34]. No difference in
risk-adjusted in-hospital mortality was found,
however, when intermediate-volume centers (per-
forming 200–399 annual PCIs) were compared
with those performing more than 400 annual
PCIs.
In a recent study from a high-volume (O4000
interventional procedures per year) institution
from 1999 to 2001, no association between oper-
ator volume and adverse outcomes (crude and
risk-adjusted in-hospital death and the composite
of death, CABG surgery, MI, or stroke) could be
291
identified [35]. These latter findings are similar to
an earlier publication by the Northern New En-
gland Cardiovascular Disease Study Group [36]
that showed that there was no statistical difference
in risk-adjusted in-hospital mortality, same-stay
CABG, MI, or clinical success among terciles of
operator volume. Centers included in this study
performed greater than 600 procedures per year.
The acute myocardial infarction experience
Risk-adjustment models have been used to
correct for differences in patient characteristics
that may affect crude rates of outcome and their
association with volume. Patients presenting with
AMI are a high-risk group, with overall higher
rates of in-hospital death compared with patients
who present with stable coronary syndromes.
Although many studies of the volume–outcome
relation have included patients who have AMI in
their analyses, few have looked at this patient
population exclusively (Table 4).
Using information from the Cooperative Car-
diovascular Project, Thiemann and colleagues [37]
noted that in 98,898 Medicare patients treated
with primary angioplasty, thrombolysis, or con-
servative medical management for AMI, crude
and risk-adjusted 30-day mortality rates were
lower in the highest-volume quartile compared
with the lowest-volume quartile. In addition, anal-
ysis of medical therapy received by the different
groups showed a higher use of aspirin, b-blockers,
and other appropriate pharmacologic therapies in
the higher-volume groups, which accounted for
about one third of the difference in outcome.
Vakili and colleagues [38] analyzed the CARS
database in 1995 and found that when hospital
volume was divided into terciles, there was no cor-
relation between volume and in-hospital mortality
for crude or risk-adjusted outcome in patients un-
dergoing primary angioplasty for AMI. A lower
mortality rate, however, was noted in the highest
tercile of individual procedural volume compared
with the lowest. When the same patient population
was reanalyzed, but the categorization of individ-
ual procedural volumes was changed to correlate
with current American College of Cardiology/
American Heart Association recommendations
(!75 cases per year considered low individual vol-
ume; !400 cases per year considered low hospital
volume), individual volume did not have an effect
on in-hospital mortality, and high-volume hospi-
tals had a statistical advantage over low-volume
 292
Table 3
Stent-era volume–outcome studies
Effect size:
Year/ Unit of Definition of Outcome lowest-minus
Stents used population analysis volume (cases (overall highest-volume
Study [ref.] in PCI studied Data source (no. in group) per year) event rate) groupa
Kastrati et al, 100% 1992–1997 Clinical Individual Quintiles 30-day combined 2.5%
1998 [31] single-center (10) Low (!90) cardiac death, MI,
High (O242) CABG (2.99%)
Malenka et al, Nearly 50% 1994–1996 Clinical Individual Terciles In-hospital mortality in
1999 [36] by 1996 multicenter (47) Low (22–84) Low risk (NA) ÿ0.26%
High (R138) High risk (NA) ÿ1.55%
Same-stay CABG (NA) 0.47%
Watanabe et al, 59% Only 1997 data Administrative Individual Low (%200) In-hospital mortality in
2002 [33] presented here (NA) High (O400) AMI (3.5%) 1.7%
multicenter Non-AMI (0.8%) 0.2%
Same-stay CABG in

SNIDER & LASKEY

AMI (2.9%) ÿ0.4%
Non-AMI (1.8%) 0.4%
McGrath et al, 48.4%–61.1% 1997 multicenter Administrative Hospital Individualb 30-day mortality Individual:ÿ0.14%
2000 [32] (1003) and Low (!30) (3.30%) Hospital: 1.14%c
individual High (O60) Same-stay CABG Individual: 0.70%
(6534) Hospitalb (1.87%) Hospital: 0.0%c
Low (!80)
High (O160)
Epstein et al, 81.7% 1998–2000 Administrative Hospital (457) Low (!200) In-hospital mortality 1.2% (low-minus very
2004 [34] multicenter Intermediate (200–399) (1.58%) high volume group)
High (400–999) 0.19% (intermediate
Very high (R1000) minus high-volume
group)
Harjai et al, 71% 1999–2001 Clinical Individual (28) Terciles In-hospital mortality ÿ0.22%
2004 [35] single-center Low (!93) (0.99%)
High (O140) Composite death, MI, 0.11%
CABG, CVA (2.59%)
Abbreviations: CVA, cerebrovascular accident; NA, not available.
a
Rates presented are the unadjusted difference between the lowest and highest volume for the outcome shown unless otherwise specified.
b
Low and high volumes correspond to individual volumes of !75 and O150 and hospital volumes of !200 and O400 if estimated Medicare volume is 40% of total volume.
c
Risk adjusted difference.
Table 4
Primary angioplasty volume–outcome studies
Definition of Effect size: lowest-
Year/population Unit of analysis volume (cases minus highest-
Study [ref.] studied Data source (no. in group) per year) Outcome volume groupa
Vakili et al, 2001 [38] 1995 multicenter Clinical Hospital (32) and Terciles In-hospital mortality Individual: 3.3%b
Hospital: 1.8%b

QUALITY IN THE CARDIAC CATHETERIZATION LABORATORY

individual (151) Individual
Low (1–10)
High (R11)
Hospital
Low (1–56)
High (R57)
Vakili et al, 2003 [39] 1995 multicenter Clinical Hospital (32) and Individualc In-hospital mortality Individual: 1.3%a
Individual (151) Low (!75) Hospital: 3.8% b
High (R75)
Hospitalc
Low (!400)
High (R400)
Magid et al, 2000 [40] June 1, 1994–July 31, Clinical Hospital (446) Low (!16) In-hospital mortality 2.8%
1999 multicenter High (R49) (1 angioplasty)
Thrombolysis 0.5%
Canto et al, 2000 [41] June 1994–March Clinical Hospital (450) Quartiles In-hospital mortality 2.0%
1998 multicenter Low (5–11) (1 angioplasty)
High (O33) Thrombolysis ÿ0.1%
Tsuchihashi et al, 1997 multicenter Clinical Hospital (129) Terciles In-hospital mortality 1.0%b
2004 [42] Low (1–16) Same-stay CABG 0.9%b
High (56–370)
a
Rates presented are the unadjusted difference between the lowest and highest volume for the outcome shown unless otherwise specified.
b
Multivariate risk adjustment did not change statistical outcome.
c
Categories are based on total annual PTCA volume shown in parentheses for individual and hospital; number of primary angioplasties not available.

293
294 SNIDER & LASKEY
hospitals [39]. Statistical significance remained
when mortality rates were risk adjusted.
Using the National Registry of Myocardial
Infarction (NRMI) database from June 1994
through July 1999, Magid and colleagues [40] com-
pared in-hospital mortality in over 62,000 patients
presenting with AMI treated with primary angio-
plasty or thrombolytic therapy as a function of hos-
pital primary angioplasty volume. Mortality rates
were significantly lower in the intermediate- and
high-volume centers for patients who underwent
primary angioplasty compared with those who un-
derwent thrombolysis. Low-volume centers, how-
ever, demonstrated no difference in mortality
when these two treatment modalities were com-
pared. Although no direct statistical comparison
of outcome was made among the low-, intermedi-
ate-, and high-volume hospitals in the primary an-
gioplasty group alone, there was a trend toward
decreasing in-hospital mortality with increasing
volume. Other investigators in the NRMI registry
[41] collected data from June 1994 to March 1998
and found an association between in-hospital mor-
tality and hospital volume in patients undergoing
primary PCI for AMI. Crude and risk-adjusted
mortality rates were lower in the highest-volume
quartile compared with the lowest-volume quartile.
These investigators could not demonstrate an asso-
ciation between hospital volume and mortality
when thrombolytic therapy was used as the reperfu-
sion strategy. In 1997, investigators of the Japanese
Coronary Intervention Study [42] evaluated 2491
patients who underwent PCI for AMI and found
no difference in crude or risk-adjusted in-hospital
mortality rate or same-stay CABG rate among ter-
ciles of hospital volume.
Longitudinal studies
A decline in overall risk-adjusted mortality and
same-stay CABG rates from 1990 to 1997 was noted
in a report from the Northern New England
Cardiovascular Disease Study Group [30]. Trends
in improving clinical success rates of PCI, MI, and
same-stay CABG were seen across the 8-year period
in unadjusted and risk-adjusted rates despite a con-
comitant increase in the prevalence of risk factors
(eg, age, diabetes, renal failure, previous CABG,
and left main disease). Although this study was not
designed to evaluate physician or hospital volume,
the investigators contended that the improved out-
comes were related, in part, to an increase in physi-
cian and hospital volume, with the most significant
contribution being from technologic advances.
As rates of adverse outcomes continue to
decline along with improvements in operator skill
and technology, the magnitude of the difference in
outcomes between high-and low-volume groups
is likely to diminish. An analysis of procedural
outcomes in California addresses the volume–
outcome relationship among low-, intermediate-,
and high-volume facilities over time [43]. Files
from the Office of Statewide Health Planning
and Development in California were reviewed,
and data were extracted from charts of 353,488
patients who were admitted after undergoing
coronary intervention between 1984 and 1996. A
trend in decreasing mortality and CABG rates
was seen as procedure volume increased; however,
the magnitude of difference among the different
volume groups decreased over time (Fig. 1). This
finding was true for crude and risk-adjusted rates.
Fig. 2 shows that with the addition of more recent
data from Epstein and colleagues [34], the differ-
ence in mortality when comparing intermediate-
volume (200–399 cases per year) to high-volume
(R400 cases per year) centers continues to de-
crease over time. Centers performing fewer than
200 cases per year, however, continue to demon-
strate higher mortality rates [34].
Summary
In this article, the authors have attempted to
summarize the factors necessary for the compre-
hensive assessment of quality and outcomes in the
CCL environment. The focus has been on the
6
5.5
Effect Size: Lowest minus highest
Same-stay CABG
In-hospital Mortality
5
volume strata (%)
4
3
2.2
2
1.2 1.3
1
0.6
0.4
0
1984 to 1987 1988 to 1992 1993 to 1996
Time Period
Fig. 1. Diminishing differences in outcome between low-
volume (!200 cases per year) and high-volume (R400
cases per year) centers in California over time. (Data
from Ho V. Evolution of the volume-outcome relation
for hospitals performing coronary angioplasty. Circula-
tion 2000;101:1806–11.)
 QUALITY IN THE CARDIAC CATHETERIZATION LABORATORY
Fig. 2. Time-dependent changes in the difference in
mortality between intermediate-volume (200–399 cases
per year) and high-volume (R400 cases per year) cate-
gories. (Data from Ho V. Evolution of the volume-
outcome relation for hospitals performing coronary
angioplasty. Circulation 2000;101:1806–11 and Epstein
AJ, Rathor SS, Volpp KG, et al. Hospital percutaneous
coronary intervention volume and patient mortality,
1998 to 2000. Does the evidence support current proce-
dure volume minimums? J Am Coll Cardiol 2004;43:
1755–62.)
association between outcome measures and pro-
cedural volume and on the methodologies used in
such studies; however, the authors have also
introduced a note of caution into unhesitating
acceptance of the quality-outcomes-volume syllo-
gism. Although there is, overall, a statistically
significant association between procedural out-
comes and volume, the magnitude of this associ-
ation is highly dependent on (1) the type and
frequency of the outcome; (2) the clinical setting;
(3) the nature of the measuring instrument (ad-
ministrative versus clinical database); (4) the unit
of analysis (individual operator versus institu-
tion); (5) the effects of temporal changes in clinical
practice and outcomes; (6) the cut points used
to define volume; (7) the robustness and trans-
portability of risk-adjustment models (for pre-
diction of outcomes); and (8) system and process
considerations.
Although the use of clinical outcomes as an
indicator of quality is certainly reasonable, out-
come assessment alone does not capture the
universe of the elements of quality or aid us in
understanding the processes of quality control
and quality improvement. The latter should be
highly visible and ‘‘real.’’ A closely related
question is, If we could impact one of the more
important aspects of the process of care in the
295
CCL, would we favorably alter outcomes? One
study suggests so [44], and the impact on out-
comes demonstrated, for the first time, that
such efforts have meaningful sequelae. Many
more studies of a similar nature under widely
varying clinical, demographic, and temporal con-
ditions will be required before we are ready to
accept procedural outcomes as the sole measure
of quality. Similar concerns arise regarding the
use of procedural volume as a surrogate of qual-
ity. As discussed herein, there is sufficient vari-
ability in the measures currently used to assess
this relationship such that considerable circum-
spection is advised for the unwary. The impor-
tance of the definition of volume is underscored
by sequential studies from the CARS registry
[38,39] of the volume–outcome relationship in
patients undergoing PCI for AMI. Using the
same patient population but varying the cut
points for low and high volume, these studies
demonstrated discordant conclusions regarding
a volume–outcome relationship.
Finally, given the interstudy variation in the
extent of an association between volume and
outcome and the decreasing absolute magnitude
of difference in outcomes across quantiles, the
inability to embrace fully a quality-outcome-
volume syllogism must be acknowledged. A
greater appreciation of the merits and limitations
of the methodology used to analyze the associ-
ation between volume and outcomes is needed.
All such associations derived from observational
studies must be scrutinized for the effects of
confounding and bias because the latter may
certainly influence (positively or negatively) the
measure of that association. Furthermore, given
the diminishing absolute differences in mortality
across strata of volumes and the ever-increasing
size of datasets, the question of statistical versus
clinical relevance assumes greater importance.
Certainly, from a population-based perspective,
small percentage differences multiplied across
a large sample size will result in a significant
number of patients overall. The extension of these
population-based data to the individual patient
becomes problematic. From an individual per-
spective, such small percentage differences may be
difficult to translate into clinical relevance.
References
[1] Grant E, Leavenworth R. Statistical quality control.
7th edition. New York: McGraw Hill; 1966.
296 SNIDER & LASKEY
[2] Available at: http://www.ge.com/sixsigma/SixSigma/
pdf. Accessed July 1, 2005.
[3] Sones FM Jr. The Society for Cardiac Angiography.
Cathet Cardiovasc Diag 1978;4:233–4.
[4] Hildner FJ. Quality is the only issue. Cathet Cardio-
vasc Diag 1990;3:216–7.
[5] Bashore TM, Bates ER, Berger PB, et al. Cardiac
catheterization laboratory standards: a report of
the American College of Cardiology Task Force
on Clinical Expert Consensus Documents (ACC/
SCAI committee to develop an expert consensus
document on catheterization laboratory standards).
J Am Coll Cardiol 2001;37:2170–214.
[6] Society for Cardiac Angiography and Interventions.
Monograph on quality management in the cardiac
catheterization laboratory. Bethesda (MD): SCAI;
1999.
[7] Ellis SG, Omoigui N, Bittl JA, et al. Analysis and
comparison of operator-specific outcomes in inter-
ventional cardiology: from a multi-center database
of 4860 quality-controlled procedures. Circulation
1996;93:431–9.
[8] Hirshfeld JW Jr, Ellis SG, Faxon DP, et al. Recom-
mendations for the assessment and maintenance of
proficiency in coronary interventional procedures:
statement of the American College of Cardiology.
J Am Coll Cardiol 1998;31:722–43.
[9] Block PC, Peterson EC, Krone R, et al. Identifica-
tion of variables needed to risk adjust outcomes of
coronary interventions: evidence-based guidelines
for efficient data collection. J Am Coll Cardiol
1998;32:275–82.
[10] Hannan EL, Wu C. Assessing quality and out-
comes for percutaneous coronary intervention:
Choosing statistical models, outcomes, time periods
and patient populations. Am Heart J 2003;145:
571–4.
[11] Kizer JR, Berlin JA, Laskey WK, et al. Limitations
of current risk-adjustment models in the era of coro-
nary stenting. Am Heart J 2003;145:683–92.
[12] Holmes DR, Selzer F, Johnston JM, et al. Modeling
and risk prediction in the current era of intervention-
al cardiology. Circulation 2003;107:1871–6.
[13] Williams DO, Holubkov R, Yeh W, et al. Percutane-
ous coronary intervention in the current era com-
pared with 1985–86: the National Heart, Lung and
Blood Institute Registries. Circulation 2000;102:
2945–51.
[14] Lindsay J, Pinnow EE, Pichard AD. Benchmarking
operator performance in percutaneous coronary in-
tervention: a novel approach using 30-day events.
Cathet Cardiovasc Interv 2001;52:139–45.
[15] Laskey WK, Selzer F, Jacobs AK, et al. Importance
of the post-discharge interval in assessing major ad-
verse clinical event rates following percutaneous cor-
onary intervention. Am J Cardiol 2005;95:1135–9.
[16] Phillips KA, Luft HS, Ritchie JL. The association of
hospital volumes of percutaneous transluminal cor-
onary angioplasty with adverse outcomes, length
of stay and charges in California. Med Care 1995;
33:502–14.
[17] Hannan EL, Racz M, Ryan TJ, et al. Coronary an-
gioplasty volume-outcome relationships for hospi-
tals and cardiologists. JAMA 1997;277:892–8.
[18] Luft HS, Bunker JP, Enthoven AC. Should opera-
tions be regionalized? The empirical relation be-
tween surgical volume and mortality. N Engl J
Med 1979;301:1364–9.
[19] Hannan EL, O’Donnell JF, Kilburn H Jr, et al. In-
vestigation of the relationship between volume and
mortality for surgical procedures performed in
New York State hospitals. JAMA 1989;262:503–10.
[20] Finlayson EV, Goodney PP, Birkmeyer JD. Hospi-
tal volume and operative mortality in cancer sur-
gery: a National Study. Arch Surg 2003;138:882–90.
[21] Birkmeyer JD, Siewers AE, Finlayson E, et al. Hos-
pital volume and surgical mortality in the United
States. N Engl J Med 2002;346:1128–37.
[22] Ritchie JL, Phillips KA, Luft HS. Coronary angio-
plasty: statewide experience in California. Circula-
tion 1993;88:2735–43.
[23] Jollis JG, Peterson ED, DeLong ER, et al. The rela-
tion between the volume of coronary angioplasty
procedures at hospitals treating Medicare beneficia-
ries and short-term mortality. N Engl J Med 1994;
331:1625–9.
[24] Jollis JG, Peterson ED, Nelson CL, et al. Relation-
ship between physician and hospital coronary angio-
plasty volume and outcome in elderly patients.
Circulation 1997;95:2485–91.
[25] Kimmel SE, Berlin JA, Laskey WK. The relation-
ship between coronary angioplasty procedure vol-
ume and major complications. JAMA 1995;274:
1137–42.
[26] Hannan EL, Racz M, Ryan TJ, et al. Coronary an-
gioplasty volume-outcome relationships for hospi-
tals and cardiologist. JAMA 1997;277(11):892–8.
[27] Kimmel SE, Localio AR, Krone RJ, et al. The effects
of contemporary use of coronary stents on in-hospi-
tal mortality. Registry Committee of the Society for
Cardiac Angiography and Interventions. J Am Coll
Cardiol 2001;37(2):499–504.
[28] Maynard C, Wright SM, Every NR, et al. Compar-
ison of outcomes of coronary stenting versus con-
ventional coronary angioplasty in the department
of veterans affairs medical centers. Am J Cardiol
2001;87:1240–5.
[29] Use of a monoclonal antibody directed against the
platelet glycoprotein IIb/IIIa receptor n high-risk
coronary angioplasty. The EPIC Investigation. N
Engl J Med 1994;330:956–61.
[30] McGrath PD, Malenka DJ, Wennberg DE, et al.
Changing outcomes in percutaneous coronary inter-
ventions. A study of 34,752 procedures in northern
New England, 1990 to 1997. J Am Coll Cardiol
1999;34:674–80.
[31] Kastrati A, Neuman FJ, Shomig A. Operator
volume and outcome of patients undergoing
 QUALITY IN THE CARDIAC CATHETERIZATION LABORATORY
coronary stent placement. J Am Coll Cardiol 1998;
32:970–6.
[32] McGrath PD, Wennberg DE, Dickens JD, et al. Re-
lation between operator and hospital volume and
outcomes following percutaneous coronary inter-
ventions in the era of the coronary stent. JAMA
2000;284:3139–44.
[33] Watanabe CT, Maynard C, Ritchie JL. Short-term
outcomes after percutaneous coronary intervention:
effects of stenting and institutional volume shifts.
Am Heart J 2002;144:309–14.
[34] Epstein AJ, Rathor SS, Volpp KG, et al. Hospital
percutaneous coronary intervention volume and pa-
tient mortality, 1998 to 2000. Does the evidence sup-
port current procedure volume minimums? J Am
Coll Cardiol 2004;43:1755–62.
[35] Harjai KJ, Berman AD, Grines CL, et al. Impact of
interventionalist volume, experience, and board cer-
tification on coronary angioplasty outcomes in the
era of stenting. Am J Cardiol 2004;94:421–6.
[36] Malenka DJ, McGrath PD, Wennberg DE, et al.
The relationship between operator volume and out-
comes after percutaneous coronary interventions
in high volume hospitals in 1994–1996. J Am Coll
Cardiol 1999;34:1471–80.
[37] Thiemann DR, Coresh J, Oetgen WJ, et al. The as-
sociation between hospital volume and survival
after acute myocardial infarction in elderly patients.
N Engl J Med 1999;340:1640–8.
[38] Vakili BA, Kaplan R, Brown DL. Volume-outcome
relation for physicians and hospitals performing
297
angioplasty for acute myocardial infarction in
New York State. Circulation 2001;104:2171–6.
[39] Vakili BA, Brown DL. Relation of total annual cor-
onary angioplasty volume of physicians and hospi-
tals on outcomes of primary angioplasty for acute
myocardial infarction (data from the 1995 Coronary
Angioplasty Reporting System of the New York
State Department of Health). Am J Cardiol 2003;
91(6):726–8.
[40] Magid DJ, Calonge BN, Rumsfeld JS, et al. Relation
between hospital primary angioplasty volume and
mortality for patients with acute MI treated with pri-
mary angioplasty vs thrombolytic therapy. JAMA
2000;284:3131–8.
[41] Canto JG, Every NR, Magid DJ, et al. The volume
of primary angioplasty procedures and survival after
acute myocardial infarction. N Engl J Med 2000;
342:1573–80.
[42] Tsuchihashi M, Tsutsui H, Tada H, et al. Volume-
outcome relation for hospitals performing angio-
plasty for acute myocardial infarction: results from
the Nationwide Japanese Registry. Circ J 2004;68:
887–91.
[43] Ho V. Evolution of the volume-outcome relation for
hospitals performing coronary angioplasty. Circula-
tion 2000;101:1806–11.
[44] Moscucci M, Share D, Kline-Rogers E, et al. The
Blue Cross Blue Shield of Michigan Cardiovascular
Consortium (BMC2) collaborative quality improve-
ment initiative in percutaneous coronary interven-
tions. J Interv Cardiol 2002;15:381–6.
 Cardiol Clin 24 (2006) xiii

Foreword

Michael H. Crawford, MD
Consulting Editor

The traditional 12-lead ECG continues to be detecting cardiac chamber enlargement or

the first-line test in the evaluation of chest pain
syndromes. Characteristic ECG changes help de-
fine myocardial infarction and ischemia. Quality
assurance algorithms usually require an ECG
within 10 minutes of hospital arrival for the com-
plaint of chest discomfort. It is also the mainstay
in the assessment of chest pain after percutaneous
coronary interventions and in the assessment of
the adequacy of reperfusion after thrombolytic
therapy. Thus, the ECG is the stalwart of ischemic
heart disease management.
An ECG is often the first step in evaluating
patients who have palpitations, syncope, or sus-
pected cardiac arrhythmias. It may detect fre-
quent or persistent arrhythmias, and it may reveal
the substrate for arrhythmias such as left atrial
enlargement, hypokalemia, long QT interval, and
pre-excitation. In addition, certain ECG abnor-
malities may indicate a higher risk of sudden car-
diac death. Finally, the ECG is useful for
hypertrophy.
Dr. Barold has assembled a world-class group
of physicians to write the articles that appear in
this issue. They are all ECG experts who keep the
flame alive in a field that is mature yet not static.
In addition to discussing the clinical issues men-
tioned above, important technical issues, pitfalls,
and artifacts are covered. Computerized ECG
interpretation is also discussed. Anyone who cares
for patients will benefit from reading this issue.
Michael H. Crawford, MD
Division of Cardiology
Department of Medicine
University of California
San Francisco Medical Center
505 Parnassus Ave., Box 0124
San Francisco, CA 94143-0124, USA
E-mail address: crawfordm@medicine.ucsf.edu

0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.005 cardiology.theclinics.com
 Cardiol Clin 24 (2006) xv–xvi

Preface
Advanced 12-Lead Electrocardiography

S. Serge Barold, MD
Guest Editor

The electrocardiogram (ECG) is the oldest and
the most commonly used cardiology procedure.
The year 2002 marked the centennial of Willem
Einthoven’s first recording of the ECG in a clini-
cally applicable fashion with a string galvanome-
ter of his design. The ECG is noninvasive,
simple to record, and its cost is minimal. Possibly
no other medical invention has had greater impact
or is so universally used all over the world. De-
spite competition from many new procedures, it
has remained in continuous use for 104 years.
Electrocardiography obviously has strengths and
weaknesses, but it remains a well-established, in-
dispensable diagnostic tool. The clinical applica-
bility and importance of the 12-lead ECG
continue to grow in patients with all kinds of
heart disease, as outlined in this issue of Cardiol-
ogy Clinics. In this context it is important to know
what the ECG can do better than other diagnostic
methods.
Many relatively recent advances in electrocardi-
ography have increased its complexity, creating
a shortage of properly trained electrocardiographers.
Charles Fisch, a renowned electrocardiographer,
has repeatedly emphasized this problem, and has
reminded us that this issue dates back to the early
days of electrocardiography as indicated by Carl
Wiggers in the preface of Principles and Practice
of Electrocardiography in 1929. Wiggers stated
that ‘‘unfortunately, the training of medical man-
power in the use of such apparatus and the intel-
ligent interpretation of the electrocardiogram has
not kept pace with the increased demand. Few
courses in electrocardiography are included in un-
dergraduate and postgraduate curricula in medi-
cal schools, so that opportunity for systematic
instruction is decidedly restricted.’’ Fisch was cor-
rect in stating that the issue of manpower ad-
dressed by Wiggers 87 years ago is still with us
today. At the beginning of the twenty-first century
there is also a loss of interest in ‘‘low-tech’’ elec-
trocardiography by younger physicians and scien-
tists. For this reason, in the United States,
specialty board certification in cardiology (Amer-
ican Board of Internal Medicine) requires passing
a separate portion of the certification examination

0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.014 cardiology.theclinics.com
xvi PREFACE

in cardiology that deals only with ECG assume more and more importance in daily clini-
interpretation. cal practice.
The contributions to this issue of Cardiology
Clinics are reminiscent of Fisch’s 1980 statement S. Serge Barold, MD
that ‘‘He who maintains new knowledge in elec- University of South Florida College of Medicine
trocardiography is no longer possible or contribu- Tampa General Hospital
tive ignores history.’’ The new information in this 5806 Mariner’s Watch Drive
volume confirms that knowledge related to the Tampa, FL 33615, USA
‘‘simple’’ 12-lead ECG continues to grow and
E-mail address: ssbarold@aol.com
 Cardiol Clin 24 (2006) 305–307
The Interpretation of Electrocardiograms:
Pretense or a Well-Developed Skill?
J. Willis Hurst, MD, MACP
Department of Medicine, Division of Cardiology, Emory University School of Medicine,
1462 Clifton Road NE, Suite 301, Atlanta, GA 30322, USA
Graduating medical students can be forgiven
for being unskilled because even our most highly
ranked medical schools only introduce students to
medical skills such as history taking, physical
examination, electrocardiography, and radiogra-
phy. Stated another waydclinical skills are not
perfected in medical school. Because this is true,
such skills must be mastered during house staff
training. However, it is patently apparent that
house officers are not mastering the skills, and
teachers are not reaching them. This is tragic,
because house officer training programs were
created so that trainees could become competent
in the use of clinical skills. Because clinical skills
are still used in the practice of medicine, it is in-
deed proper to be concerned about the obvious
deterioration of house officer training.
This essay deals exclusively with the decline in
interest and clinical competence in electrocardi-
ographydthe most commonly performed proce-
dure by internists and family practitioners. The
decline is also true for the subspecialties of
internal medicine including cardiology. A few
suggestions for improving the interpretive process
are also discussed [1].
To perform a procedure and not be able to
interpret the information yielded by the procedure
is pretense. The patient believes the physician
knows how to interpret the tracings. It is not a will-
ful act of pretense on the part of the physician. As
the author sees itdmany physicians do not know
that they don’t know, and believe they can interpret
electrocardiograms. After all, they have memorized
a few patterns and a little of the language and seem
E-mail address: jhurst@emory.edu
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.03.001
to get by. Unfortunately, their memory is limited,
and they do not identify all the abnormalities in
the electrocardiogram. This, of course, leads to nu-
merous errors of omission and commission.
Why has the interpretations of electrocardiograms
deteriorated and what is responsible for the
mistaken view that electrocardiography is no
longer needed
There are those who believe that the informa-
tion yielded by an electrocardiogram has not
changed in the last few decades. They argue that
little new has been discovered, and that an
electrocardiogram is no longer needed in the
examination of the heart. They naively believe
that other technology has replaced electrocardi-
ography. I ask such skeptics if they believe we are
no longer dependent on the electrocardiogram for
the recognition of cardiac arrhythmia. They then
admit that the electrocardiogram is needed to
identify certain cardiac arrhythmias. I then point
out that the treatment of acute coronary syn-
dromes is based entirely on the abnormalities in
the electrocardiogram. Therefore, the correct in-
terpretation of the electrocardiogram is needed
more now than anytime in the history of medicine.
I ask the skeptic when did he or she discover an
epsilon [2] wave, a Brugada wave, [3] or an Osborn
wave [4] in a tracing? I ask how do you identify
primary and secondary T waves or primary and
secondary S-T segments? I may also ask them to
discuss the ventricular gradient and when do they
use it.
It usually becomes apparent that those who
downgrade the use of the electrocardiogram, but
keep recording tracings, are arguing from a poorly
cardiology.theclinics.com
306
thought out, but obviously defensive position.
The fact is, they have lost touch with the skill.
Dependence on the computer interpretation
Many of us hailed the new computer interpre-
tation as a great advance. Therefore, it is sad to
report that it has failed us [1]. We hoped it would be
accurate and would serve as a teacher to trainees.
Regrettably, it is not accurate and has failed as
a teacher, leaving those who wish to learn actually
learning very little. Then, too, the computer
software in many hospitals is 2 or more decades
old, so the new information created during the
last 2 decades is not recognized by the computer.
In addition, although most computers calcu-
late the frontal plane direction of the P, QRS, and
T vectors, they do not compute the spatial
orientation of the vectors.
Finally, and most important, the computer
readout rarely gives the physician a differential
diagnosis, indicating the probability that certain
types of heart disease are present. After all, that is
what the clinician needs.
Poor teaching
It is clear that the deterioration of the teaching
of electrocardiography is the major cause of the
poor performance of graduates of house staff
programs. Trainees do not rotate through
a 2-month period of intensive study under the
supervision of a true teacher as they did formerly.
Simultaneously, teaching ward rounds have
changed considerably. It is uncommon for a teach-
ing attending to discuss and demonstrate the sub-
tleties of the history, physical abnormalities,
radiologic abnormalities, and electrocardiographic
abnormalities that are germane to the diagnosis
and treatment of the patient. The teaching attend-
ing often simply states what high-tech procedure is
needed rather than teach electrocardiography or
any other skill. Therefore, the trainee is left with the
view that the electrocardiogram is no longer used in
any obvious way. This actually teaches the trainee
to become a subcontractor. Knowing that the
teaching attending will order an echocardiogram,
a nuclear test, and cardiac catherization, the trainee
simply orders several procedures, including an
electrocardiogram, and lets someone else make
decisions based on the results of these procedures.
Now there is less teaching time than ever. The
Residency Review Committee has decreased the
HURST
number of hours a trainee can remain in the
hospital, and this has decreased the time a true
teacher can discuss the basis of electrocardiogra-
phy, leaving the trainee to memorize a few pat-
terns. Not knowing what they mean, the trainee
quickly forgets them.
The method used in teaching
Here this author admits a strong bias. I have
tried having trainees memorize patterns of elec-
trocardiographic abnormalities, and have been
miserably disappointed. I strongly recommend
the Grant method of interpretation [5,6]. His basic
work has been updated, and those who master it
have a tool that can be used to interpret tracings
they have never seen before [7]. The use of vector
concepts leads the user to understand the cause of
abnormalities found in the tracing and permits
one to measure in degrees the magnitude of the
abnormality. Also, the pattern worshippers
should realize that the best way to learn patterns
is to become proficient in the Grant method of
interpretation of electrocardiograms.
How can skilled interpreters be identified?
The answer to this question is obviousdan
interpreter is skilled when, after viewing an elec-
trocardiogram, he or she can answer the following
question. What heart disease, or diseases, can
produce this tracing? A skilled interpreter can
either identify the cardiac diagnosis or give a
carefully thought out differential diagnosis.
Suggestions for improvement
Improvement will not come unless house staff
training programs are restructured so that trainees
can learn skills, including cognitive skills, from
a true teacher of medicine. Protected time for
teaching is necessary, just as protected time for
research is needed by research-oriented faculty.
The updated Grant methods should be taught in
specially designed seminars and implemented by
teaching attendings when they see patients with
trainees. Trainees should become knowledgeable
of the Grant method of interpretation by mid-year
of their internship, and use it on every patient they
see in whom an electrocardiogram has been
performed. They should apply the method to
each tracing they see and correlate their perception
 INTERPRETATION OF ELECTROCARDIOGRAMS
of the abnormalities with the other data collected
from the patient.
The wise men and women who determine the
content of board examinations should consider
making the current interpretation of 12 carefully
chosen electrocardiograms mandatory to pass the
examination. This is justified, because many
physicians other that cardiologists order
electrocardiograms.
Summary
There must be no pretense in medicine. When
a physician orders an electrocardiogram on a
patient he or she must be able to interpret it.
Not only that, he or she must understand the
mechanism responsible for the abnormality and
how the abnormality fits, or does not fit, with
the other data collected from the patient.
Some of the reasons the teaching of electro-
cardiography has declined have been discussed,
and a few suggestions have been offered to
improve a rather serious educational problem.
307
References
[1] Hurst JW. Current status of clinical electrocardio-
graphy with suggestions for the improvement of the
interpretive process. Am J Cardiol 2003;92:1072–9.
[2] Hurst JW. Naming of the waves in the ECG, with
a brief account of their genesis. Circulation 1998;98:
1937–42.
[3] Brugada P, Brugada J. Right bundle ranch block, per-
sistent ST segment elevation and sudden cardiac
death: a distinct clinical and electrocardiography syn-
drome. A multimember report. J Am Coll Cardiol
1992;20:1391–6.
[4] Osborn JJ. Experimental hypothermia; respiratory
and blood pH changes in relation to cardiac function.
Am J Physiol 1953;175:389–98.
[5] Grant RP. Spatial vector electrocardiography:
a method for calculating the spatial electrical vectors
of the heart from conventional leads. Circulation
1950;2:676–95.
[6] Grant RP. Clinical electrocardiography: the spatial
vector approach. New York: McGraw-Hill Book
Company; 1957. p. 1–218.
[7] Hurst JW. Interpreting electrocardiograms: using ba-
sic principles and vector concepts. New York: Marcel
Dekker; 2001. p. 119–20.
 Cardiol Clin 24 (2006) 309–315
Pitfalls and Artifacts in Electrocardiography
Barbara J. Drew, RN, PhD
Department of Physiological Nursing, University of California, University of California, San Francisco,
2 Koret Way, San Francisco, CA 94143
The ECG is a noninvasive technique that is
inexpensive, simple, and reproducible. It is one of
the most commonly used diagnostic tests that can
be recorded rapidly with extremely portable
equipment and generally is always obtainable.
The ECG contains a wealth of diagnostic in-
formation routinely used to guide clinical decision
making. For some conditions, such as transient
myocardial ischemia, the ECG remains the refer-
ence standard for diagnosis despite the advance of
many other diagnostic techniques.
ECG diagnoses are valid, however, only if
electrodes are placed in correct anatomic loca-
tions, lead wires are attached to the appropriate
electrode, and the recording is of good technical
quality (eg, proper filtering, absence of extraneous
electrical noise). In addition, if comparisons of
serial ECGs are required to determine trends over
time (eg, to determine whether ST-segment
changes of ischemia are resolving after early
reperfusion therapy), all recordings must be
made using a consistent technique. For example,
each ECG should be recorded with electrodes in
the same location, with the patient’s body in the
same (supine) position, and using the same lead
configuration.
This article reviews commonly encountered
errors in clinical electrocardiography. Errors re-
lated to inaccurate lead placement, lead wire
reversals, noisy ECG signals, inappropriate filter
settings, and serial comparisons between incon-
sistent lead sets are reviewed. In addition, recom-
mendations are made for preventing these pitfalls
and artifacts that may lead to inappropriate
E-mail address: barbara.drew@nursing.ucsf.edu
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.006
therapy and adverse patient outcomes in clinical
electrocardiography.
Inaccurate lead placement
Limb leads
When Einthoven [1] introduced the standard
limb leads nearly 100 years ago, he measured elec-
trical potentials from the lower left leg and right
and left wrists. When the arms are extended, these
three electrode sites form a triangle with points
remote and nearly equidistant from the heart.
This equilateral triangle is the basis for Eint-
hoven’s formula, which states that, at any given
instant, the amplitude of any wave in lead II is
equal to the sum of the amplitudes of the corre-
sponding wave in leads I and III. To this day,
ECG machines record leads I and II and then
mathematically derive lead III using Einthoven’s
formula.
Wilson [2] later pointed out that because all
points on a given extremity have the same poten-
tial, it does not matter whether the electrode is
connected to the limb end or the trunk end of
the extremity. In fact, Wilson referred to the limbs
as simply extension cords of the lead wires. Wil-
son went on to warn, however, that if limb elec-
trodes were placed off the limbs on to the body
torso, particularly if placed within 10 to 15 inches
of the heart, ECG waveforms would be altered
significantly [2].
These principles are important to appreciate
today in recording a standard ECG. In general,
placement of the electrodes for the limb leads (right
arm, left arm, left leg) can be anywhere on the limb
without substantially affecting ECG waveforms
and diagnostic interpretation. When limb elec-
trodes are placed on the torso, however,
cardiology.theclinics.com
310 DREW

particularly if they are moved closer to the heart
(eg, arm electrodes placed closer to the sternum
than the shoulders), the result may be clinically
important misdiagnoses in computerized and phy-
sician interpretations [3,4].
In most cases, the change in diagnosis between
a torso-positioned versus a standard ECG is either
loss or appearance of Q waves of infarction in the
inferior leads. Changes in ST-T waves (ST de-
pression, T-wave inversion or flattening) also may
occur and could be interpreted as acute myocar-
dial ischemia when serial comparisons are made
between standard and torso-positioned ECGs.
Finally, P-, R-, and S-wave voltages may be
different between standard and torso-positioned
ECGs, resulting in differences regarding presence
or absence of atrial overload or left ventricular
hypertrophy.
The practical application of these principles is
that a consistent method for placing limb elec-
trodes should be used throughout a hospital
system, especially if 12-lead ECGs are stored in
a computer system for subsequent serial compar-
ison. In some critical care and emergency hospital
units, a torso-positioned 12-lead ECG is recorded
continuously. Some manufacturers tout the ad-
vantage of 12-lead cardiac monitors as a substan-
tial financial savings because cardiac monitor
tracings can be substituted for ECGs previously
recorded with standard ECG machines by spe-
cially trained ECG technicians. Serial compari-
sons should not be made between limb-positioned
and torso-positioned ECGs, however, because
observed differences may be caused entirely by
the different lead methods.
Another potential difference in 12-lead ECGs
printed at a central nurses’ station in a monitored
hospital unit is that the patient may not be in
a supine position during the tracing. Body posi-
tion changes have been reported to be the most
common cause of false ST-segment monitor
alarms [5] and have resulted in inappropriate ther-
apy and adverse patient outcome [6].
Precordial leads
According to the principles laid out by Wilson
[2], slight misplacement of leads that are located
near the heart can produce substantial alterations
in ECG waveforms. Fig. 1 shows how valuable
QRS morphologic criteria to distinguish ventricu-
lar tachycardia from supraventricular tachycardia
with aberrant conduction disappear when the lead
V1 electrode is moved just one intercostal space
from its correct anatomic location.
Correct placement of the lead V1 electrode is
especially important because it is the first precor-
dial lead to be positioned, and the remaining five
precordial leads are placed in relation to it. There-
fore, if the V1 electrode is placed superior to its
correct location, leads V2–6 also may be misplaced
superiorly. Wenger and Kligfield [7] reported clin-
ically important differences in precordial lead
placement when serial ECGs were recorded in dif-
ferent hospital units by different personnel. These
investigators found that leads V1 and V2 were mis-
placed superiorly in 50% of the recordings, occa-
sionally reaching the second intercostal space.
Zema and colleagues [8] reported that high and
low placement of precordial electrodes produced
a change in R-wave amplitude of about 1 mm
per interspace, with smaller R waves when V1
and V2 were placed superior to the correct fourth
intercostal space. Thus, the misdiagnosis of poor
R-wave progression or anterior myocardial

Fig. 1. (A) Onset of wide QRS complex tachycardia shows an rR0 pattern in lead V1, which is unhelpful in distinguishing
between ventricular tachycardia and supraventricular tachycardia with aberrant conduction. Examination of the patient
revealed that the V1 electrode was misplaced to the fifth rather than the fourth intercostal space. (B) After lead placement
was corrected, another episode of wide QRS complex tachycardia showed an Rr0 pattern, which is strongly suggestive of
ventricular tachycardia. Subsequent invasive cardiac electrophysiologic study confirmed the patient had recurrent mono-
morphic ventricular tachycardia. (From Drew BJ, Califf RM, Funk M, et al. Practice standards for electrocardiographic
monitoring in hospital settings. Circulation 2004;110:2721–46; with permission.)
 PITFALLS AND ARTIFACTS IN ELECTROCARDIOGRAPHY 311

infarction may result from superiorly displaced or for other reasons. In such circumstances, the
precordial leads. alternative electrode sites should be clearly iden-
An important factor that complicates place- tified on the final ECG.
ment of precordial leads in women is large,
pendulous breasts, particularly in obese and older
women. In young women, the V4 location in the Inappropriate serial comparisons using different
fifth intercostal space at the midclavicular line of- lead sets
ten falls at the inferior margin of the breast where
A phenomenon occurring in current hospital
the fleshy part of the breast joins the chest wall.
settings that may make serial ECG comparisons
Thus, in young women, leads V4 and V5 can be
misleading is that patients may have cardiac
placed accurately on the chest wall below the
monitors that provide 12-lead ECG tracings that
breast. In older women, however, the nipple end
are mathematically derived from a reduced num-
of the breast sags below the breast attachment
ber of leads [11,12]. Such derived 12-lead ECGs
to the chest wall such that visualization of the
are comparable to the standard ECG for diagno-
V4 site requires lifting up the breast. In such in-
ses important in the immediate care setting, such
stances, electrodes could be placed accurately in
as distinguishing ventricular tachycardia from
the correct intercostal space either by placing elec-
supraventricular tachycardia with aberrant con-
trodes over the top of breast tissue or by lifting up
duction, detection of new bundle-branch block,
the breast and placing them under the breast
and ST-segment changes of acute myocardial is-
against the chest wall.
chemia [13–16]. For serial comparisons that re-
Research reports about which of these two
quire measurement of precise amplitudes (eg, to
options are best are conflicting. Rautaharju and
determine whether voltages indicative of left ven-
colleagues [9] recommend that precordial leads be
tricular hypertrophy have decreased or whether
placed over the top of the breast, because in so
ST-segment elevations have resolved following re-
doing they found only negligible attenuation of
perfusion therapy), however, derived and stan-
R-wave voltage. Colaco and colleagues [10], how-
dard ECG comparisons should not be made.
ever, found that women who had large breasts
Fig. 2 illustrates ST-segment amplitude differences
(bra size D or greater) had clinically significant at-
between simultaneously recorded derived and
tenuation of R waves in the anterior precordial
standard leads in a patient who had acute myocar-
leads V1–4. They also reported that if precordial
dial infarction enrolled in a prospective clinical
leads were misplaced inferiorly (ie, below pendu-
trial to compare the two lead methods in patients
lous breasts), R-wave voltage in leads V5–6 were
presenting to the emergency department with
attenuated. These investigators also reported the
chest pain [15].
prevalence of poor R-wave progression was 19%
Therefore, as in the case for torso-positioned
in women compared with 11% in men. They con-
ECGs, 12-lead ECGs recorded from reduced lead
cluded that this 8% gender difference might have
sets should not be compared serially with 12-lead
been caused in part by electrodes being placed on
ECGs recorded in the standard way [17].
top of breast tissue because there was no evidence
of prior anterior infarction in most of the women
who had poor R-wave progression. Limb lead wire reversals
Because of these conflicting research findings, There are five limb lead wire reversals that
it seems reasonable to place electrodes under the produce abnormal ECG findings. A sixth error in
breast, especially in large-breasted or obese connecting the ECG cable, reversal of the right leg
women, to minimize the likelihood of altering R- (RL) and left leg (LL) electrodes, does not alter
wave voltage. If technicians are uncomfortable the ECG because potentials recorded from the
lifting up breast tissue in women, placement over right and left legs are practically the same.
the breast in the correct anatomic location is
acceptable, however. An unacceptable method is
Right arm–left arm reversal
to place precordial leads superior or inferior to the
breast, because R-wave voltage and progression The most common error when attaching an
will be altered. ECG cable to a patient is right arm (RA)–left arm
Occasionally, inaccurate precordial lead place- (LA) lead wire reversal, which means that lead I
ment is necessary in hospitalized patients because looks upside down (inverted P, QRS, and T
of chest wounds, placement of defibrillator pads, waves). If the P wave is upright and only the
312 DREW

Fig. 2. Simultaneous recordings of standard leads (top) and the EASI system that derives a 12-lead ECG using just 5
electrodes (Philips Co., Andover, MA) (bottom). Although both ECGs indicate acute inferior myocardial infarction,
ST-segment amplitudes differ considerably between the two methods. Thus, if the top tracing had been recorded in
the emergency department with a standard ECG machine before fibrinolytic therapy, and the bottom tracing had
been recorded minutes later in the coronary care unit with a bedside cardiac monitor, serial comparison between the
two lead methods would have suggested erroneously that ST-segment deviation was getting worse. (From Philips Com-
pany, Andover, MA; with permission [EASI recording].)

QRS is negative, the problem is not lead wire
reversal but an abnormal QRS axis, most likely
resulting from some pathology causing extreme
right axis deviation. If, however, the P wave is
inverted also, there are two possible causes: (1)
RA-LA lead wire reversal, or (2) dextrocardia. It
is a simple matter to distinguish between these two
conditions by examining a set of leads that do not
depend upon proper placement of arm electrodes
(ie, the precordial leads). In RA-LA lead wire
reversal, the precordial leads will look normal
with a small R wave in V1 that progressively in-
creases in consecutive V leads until it is maximally
positive in V5 or V6. In contrast, an individual
born with the heart positioned on the right side
in mirror-image fashion (sinus inversus) will
have no R-wave progression, with a negative
QRS complex in lead V6 (Fig. 3).
 PITFALLS AND ARTIFACTS IN ELECTROCARDIOGRAPHY 313

Fig. 3. Method to distinguish RA-LA lead wire reversal from dextrocardia. When there is a negative P and QRS com-
plex in lead I, examine the precordial leads to determine whether they have normal R-wave progression (RA-LA reversal
does not affect precordial leads) or abnormal R-wave progression with a negative QRS in V6 (dextrocardia affects the
precordial leads because the heart is not on the left side).

Left arm–left leg reversal
LA-LL reversal means that lead I on the ECG
is actually lead II, lead II is actually lead I, and
lead III is upside down because the positive and
negative poles are reversed. In addition, leads aVL
and aVF are reversed. This type of reversal may
be difficult to identify because it may not appear
out of the ordinary except for left axis deviation,
which is common in hospitalized patients. One
clue is the appearance of a P wave in ‘‘lead I’’ that
is larger in amplitude than in ‘‘lead II.’’ The ‘‘lead
I’’ P wave actually represents the P wave of the
true lead II, which typically has the largest P-wave
amplitude of any limb lead.
Right arm–left leg reversal
RA-LL reversal causes lead I to be the inverse
of lead II, lead II to be the inverse of lead I, and
lead II is upside down because the positive and
negative poles are reversed. In addition, aVR
and aVF are reversed. This situation produces
highly abnormal-looking limb leads, with leads I,
II, III, and aVF being negative and aVR being
upright. During sinus rhythm, it is highly unlikely
to have a QRS axis in the bizarre quadrant of
90
to  180 , which is typical of this type of lead
reversal.
Reversals involving the right leg and right
or left arm
In reversals involving one of the arm lead wires
switched with the RL lead wire, one lead (either
lead II in the case of RA-RL reversal or lead III in
the case of LA-RL reversal) records a zero
potential difference between the legs, resulting in
a flat line (termed a ‘‘far-field’’ signal).
Technically unacceptable ECG recordings
Excessively noisy signals
Hospitals are filled with equipment that can be
a source of electrical artifact during the recording
of an ECG. Moreover, ECGs recorded in the
emergency department or other immediate care
setting where patients may be restless or confused
may be plagued with a noisy signal. Fig. 4 shows
an example of a noisy signal that simulates ven-
tricular tachycardia. Knight and colleagues [18]
reported on the clinical implications of the mis-
diagnosis of artifact as ventricular tachycardia
in 12 patients. Clinical consequences of the
misdiagnoses in these patients included unneces-
sary cardiac catheterization in three patients,
unnecessary medical therapies including intrave-
nous antiarrhythmic agents in nine patients,
314
Fig. 4. False cardiac monitor alarm showing a noisy signal simulating ventricular tachycardia. Arrows point out the
normal QRS complexes that can be discerned if one looks carefully and uses calipers to confirm their timing.
precordial thumps in two patients, implantation
of a permanent pacemaker in one patient, and in-
sertion of an implantable cardioverter-defibrilla-
tor in one patient. Moreover, hospital costs were
high because of unnecessary testing, and patients
were transferred to a higher level of intensive
care or treated longer than necessary in the
hospital.
Inappropriate filter settings
The problem of unacceptably noisy ECG re-
cordings makes it tempting for clinicians to re-
program high-frequency filter settings to a lower
number on ECG machines, especially in emer-
gency departments and other immediate care
settings. Clinical guidelines for recording a stan-
dard 12-lead ECG specify a high-frequency filter
setting of no lower than 100 Hz [19]. Lowering the
high-frequency filter from 100 to 40 Hz will elim-
inate noise caused by 60-cycle interference and
other artifact. Clinically important high-frequency
signals (eg, pacemaker stimuli or notches in the
QRS complex) will be eliminated also, however.
The low-frequency filter should be set no
higher than 0.05 Hz to avoid distortion of the
ST segment [19]. Clinicians may violate this rule
to minimize baseline wander and other artifacts,
however.
In patients who have severe pain, tremor, or
some other cause of an unacceptably noisy ECG
signal, the filter switch on the ECG machine can
be used after all attempts to eliminate the in-
terference have failed. The change in filter setting
should be documented on the final ECG.
Summary/clinical implications
The value of the ECG depends upon the
accuracy of how it is obtained. Currently, many
different types of health personnel record ECGs,
including physician office workers, minimally
trained hospital aides, ECG technicians, nurses
from a variety of hospital units, and house staff.
Thus, the potential exists for inconsistent
DREW
recording methods that are especially problematic
when accurate diagnosis depends on comparison
of serial ECG tracings. It is important to have
a hospital/clinic policy for the recording of ECGs
that includes appropriate filter settings, diagrams
of accurate lead placement, and most commonly
encountered errors. Training of personnel should
include a return demonstration to identify diffi-
culties identifying anatomic landmarks and to
clear up any other inconsistent techniques or
misconceptions. Finally, a quality improvement
program should monitor the incidence of common
errors such as RA-LA lead wire reversal and
provide retraining when indicated.
References
[1] Einthoven W. The different forms of the human elec-
trocardiogram and their signification. Lancet 1912;
1:853.
[2] Wilson FN. The distribution of the potential differ-
ences produced by the heart beat within the body
and at its surface. Am Heart J 1929–1930;5:599–616.
[3] Gamble P, McManus H, Jensen D, et al. A compar-
ison of the standard 12-lead electrocardiogram to ex-
ercise electrode placements. Chest 1984;85(5):
616–22.
[4] Bennett FT, Bennett KR, Markov AK. Einthoven’s
triangle: lead errors and an algorithm for solution.
Am J Med Sci 2005;329(2):71–7.
[5] Drew BJ, Wung SF, Adams MG, et al. Bedside diag-
nosis of myocardial ischemia with ST-segment mon-
itoring technology: measurement issues for real-time
clinical decision-making and trial designs. J Electro-
cardiol 1998;30:157–65.
[6] Drew BJ, Adams MG. Clinical consequences of ST-
segment changes caused by body position mimicking
transient myocardial ischemia: hazards of ST-
segment monitoring? J Electrocardiol 2001;34:261–4.
[7] Wenger W, Kligfield P. Variability of precordial
electrode placement during routine electrocardiog-
raphy. J Electrocardiol 1996;29:179–84.
[8] Zema MJ, Luminais SK, Chiaramida S, et al. Elec-
trocardiographic poor R wave progression III: the
normal variant. J Electrocardiol 1980;13:135–42.
[9] Rautaharju PM, Park L, Rautaharju FS, et al.
A standardized procedure for locating and
 PITFALLS AND ARTIFACTS IN ELECTROCARDIOGRAPHY
documenting ECG chest electrode positions. J Elec-
trocardiol 1998;31(1):17–29.
[10] Colaco R, Reay P, Beckett C, et al. False posi-
tive ECG reports of anterior myocardial infarc-
tion in women. J Electrocardiol 2000;33(Suppl):
239–44.
[11] Dower GE, Yakush A, Nazzal SB, et al. Deriving the
12-lead electrocardiogram from four (EASI) elec-
trodes. J Electrocardiol 1988;21:S182–7.
[12] Nelwan SP, Kors JA, Meij SH, et al. Reconstruction
of the 12-lead electrocardiogram from reduced lead
sets. J Electrocardiol 2004;37:11–8.
[13] Drew BJ, Scheinman MM, Evans GT. Comparison
of a vectorcardiographically derived 12-lead electro-
cardiogram with the conventional electrocardio-
gram during wide QRS complex tachycardia, and
its potential application for continuous bedside
monitoring. Am J Cardiol 1992;69:612–8.
[14] Drew BJ, Adams MG, Pelter MM, et al. ST segment
monitoring with a derived 12-lead electrocardio-
gram is superior to routine CCU monitoring. Am J
Crit Care 1996;5:198–206.
315
[15] Drew BJ, Pelter MM, Wung SF, et al. Accuracy of
the EASI 12-lead electrocardiogram compared to
the standard 12-lead electrocardiogram for diagnos-
ing multiple cardiac abnormalities. J Electrocardiol
1999;32(Suppl):38–47.
[16] Drew BJ, Pelter MM, Brodnick DE, et al. Compar-
ison of a new reduced lead set ECG with the stan-
dard ECG for diagnosing cardiac arrhythmias and
myocardial ischemia. J Electrocardiol 2002;
35(Suppl):13–21.
[17] Drew BJ, Califf RM, Funk M, et al. Practice stan-
dards for electrocardiographic monitoring in hospi-
tal settings. Circulation 2004;110:2721–46.
[18] Knight BP, Pelosi F, Michaud GF, et al. Clinical
consequences of electrocardiographic artifact mim-
icking ventricular tachycardia. N Engl J Med 1999;
341:1270–4.
[19] Society for Cardiological Science and Technology.
Clinical guidelines by consensus, number 1: record-
ing a standard 12-lead electrocardiogram. London,
England: Society for Cardiological Science and
Technology; 2005.
 Cardiol Clin 24 (2006) 317–330
How many ECG leads do we need?
Elin Trägårdh, MD, PhD candidate*,
Henrik Engblom, MD, PhD candidate, Olle Pahlm, MD, PhD
Department of Clinical Physiology, Lund University Hospital, SE-221 85 Lund, Sweden
The number of leads used and the electrode
placement in a standard ECG have remained the
same for over half a century although the technol-
ogy behind the method has developed greatly
(Fig. 1). It is well known that the sensitivities of cur-
rent standard 12-lead ECG criteria for detecting
many cardiac diseasesdsuch as acute myocardial
infarction (MI) and ventricular hypertrophydare
poor. To increase the diagnostic ability of the
ECG, many alternative lead systems have been
proposed, but none has yet received general accep-
tance in the cardiology community.
The issue of the optimal number of leads is of
great interest to researchers, as becomes apparent
when researching the literature on the subject. Some
investigators argue that a small number of leads will
suffice; others maintain that the greater the number
of leads, the better. This article aims to answer the
question of how many ECG leads actually are
necessary by presenting some of the most widely
studied lead sets. Also discussed are researchers’
efforts to enhance the sensitivity and specificity
of the ECG in detecting acute MI and the capability
of the ECG to diagnose other cardiac conditions.
The development of the standard 12-lead ECG
The first surface ECG recorded on humans was
performed in 1887 by Waller [1], but the leads that
were first used for diagnostic electrocardiography
were three-limb leads introduced by Einthoven [2].
In the earliest ECG recording systems, the pa-
tient’s arms were inserted into jars of conducting
solution, which were connected to a sensitive
galvanometer. This method made it possible to
measure the potential differences between the
* Corresponding author.
E-mail address: elin.tragardh@med.lu.se (E. Trägårdh).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.005
two arms (lead I), the left leg and the right arm
(lead II), and the left leg and the left arm (lead III).
In the early 1930s, Wilson [3,4] introduced
a ‘‘central terminal,’’ which he hypothesized
would allow measurements of the potential varia-
tion at a single point, giving rise to the incorrect
term ‘‘unipolar leads.’’ Indeed, all leads measure
the potential difference between two poles. The
original Einthoven leads were ‘‘two-electrode
leads,’’ whereas other standard leads used a single
electrode as one pole and averaged inputs from
multiple electrodes as the second pole. When in-
puts from all three limb electrodes were averaged,
the resulting ‘‘central terminal’’ served as the sec-
ond pole, thus creating a ‘‘V’’ lead. The lead sys-
tem defined six additional chest leads to be
recorded, labeled V1, V2, V3, V4, V5, and V6.
In the early years of ECG recordings, there was
no standardization of placement of the precordial
leads, making it difficult to compare studies. In
1938, the first standardization was made by the
Cardiac Society of Great Britain and Ireland in
conjunction with the American Heart Association
[5]. Later that same year, the electrode positions
of leads V1 to V6 as known today were described
[6,7]. The standardized placement of the precor-
dial leads was a committee decision, made with
the specific intent of standardizing research.
The Wilson central terminal also enabled three
other limb leads: VR, VL, and VF. These leads,
however, were of low voltage in general and were
replaced by augmented (aV) leads by removing
input from the electrode that provided the first
pole from the averaged second pole. Augmented
limb leads, developed in 1942, measure the poten-
tial differences between the left arm and the
average of the potentials at the right arm and
left leg (lead aVL). In a similar fashion, it was
possible to design two other augmented limb
cardiology.theclinics.com
318
Fig. 1. The placement of precordial and limb electrodes. In the frontal view (left) the electrode placement for the standard 12-
lead ECG is shown: V1 to V6. The right-sided electrodes are also shown, which are mirror images of leads V2 to V6. On the
back view (right) the posterior electrodes V7 to V9 and V7R to V9R are shown. LA, left arm; LL, left leg; RA, right arm.
leads, aVR and aVF [8], creating the standard 12-
lead ECG used today.
In some situations it is not feasible to use the
standard electrode placement for limb leads be-
cause of noise from skeletal muscle (eg, during
percutaneous transluminal coronary angioplasty
and during stress tests). In 1966, Mason and Likar
[9] introduced a system in which limb electrodes
are instead placed on the torso (Fig. 2). In this sys-
tem, the arm electrodes are placed in the left or
right infraclavicular fossa, medial to the border
of the deltoid muscle and 2 cm below the lower
border of the clavicle. The left leg electrode is
placed in the anterior axillary line, halfway be-
tween the costal margin and the crest of the ilium.
Studies have shown, however, that the electrical
axis may change and that ECG findings indicating
inferior or posterior infarcts are lost in 69% and
31% of patients, respectively, compared with find-
ings in ECGs recorded with standard electrode
placement [10,11]. Hence, when interpreting an
ECG, it is important to know if it has been re-
corded using Mason-Likar electrode placement.
Additional precordial leads
It is well known that the sensitivity of current
12-lead ECG criteria is poor for detecting acute MI
TRÄGÅRDH et al
[12,13]. There are three principal reasons. First, the
infarct may be too small to affect the cardiac elec-
trophysiology reflected by an ECG. Second, the lo-
cation of the infarction might not be ‘‘observed’’ by
a 12-lead ECG because of its limited coverage of the
precordium. It also is possible that the precordial
electrodes in a standard 12-lead ECG are not opti-
mally placed to detect acute MI and ischemia. For
example, Saetre and colleagues [14] found that
leads derived from electrodes placed in the left axilla
and on the back could differentiate ECG changes
that occurred during percutaneous coronary inter-
vention of the first diagonal and left circumflex
(LCX) arteries, respectively.
Third, the location of ischemia in acute in-
farction might be observed by the 12-lead ECG as
ST depression when, in fact, it should be consid-
ered ‘‘ST elevation equivalent.’’ This consideration
of the inverse of a standard lead as the equivalent
of an additional lead (eg, lead aVR and lead
aVR)
may negate the need to create additional leads by
placement of additional electrodes [15].
In addition to precordial leads V1 to V6, other
precordial torso leads can be used as well. Posterior
leads V7 (left posterior axillary line), V8 (left
midscapular line), and V9 (left border of the spine)
are all located in the same horizontal plane as lead
 HOW MANY ECG LEADS DO WE NEED?
Fig. 2. Mason-Likar electrode placement. In compari-
son with the electrode placement for the standard
12-lead ECG, the limb leads are moved proximally when
recording the 12-lead ECG according to the Mason-Likar
system. The upper limb leads are placed in the infracla-
vicular fossa medial to the border of the deltoid muscle
and 2 cm below the lower border of the clavicle. The lower
limb lead is moved to the anterior axillary line, halfway
between the costal margin and the crest of the ilium.
LA, left arm; LL, left leg; RA, right arm.
V6. Leads V3R to V9R are located on the right
side of the chest in the same location as corre-
sponding to the location of the left-sided leads V3
to V9. V2R is therefore the same as V1 (Fig. 1).
Posterior leads
Posterior MI occurs when either the LCX or
a branch of the right coronary artery (RCA)
supplying blood to the posterior wall of the left
ventricle is occluded. Because there are no poste-
rior electrodes in a standard 12-lead ECG, the
diagnosis of posterior MI depends on ST de-
pression in anterior leads [16]. Currently, patients
who do not have ST-segment elevation in a stan-
dard ECG are not recommended for thrombolytic
therapy [17]. Moreover, there are reports of ST el-
evation in posterior leads in the absence of changes
in anterior leads [18]. It is important to diagnose
posterior MI because it often is associated with
complications such as mitral valve regurgitation
[19] and poor patient prognosis [20,21]. Accord-
ingly, there has been much interest in the study of
posterior chest leads in the past few years.
In one such study, ECG patterns in posterior
chest leads (V7–V9) were studied in 225 normal
male subjects about to undergo military training
319
[22]. The investigators found that none of the
subjects had ST elevation at the J point exceeding
0.5 mm. The prevalence of 0.5- to 1.0-mm ST ele-
vation at 80 milliseconds after the J point in lead
V7 was 8.9%, 5.8% and 3.1% in leads V7, V8,
and V9, respectively. In only two subjects did
the ST elevation reach 1 mm in amplitude. The re-
sults indicate that ST elevation in leads V7 to V9
at the J point or of more than 1 mm 80 millisec-
onds after the J point can be considered abnormal.
In acute MI, it is important to open the
occluded artery as soon as possible after the onset
of symptoms. It is desirable to identify the culprit
artery by surface ECG to facilitate coronary
intervention. A prospective ECG analysis during
angioplasty for single-vessel disease involving
LCX or RCA found that the most common
ECG change during RCA occlusion was ST-
segment elevation in leads II, aVF, and III
(95%), and that the most common change during
LCX occlusion was posterior ST elevation in leads
V7 to V9 (68%) [23]. Therefore, the use of poste-
rior leads may help in distinguishing between oc-
clusion of the RCA and occlusion of the LCX.
Many studies have demonstrated that posterior
leads are useful for identifying acute posterior MI
not detected by the standard ECG [19,21,24–26].
Zalenski and colleagues [26] found that the addi-
tion of V4R, V8, and V9 in patients admitted to
a cardiac-monitored unit with a provisional diag-
nosis of MI or unstable angina increased the sensi-
tivity of ST-segment elevation detection from
47.1% to 58.8%, with no loss of specificity. Bio-
chemical markers were used to diagnose MI, and
percutaneous coronary intervention was not per-
formed in the study.
In another study, however, Zalenski and col-
leagues [27] found that the accuracy of detecting
ST-segment elevation in acute MI was improved
only modestly by the addition of leads V7 to V9.
The sensitivity increased from 57.7% to 59.7%,
whereas the specificity decreased from 91% to
89.4%.
In a study of low-risk patients presenting with
chest pain suggestive of acute coronary syndromes,
posterior leads were not found to increase the
detection rate of ischemia [28]. Low-risk patients
were identified by a normal 12-lead ECG, absence
of arrhythmias or hemodynamic instability, and
one negative serum cardiac troponin I assay. Dif-
ferent studies have used different levels of ST eleva-
tion in posterior leads as diagnostic of acute MI.
Some authors have used a criterion of 1 mm,
whereas studies with an angiography gold
320 TRÄGÅRDH
standard have considered 0.5 mm of ST elevation
to be diagnostic [29].
Right-sided leads
Right ventricular infarction rarely occurs in
isolation [30] but occur in more than 30% of cases
of inferior left ventricular MI [31]. Patients who
have right ventricular infarct involvement have
a worse prognosis [32,33], a significantly higher in-
cidence of depressed right ventricular function
[34], and a high risk of developing high-degree
atrioventricular nodal block compared to infarcts
with no right ventricular involvement [35]. Usu-
ally, there are only inconclusive signs of right ven-
tricular infarction in the standard 12-lead ECG.
Chou and colleagues [36] concluded that ST eleva-
tion of 1 mm or more in lead V1, in the absence of
other explanations for the ST elevation, should
raise suspicion of right ventricular involvement.
Other studies [37–39], however, do not support
this criterion because of its low sensitivity. One
method that can be used to observe the right ven-
tricle better is to apply right-sided leads. Andersen
and colleagues [40] described ST deviations in the
right chest leads V3R to V7R during transient bal-
loon occlusion of the coronary arteries. They
found that ST elevation always evolved in leads
V5R and V6R when the RCA was occluded and
was 100% discriminative for RCA occlusion com-
pared with occlusion of either of the two left cor-
onary arteries. ST elevation within normal range
or ST depression could be seen in leads V3R
and V4R during RCA occlusion.
Morgera and colleagues [41] studied 82 normal
subjects and found that an rS pattern was always
present in V3R and was present in V4R in 91% of
subjects. qR or qS patterns were seen in V6R in
26% of the subjects, in V5R in 10% of subjects,
and in V4R in 2% of subjects. In a study by An-
dersen and colleagues [42], 109 normal subjects
were studied. An rS configuration was found in
V3R in 98% and in V4R in 91%. A qS configura-
tion was found most frequently in V6R (16%).
They also investigated the ST deviation 40 and
80 milliseconds after the J point. Measurements
of the ST segment made 80 milliseconds after
the last QRS deflection showed significantly
more deviation from the isoelectric line than mea-
surements made 40 milliseconds after the last
QRS deflection. They concluded that ST-segment
deviation should be measured 40 milliseconds af-
ter the last QRS deflection and should be at least
0.6 mm to be considered abnormal.
et al
Many studies have evaluated right-sided pre-
cordial leads in acute MI and have found a high but
variable sensitivity for detecting right ventricular
MI [27,37–39,41,43]. Croft and colleagues [37]
studied 33 patients who had enzymatically con-
firmed MI, using myocardial scintigraphy as the
criterion gold standard. They found that ST eleva-
tion of 1 mm or greater in one or more of leads V4R
to V6R is both highly sensitive (90%) and specific
(91%) in identifying acute right ventricular infarc-
tion. Braat and colleagues [38] reported similar re-
sults, with a sensitivity of 93% and a specificity of
95% in lead V4R, which was found to be the best
right-sided lead for detecting right ventricular MI.
Later, they also performed a study using angiogra-
phy as the criterion gold standard [43]. Forty-two
patients who had inferior MI were included, 17 of
whom had right ventricular MI. The sensitivity
for detecting right ventricular MI was 100% for
lead V4R, and the specificity was 87%. Lopez-Sen-
don and colleagues [39] found 100% sensitivity and
68% specificity in detecting right ventricular MI in
lead V4R.
Another study compared the standard 12-lead
ECG with 12-lead ECG plus the addition of V4R
to V6R in 533 patients who had chest pain [27].
They found that the sensitivity increased from
57.7% to 64.4% with the addition of V4R to
V6R. The specificity, however, decreased from
91% to 85.6%. Most of the studies used 1 mm
as the cut-off value [27,37,38,41,43], but others
have used 0.5 mm [39,41]. An autopsy study by
Morgera and colleagues [41] compared sensitivity
and specificity when using a 1-mm or 0.5-mm cut-
off in lead V4R. With 0.5 mm as the cut-off, the
sensitivity was 76%, and the specificity was
86%. With 1 mm as the cut-off, the sensitivity
was 57%, and the specificity was 100%.
The use of right-sided precordial leads has also
been investigated during exercise tests, with di-
vergent results. Some studies have shown that the
addition of right-sided leads does not increase the
sensitivity of the ECG for the detection of
myocardial ischemia [44,45], but another study
showed that the sensitivity improved greatly [46].
Some studies indicate that the addition of right-
sided chest leads could be of use in the early diag-
nosis of acute pulmonary embolism [47,48].
Body surface potential mapping
The standard 12-lead ECG is derived from
only 10 electrodes. Hence, it offers a limited
 HOW MANY ECG LEADS DO WE NEED?
coverage of the body surface potential distribu-
tion caused by the depolarization and repolariza-
tion of the myocardium. Therefore, local electrical
events in the myocardium may not be reflected in
the standard 12-lead ECG. Body surface potential
mapping (BSPM) is a method that provides better
coverage of the body surface potential distribu-
tion by performing recordings at multiple sites
(24–240) on the body surface. This approach is
not novel; in fact, the first example of a potential
map was published in 1889 by Waller [49], who
performed 10 to 20 ECG recordings at various
points on the surface of the human body. Since
then, normal BSPM patterns have been estab-
lished [50,51], and abnormal patterns of BSPM
have been explored. In addition to this empiric
use of BSPM, aimed at obtaining descriptive in-
formation of cardiac electrical behavior, BSPM
has evolved from the so-called ‘‘classical forward’’
and ‘‘inverse’’ problems of electrocardiography.
The former focuses on how the electrical activity
from electrical sources present in the heart is
propagated to the epicardium (near-field problem)
or to the body surface (far-field problem) [52]. The
latter focuses on how the electrical activity re-
corded at the body surface can be used to derive
details about the electrical activity in the heart
(ie, epicardial potentials) [53]. Both problems re-
quire information of high spatial resolution from
the body surface, which can be obtained through
BSPM. The theoretical aspects of BSPM are not
discussed further in this article. Instead, the
more clinical aspects of BSPM are discussed.
Clinical implications
Several studies involving BSPM in patients
with MI have been published [54–74]. A study
by Kornreich and colleagues [55] included 177 pa-
tients who had MI and 184 normal subjects. Cri-
teria based on six features from three locations
(mostly ST-T measurements) derived from 120-
lead BSPM data yielded a specificity of 95% and
sensitivity of 95% for diagnosing MI. The three
locations were the right subclavian area, the left
posterior axillary region, and the left leg. The
12-lead ECG had the same specificity but a sensi-
tivity of 88% for the same number of features.
The leads with the greatest discriminating power
were found outside the 12-lead ECG. Further-
more, it has been shown that the sensitivity for
right ventricular MI increased from 42% for right
ventricular leads to 58% for BSPM [57]. In the
321
same study, the sensitivity increased from 1%
for posterior leads to 27% for BSPM.
The 12-lead ECG has a sensitivity of about
50% for detection of acute MI. Menown and
colleagues [56] studied 635 patients who had chest
pain, of whom 325 had MI (according to the
World Health Organization criteria for acute
MI), and 125 controls. Subjects were randomly
assigned to either a learning or a test set. The
64-lead anterior BSPM data showed an overall
correct acute MI classification of 74%.
The standard 12-lead ECG is known to have
low sensitivity for detection of acute MI in the
presence of left bundle branch block. Sgarbossa
and colleagues [75] showed a sensitivity of 33%,
and Hands and colleagues [76] showed a sensitivity
of 17% for detecting acute MI in the presence of
left bundle branch block. Maynard and colleagues
[58], however, have shown that the sensitivity can
be increased to 67% using BSPM.
BSPM also has been shown useful in evaluat-
ing the clinical efficacy of reperfusion therapy
[77,78]. In a study of 67 patients who had acute
MI undergoing coronary angiography 90 minutes
after fibrinolytic therapy, Menown and colleagues
[77] showed that BSPM had a sensitivity of 97%
and a specificity of 100% for detecting patients
who had established reperfusion (Thrombolysis
in Myocardial Infarction [TIMI] grade 2–3
flow). In comparison, all ST-resolution patterns
on the standard 12-lead ECG showed a sensitivity
of 59% and a specificity of 50%.
The 12-lead ECG is used frequently for assess-
ing stress-induced ischemia in coronary artery
disease. Using BSPM, Manninen and colleagues
[79] showed that regional stress-induced ischemia
can be identified at sites on the body surface not
covered by the standard 12-lead ECG. Montague
and colleagues [80] demonstrated that ischemic re-
polarization changes were detectable and quantifi-
able by BSPM at low levels of cardiac stress in
patients who had one-vessel disease when the
usual ECG criteria of myocardial ischemia fre-
quently were absent. Furthermore, Boudik and
colleagues [81] showed that BSPM during dipyri-
damole stress can be used to differentiate between
normal subjects, patients who have syndrome X,
and patients who have coronary artery disease.
The sensitivity and specificity for detecting syn-
drome X were 71% and 78%, respectively.
BSPM also has been shown to increase the
diagnostic accuracy of left ventricular hypertro-
phy criteria [82] as well as the ability to estimate
left ventricular mass from an ECG [83,84] in
322 TRÄGÅRDH et al

comparison to a standard 12-lead ECG. In addi- problems might occur even when using only the
tion, BSPM has been shown to be useful in iden- 10 electrodes of the standard 12-lead ECG.
tifying accessory pathways in patients who have Multiple electrodes and wires interfere with aus-
Wolff-Parkinson-White syndrome [85–91]. cultation, echocardiograms, resuscitation efforts,
and chest radiographs. Noise levels created by
Why is body surface potential mapping limb movement detected by multiple leads make
not clinical routine? interpretation difficult, and the discomfort to
patients caused by so many electrodes tends to
It is clear that BSPM provides more informa-
be high. Additionally, rapid and accurate elec-
tion regarding the cardiac electrical field than the
trode placement can be difficult in emergency
standard 12-lead ECG. Why then is BSPM not
situations. Fewer electrodes placed at more easily
more clinically established? First, recording the
accessed locations could facilitate ECG acquisi-
large number of leads and displaying them ap-
tion for both patient and staff.
propriately is time consuming, making the pro-
cedure unsuitable for clinical settings. This
obstacle has, to a large extent, been overcome Frank leads
by the development of electrode strips that are
Today, the standard 12-lead ECG is the most
easier to apply and by modern computer technol-
common method for studying the electrophysiol-
ogy. Modern BSPM systems can acquire and
ogy of the heart. Vectorcardiography is another
process maps in less than 30 seconds. Second,
method that was first described in the early
the great amount of information received from
twentieth century. A vectorcardiogram traces the
BSPM has been difficult to interpret and classify
sum of the heart’s electrical activity in a three-
correctly; however, a variety of statistical methods
dimensional space throughout the cardiac cycle.
have been devised that allow measurements of
Three orthogonal leads, Vx, Vy, and Vz, are used.
cardiac patients to be compared with those of
The most commonly used and best-studied or-
healthy subjects of various ages, sex, and body
thogonal lead system is that of Frank [93]. In this
habitus. Interpretation schemes are now available,
lead system, seven electrodes are used (Fig. 3, up-
also. Nonetheless, correct interpretation and clas-
per panel). Five of the electrodes are positioned
sification of maps remain challenging tasks and
around the heart on the torso (on the left and
require a deep understanding of body surface
right side of the thorax, on the sternum, on the
potential distribution. In a recent study of 389
back, and on the left side of the chest, at an angle
patients presenting to an emergency department
of 45 with respect to the center of the thorax),
with chest pain, Carley and colleagues [92] showed
one electrode is positioned on the back of the
that BSPM offered a relatively small increase in
neck, and one on the left foot. From the contrib-
sensitivity (from 40% to 47.1%), and a compara-
uting measurements of these leads, the orthogonal
ble decrease in specificity (from 93.7% versus
leads Vx, Vy, and Vz are derived. Typically, the
85.6%) for detecting acute MI in comparison to
leads are analyzed in pairs and presented as loops.
a standard 12-lead ECG. In this study, the
The amplitudes of the two leads are measured at
BSPM maps were interpreted by emergency physi-
the same time and plotted in a diagram, with the
cians trained in BSPM, not by cardiologists with
amplitude of one lead appearing on the x-axis
a special interest in BSPM. Additionally, this
and the amplitude of the other lead appearing
study was performed in an emergency setting
on the y-axis (Fig. 4). The leads are combined
with unselected patients, thus reflecting the clini-
pair-wise to a frontal plane (x and y), a transver-
cal reality of a typical emergency department.
sale plane (x and z), and a sagittal plane (y and z).
Although BSPM offers better coverage of the
The shape, direction of rotation, and area of the
body surface potential distribution, further studies
loop are then analyzed.
of the diagnostic capability of BSPM in different
Continuous vectorcardiographic registration
clinical settings are needed to establish it as
with Frank leads has been shown to have sub-
a clinical routine examination.
stantial potential for monitoring patients who
have acute MI [94,95] and may help in identifying
candidates for emergency coronary angiography
Reduced lead sets
[95]. It also has been shown to correlate more
Although there are clear advantages in using closely with enzymatically estimated infarct size
additional leads in certain circumstances, in patients who have Q-wave infarction than
 HOW MANY ECG LEADS DO WE NEED? 323

Fig. 3. Comparison between Frank and EASI electrode placement. The E, A, and I electrodes are identical for the two
placements. In the EASI setting, however, no electrodes are required on the back.

does QRS scoring of the standard 12-lead ECG
[96]. The Common Standards for Quantitative
Electrocardiography working party, however,
concludes that vectorcardiography has slightly
lower specificity and although a slightly higher
sensitivity than standard 12-lead ECG [97].
Derived 12-lead ECG
In 1988 Dower [98] introduced a ‘‘derived’’ 12-
lead ECG, using five torso electrodes, four record-
ing and one ground. Because three electrodes
from the original Frank lead configuration
(E, A, and I) were used, he labeled it the ‘‘EASI
12-lead ECG’’ (Fig. 3, lower panel). In addition,
he used a fourth electrode (S). The fifth ground
electrode can be placed anywhere. The electrodes
are placed on easily identified anatomic points
on the torso: the E electrode on the lower sternum,
the A and I electrodes on the left and right mid-
axillary lines, respectively, at the same transverse
level as E, and the S electrode on the sternum
manubrium. The electrodes record three ECG
leads: A-I (horizontal vector), E-S (vertical vector),
and A-S (anterior-posterior vector). Each of the 12
standard ECG leads is derived as a weighted linear
324 TRÄGÅRDH
Fig. 4. A vectorcardiogram from a patient who had a large anterior myocardial infarction. To the left the three orthog-
onal leads Vx, Vy, and Vz are shown. From these leads the QRS loops shown to the right in the frontal, horizontal, and
left sagittal planes were derived. As seen in the horizontal and left sagittal planes, the anterior electrical forces are com-
pletely absent as a result of the large anterior infarction.
sum of the signals from these three leads using the
following formula:
Lderived ¼ aðA
IÞ þ bðE
SÞ þ cðA
SÞ;
where L represents any ECG lead and a, b, and c
represent coefficients determined to optimize the
fit between the standard 12-lead ECG and the
derived ECG [99].
Basic comparisons between standard 12-lead
ECG and EASI-derived 12-lead ECG have con-
cluded that differences between PR, QRS, QT,
and QTc intervals are small. Some differences in
cardiac axes, especially in T-wave axes, have been
found, but EASI still accurately detected right/left
bundle branch block and fascicular blocks
[100,101].
Accurate diagnosis of both myocardial ar-
rhythmias and ischemia often requires analysis
of multiple leads. Many hospital wards, however,
rely on only two leads (most often leads II and
V1) for ECG monitoring. Continuous monitoring
of the EASI-lead ECG has been found to be
superior to monitoring with leads II and V1 for
diagnosing transient myocardial ischemia in pa-
tients who have unstable coronary syndromes
[102]. Leads II and V1 were shown to miss 64%
of ischemic ST changes detected by EASI; 75%
et al
of the missed events were clinically silent. More-
over, EASI-derived ECG has been found to be di-
agnostically equivalent to standard 12-lead ECG
for the detection of cardiac arrhythmias [100,101].
The diagnostic capabilities of the EASI system
have been found to be equivalent to the standard
12-lead ECG for the detection of acute ischemia
[101–104], acute MI [105], and prior MI
[100,101,103,106]. Another application might be
in the ambulance setting. Sejersten and colleagues
[107] showed that the EASI lead system can pro-
vide an alternative to the standard 12-lead ECG
to facilitate data acquisition and possibly save
time in emergency situations. So far, few studies
have been performed in children, where it also is
important to simplify ECG recordings. A basic
study by Pahlm and colleagues [108] concluded
that EASI leads in children have the same high
levels of ‘‘goodness-of-fit’’ to replicate conven-
tional 12-lead ECG waveforms as reported in
adults. The EASI lead system has been approved
by the US Food and Drug Administration for as-
sessing normal, abnormal, and paced cardiac
rhythms and for detecting myocardial ischemia
in patients with chest pain or silent ischemia by
monitoring ST-segment elevation or depression.
Because there is individual variation between
standard 12-lead ECG and EASI ECG it is
 HOW MANY ECG LEADS DO WE NEED? 325

advised that clinicians only use one method when One advantage to using this system is that all
making serial ECG comparisons. limb leads and two precordial leads are known
to be true leads, although the system uses more
Derived precordial leads electrodes than the EASI lead system.

Another approach to a reduced number of

leads is to exclude certain leads from the ECG
recording and instead reconstruct the excluded
leads from the existing leads, using either general
or patient-specific reconstruction. Nelwan and
colleagues [109] studied how well the 12-lead
ECG can be reconstructed from different lead
subsets, always including limb leads I and II and
at least one precordial lead. Patients who had un-
stable angina were monitored with extremity elec-
trodes at the Mason-Likar locations and
precordial leads at the standard positions. They
concluded that general construction allows recon-
struction of one or two precordial leads, whereas
patient-specific construction allows up to four
leads to be reconstructed. The best lead subset
with four removed precordial leads was I, II,
V2, and V5. The patient-specific reconstruction, how-
ever, presupposes a previously recorded ECG
and is affected over time, probably by postural
changes. In a comparison of general construction,
patient-specific construction (using leads I, II, V2,
and V5), and EASI leads, it was found that,
when comparing ST60, patient-specific construction
showed a more accurate reconstruction of the stan-
dard ECG than did the other two methods [110].
24-view ECG
Some investigators propose that there is no
need for additional leads if the standard 12-lead
ECG system could be used more effectively [15]. It
is widely considered that ST elevation occurs
when an injury current develops between normal
and transmurally ischemic myocardium [111].
The flow of this current creates a vector toward
leads with positive poles facing that area of the
myocardium, seen as ST elevation. Any electrode
on the other side of the heart will instead record
an ST depression. For example, to see ST eleva-
tion in patients who have posterior infarcts, one
would have to apply posterior thoracic electrodes
(as discussed previously) or to produce a 24-view
ECG in which displays of negative views of all
standard leads are regarded as well (Fig. 5).
The classical standard 12-lead display includes
two subsets of limb leads (I, II, III, and aVL,
aVR, aVF) and the chest leads V1 to V6. It is,
however, equally possible to regroup the limb
leads into an orderly sequence [112]. This se-
quence displays the cardiac electrical activity in
the frontal plane in a logical order: aVL, I,
aVR,
II, and aVF, III, with each separated by

Fig. 5. The 24 views of the standard 12-lead ECG are shown as recorded from a patient receiving angioplasty balloon
occlusion in a nondominant left circumflex coronary artery. QRS complexes and T waves from single cardiac cycles re-
corded simultaneously in the six frontal plane leads viewed from the front (left) and the six transverse plane leads viewed
from below (right) are arranged like numbers around a clock face. (From: Pahlm-Webb U, Pahlm O, Sadanandan S,
et al. A new method for using the direction of ST-segment deviation to localize the site of acute coronary occlusion:
the 24-view standard electrocardiogram. Am J Med 2002;113(1):75–8; with permission.)
326 TRÄGÅRDH
30 . This sequence could be enlarged further to in-
clude all inverted leads in the frontal plane, just
as
aVR has been inverted in the orderly se-
quence. In the same manner, an inversion of the
precordial leads V1 to V6 could also be used.
Thus, an ST depression in lead V1 would present
as an ST elevation in lead
V1, just as an ST de-
pression in lead aVR would present as an ST ele-
vation in lead
aVR. The 24-view ECG does not
require the placement of additional electrodes, be-
cause all information is available in the 12-lead
ECG. An alternative of the 24-view ECG is simply
to allow ST deviation (elevation or depression) in
the standard 12-lead ECG to be diagnostic of acute
MI. Further studies are needed, however, before
any ECG diagnostic guidelines can be changed.
Summary
There is no single right answer to the question
of how many leads are needed in clinical electro-
cardiography. This question must be answered in
the context of the clinical problem to be solved.
The standard 12-lead ECG is so well established
that alternative lead systems must prove their
advantage through well-conducted clinical studies
to achieve clinical acceptance. Certain additional
leads seem to add valuable information in specific
patient groups. The use of a large number of leads
(such as in body surface potential mapping) may
add clinically relevant information; however, it is
still cumbersome, and its clinical advantage is yet
to be proven. Reduced lead sets emulate the 12-
lead ECG reasonably well and are especially
advantageous in the emergency situation.
References
[1] Waller A. A demonstration on man of electromo-
tive changes accompanying the hearts beat. J Phys-
iol 1887;8:229–34.
[2] Einthoven W. The different forms of the human
electrocardiogram and their signification. Lancet
1912;1:853–61 [reprinted in Am Heart J 1950;40:
195–211].
[3] Wilson F, MacLeod A, Barker P. The potential
variations produced by the heart beat at the apices
of Einthovens triangle. Am Heart J 1931;7:207–11.
[4] Wilson F, MacLeod A, Barker P, et al. The electro-
cardiogram in myocardial infarction with particu-
lar reference to the initial deflections of the
ventricular complex. Heart 1933;16:155–99.
[5] Committee of the Cardiac Society of Great Britain
and Ireland and Committee of the American Heart
Association. Standardisation of precordial leads.
Lancet 1938; 221.
et al
[6] Barnes A, Pardee H, White P, et al. Standardiza-
tion of precordial leads. Am Heart J 1938;15:107–8.
[7] Barnes A, Pardee H, White P, et al. Standardiza-
tion of precordial leads: supplementary report.
Am Heart J 1938;15:235–9.
[8] Goldberger E. A simple, indifferent, electrocardio-
graphic electrode of zero potential and a technique
of obtaining augmented, unipolar, extremity leads.
Am Heart J 1942;23:483–92.
[9] Mason RE, Likar I. A new system of multiple-lead
exercise electrocardiography. Am Heart J 1966;
71(2):196–205.
[10] Papouchado M, Walker PR, James MA, et al. Fun-
damental differences between the standard 12-lead
electrocardiography and the modified (Mason-
Likar) exercise lead system. Eur Heart J 1987;
8(7):725–33.
[11] Sevilla DC, Dohrmann ML, Somelofski CA, et al.
Invalidation of the resting electrocardiogram
obtained via exercise electrode sites as a standard
12-lead recording. Am J Cardiol 1989;63(1):35–9.
[12] Speake D, Terry P. Towards evidence based emer-
gency medicine: best BETs from the Manchester
Royal Infirmary. First ECG in chest pain. Emerg
Med J 2001;18(1):61–2.
[13] Blanke H, Cohen M, Schlueter GU, et al. Electro-
cardiographic and coronary arteriographic correla-
tions during acute myocardial infarction. Am J
Cardiol 1984;54(3):249–55.
[14] Saetre HA, Selvester RH, Solomon JC, et al. 16-
lead ECG changes with coronary angioplasty.
Location of ST-T changes with balloon occlusion
of five arterial perfusion beds. J Electrocardiol
1992;24(Suppl):153–62.
[15] Pahlm-Webb U, Pahlm O, Sadanandan S, et al.
A new method for using the direction of ST-seg-
ment deviation to localize the site of acute coronary
occlusion: the 24-view standard electrocardiogram.
Am J Med 2002;113(1):75–8.
[16] Brady WJ. Acute posterior wall myocardial infarc-
tion: electrocardiographic manifestations. Am J
Emerg Med 1998;16(4):409–13.
[17] The Joint European Society of Cardiology/
American College of Cardiology Committee.
Myocardial infarction redefined – A consensus
document of The Joint European Society of
Cardiology/American College of Cardiology Com-
mittee for the Redefinition of Myocardial Infarc-
tion. Eur Heart J 2000;21:1502–13.
[18] Khaw K, Moreyra AE, Tannenbaum AK, et al.
Improved detection of posterior myocardial wall
ischemia with the 15-lead electrocardiogram. Am
Heart J 1999;138(5 Pt 1):934–40.
[19] Matetzky S, Freimark D, Feinberg MS, et al.
Acute myocardial infarction with isolated ST-seg-
ment elevation in posterior chest leads V7–9:
‘‘hidden’’ ST-segment elevations revealing acute
posterior infarction. J Am Coll Cardiol 1999;
34(3):748–53.
 HOW MANY ECG LEADS DO WE NEED?
[20] Oraii S, Maleki M, Tavakolian AA, et al. Preva-
lence and outcome of ST-segment elevation in
posterior electrocardiographic leads during acute
myocardial infarction. J Electrocardiol 1999;
32(3):275–8.
[21] Matetzky S, Freimark D, Chouraqui P, et al. Sig-
nificance of ST segment elevations in posterior
chest leads (V7 to V9) in patients with acute inferior
myocardial infarction: application for thrombo-
lytic therapy. J Am Coll Cardiol 1998;31(3):506–11.
[22] Chia BL, Tan HC, Yip JW, et al. Electrocardio-
graphic patterns in posterior chest leads (V7, V8,
V9) in normal subjects. Am J Cardiol 2000;85(7):
911–2.
[23] Kulkarni AU, Brown R, Ayoubi M, et al. Clinical
use of posterior electrocardiographic leads: a pro-
spective electrocardiographic analysis during coro-
nary occlusion. Am Heart J 1996;131(4):736–41.
[24] Agarwal JB. Routine use of a 15-lead electrocardio-
gram for patients presenting to the emergency de-
partment with chest pain. J Electrocardiol 1998;
31(Suppl):172–7.
[25] Agarwal JB, Khaw K, Aurignac F, et al. Impor-
tance of posterior chest leads in patients with sus-
pected myocardial infarction, but nondiagnostic,
routine 12-lead electrocardiogram. Am J Cardiol
1999;83(3):323–6.
[26] Zalenski RJ, Cooke D, Rydman R, et al. Assessing
the diagnostic value of an ECG containing leads
V4R, V8, and V9: the 15-lead ECG. Ann Emerg
Med 1993;22(5):786–93.
[27] Zalenski RJ, Rydman RJ, Sloan EP, et al. Value of
posterior and right ventricular leads in comparison
to the standard 12-lead electrocardiogram in evalu-
ation of ST-segment elevation in suspected acute
myocardial infarction. Am J Cardiol 1997;79(12):
1579–85.
[28] Ganim RP, Lewis WR, Diercks DB, et al. Right
precordial and posterior electrocardiographic leads
do not increase detection of ischemia in low-risk
patients presenting with chest pain. Cardiology
2004;102(2):100–3.
[29] Wung SF, Drew BJ. New electrocardiographic cri-
teria for posterior wall acute myocardial ischemia
validated by a percutaneous transluminal coronary
angioplasty model of acute myocardial infarction.
Am J Cardiol 2001;87(8):970–4.
[30] Andersen HR, Falk E, Nielsen D. Right ventricular
infarction: frequency, size and topography in coro-
nary heart disease: a prospective study comprising
107 consecutive autopsies from a coronary care
unit. J Am Coll Cardiol 1987;10(6):1223–32.
[31] Haji SA, Movahed A. Right ventricular infarction–
diagnosis and treatment. Clin Cardiol 2000;23(7):
473–82.
[32] Correale E, Battista R, Martone A, et al. Electro-
cardiographic patterns in acute inferior myocardial
infarction with and without right ventricle involve-
ment: classification, diagnostic and prognostic
327
value, masking effect. Clin Cardiol 1999;22(1):
37–44.
[33] Andersen HR, Nielsen D, Lund O, et al. Prognostic
significance of right ventricular infarction diag-
nosed by ST elevation in right chest leads V3R to
V7R. Int J Cardiol 1989;23(3):349–56.
[34] Braat SH, Brugada P, De Zwaan C, et al. Right and
left ventricular ejection fraction in acute inferior
wall infarction with or without ST segment eleva-
tion in lead V4R. J Am Coll Cardiol 1984;4(5):
940–4.
[35] Braat SH, de Zwaan C, Brugada P, et al. Right
ventricular involvement with acute inferior wall
myocardial infarction identifies high risk of devel-
oping atrioventricular nodal conduction distur-
bances. Am Heart J 1984;107(6):1183–7.
[36] Chou TC, Van der Bel-Kahn J, Allen J, et al. Elec-
trocardiographic diagnosis of right ventricular in-
farction. Am J Med 1981;70(6):1175–80.
[37] Croft CH, Nicod P, Corbett JR, et al. Detection of
acute right ventricular infarction by right precor-
dial electrocardiography. Am J Cardiol 1982;50(3):
421–7.
[38] Braat SH, Brugada P, de Zwaan C, et al. Value of
electrocardiogram in diagnosing right ventricular
involvement in patients with an acute inferior wall
myocardial infarction. Br Heart J 1983;49(4):
368–72.
[39] Lopez-Sendon J, Coma-Canella I, Alcasena S, et al.
Electrocardiographic findings in acute right ven-
tricular infarction: sensitivity and specificity of elec-
trocardiographic alterations in right precordial
leads V4R, V3R, V1, V2, and V3. J Am Coll Cardiol
1985;6(6):1273–9.
[40] Andersen HR, Thomsen PE, Nielsen TT, et al. ST
deviation in right chest leads V3R to V7R during
percutaneous transluminal coronary angioplasty.
Am Heart J 1990;119(3 Pt 1):490–3.
[41] Morgera T, Alberti E, Silvestri F, et al. Right pre-
cordial ST and QRS changes in the diagnosis of
right ventricular infarction. Am Heart J 1984;
108(1):13–8.
[42] Andersen HR, Nielsen D, Hansen LG. The normal
right chest electrocardiogram. J Electrocardiol
1987;20(1):27–32.
[43] Braat SH, Brugada P, den Dulk K, et al. Value of
lead V4R for recognition of the infarct coronary ar-
tery in acute inferior myocardial infarction. Am J
Cardiol 1984;53(11):1538–41.
[44] Ueshima K, Kobayashi N, Kamata J, et al. Do the
right precordial leads during exercise testing con-
tribute to detection of coronary artery disease?
Clin Cardiol 2004;27(2):101–5.
[45] Shry EA, Eckart RE, Furgerson JL, et al. Addition
of right-sided and posterior precordial leads during
stress testing. Am Heart J 2003;146(6):1090–4.
[46] Michaelides AP, Psomadaki ZD, Dilaveris PE,
et al. Improved detection of coronary artery disease
by exercise electrocardiography with the use of
328 TRÄGÅRDH
right precordial leads. N Engl J Med 1999;340(5):
340–5.
[47] Akula R, Hasan SP, Alhassen M, et al. Right-sided
EKG in pulmonary embolism. J Natl Med Assoc
2003;95(8):714–7.
[48] Chia BL, Tan HC, Lim YT. Right sided chest lead
electrocardiographic abnormalities in acute pulmo-
nary embolism. Int J Cardiol 1997;61(1):43–6.
[49] Waller AD. On the electromotive changes con-
nected with the beat of the mammalian heart, and
of the human heart in particular. Philos Trans R
Soc 1889;180:169–94.
[50] Taccardi B. Distribution of heart potentials on the
thoracic surface of normal human subjects. Circ
Res 1963;12:341–52.
[51] Nahum LH, Mauro A, Chernoff HM, et al. Instan-
taneous equipotential distribution on surface of the
human body for various instants in the cardiac cy-
cle. J Appl Physiol 1951;3(8):454–64.
[52] Gulrajani RM, Roberge FA, Mailloux GE. The
forward problem of electrocardiography. In:
Macfarlane P, Lawrie T, editors. Comprehensive
electrocardiology: theory and practice in health
and disease. New York: Pergamon Press; 1989. p.
197–236.
[53] Gulrajani RM, Roberge FA, Savard P. The inverse
problem of electrocardiography. In: Macfarlane P,
Lawrie T, editors. Comprehensive electrocardiology:
theory and practice in health and disease. New York:
Pergamon Press; 1989. p. 237–88.
[54] Montague TJ, Smith ER, Spencer CA, et al. Body
surface electrocardiographic mapping in inferior
myocardial infarction. Manifestation of left and
right ventricular involvement. Circulation 1983;
67(3):665–73.
[55] Kornreich F, Rautaharju PM, Warren J, et al.
Identification of best electrocardiographic leads
for diagnosing myocardial infarction by statistical
analysis of body surface potential maps. Am J Car-
diol 1985;56(13):852–6.
[56] Menown IB, Patterson RS, MacKenzie G, et al.
Body-surface map models for early diagnosis of
acute myocardial infarction. J Electrocardiol
1998;31(Suppl):180–8.
[57] Menown IB, Allen J, Anderson JM, et al. Early di-
agnosis of right ventricular or posterior infarction
associated with inferior wall left ventricular acute
myocardial infarction. Am J Cardiol 2000;85(8):
934–8.
[58] Maynard SJ, Menown IB, Manoharan G, et al.
Body surface mapping improves early diagnosis
of acute myocardial infarction in patients with
chest pain and left bundle branch block. Heart
2003;89(9):998–1002.
[59] Yamada K, Toyama J, Sugenoya J, et al. Body
surface isopotential maps. Clinical application to
the diagnosis of myocardial infarction. Jpn Heart
J 1978;19(1):28–45.
et al
[60] Vincent GM, Abildskov JA, Burgess MJ, et al. Di-
agnosis of old inferior myocardial infarction by
body surface isopotential mapping. Am J Cardiol
1977;39(4):510–5.
[61] Vesterinen P, Hanninen H, Karvonen M, et al.
Temporal analysis of the depolarization wave of
healed myocardial infarction in body surface
potential mapping. Ann Noninvasive Electrocar-
diol 2004;9(3):234–42.
[62] Toyama S, Suzuki K, Koyama M, et al. The body
surface isopotential mapping of the QRS wave in
myocardial infarction. a comparative study of the
scintigram with thallium-201. J Electrocardiol
1982;15(3):241–8.
[63] Toyama S, Suzuki K, Koyama M. The isopotential
mapping of the T wave in myocardial infarction.
J Electrocardiol 1980;13(4):323–30.
[64] Pham-Huy H, Gulrajani RM, Roberge FA, et al.
A comparative evaluation of three different
approaches for detecting body surface isopotential
map abnormalities in patients with myocardial
infarction. J Electrocardiol 1981;14(1):43–55.
[65] Osugi J, Ohta T, Toyama J, et al. Body surface
isopotential maps in old inferior myocardial infarc-
tion undetectable by 12 lead electrocardiogram.
J Electrocardiol 1984;17(1):55–62.
[66] Ohta T, Toyama J, Ohsugi J, et al. Correlation
between body surface isopotential maps and left
ventriculograms in patients with old anterior myo-
cardial infarction. Jpn Heart J 1981;22(5):747–61.
[67] Ohta T, Kinoshita A, Ohsugi J, et al. Correlation
between body surface isopotential maps and left
ventriculograms in patients with old inferoposte-
rior myocardial infarction. Am Heart J 1982;
104(6):1262–70.
[68] Medvegy M, Preda I, Savard P, et al. New body
surface isopotential map evaluation method to de-
tect minor potential losses in non-Q-wave myocar-
dial infarction. Circulation 2000;101(10):1115–21.
[69] Kornreich F, Montague TJ, Rautaharju PM. Iden-
tification of first acute Q wave and non-Q wave
myocardial infarction by multivariate analysis of
body surface potential maps. Circulation 1991;
84(6):2442–53.
[70] Hirai M, Ohta T, Kinoshita A, et al. Body surface
isopotential maps in old anterior myocardial in-
farction undetectable by 12-lead electrocardio-
grams. Am Heart J 1984;108(4 Pt 1):975–82.
[71] Flowers NC, Horan LG, Sohi GS, et al. New evi-
dence for inferoposterior myocardial infarction
on surface potential maps. Am J Cardiol 1976;
38(5):576–81.
[72] Flowers NC, Horan LG, Johnson JC. Anterior
infarctional changes occurring during mid and late
ventricular activation detectable by surface map-
ping techniques. Circulation 1976;54(6):906–13.
[73] De Ambroggi L, Bertoni T, Rabbia C, et al. Body
surface potential maps in old inferior myocardial
 HOW MANY ECG LEADS DO WE NEED?
infarction. Assessment of diagnostic criteria.
J Electrocardiol 1986;19(3):225–34.
[74] Ackaoui A, Nadeau R, Sestier F, et al. Myocardial
infarction diagnosis with body surface potential
mapping, electrocardiography, vectorcardiogra-
phy and thallium-201 scintigraphy: a correlative
study with left ventriculography. Clin Invest Med
1985;8(1):68–77.
[75] Sgarbossa EB, Pinski SL, Barbagelata A, et al.
Electrocardiographic diagnosis of evolving acute
myocardial infarction in the presence of left bun-
dle-branch block. GUSTO-1 (Global Utilization
of Streptokinase and Tissue Plasminogen Activator
for Occluded Coronary Arteries) Investigators.
N Engl J Med 1996;334(8):481–7.
[76] Hands ME, Cook EF, Stone PH, et al. Electrocar-
diographic diagnosis of myocardial infarction in
the presence of complete left bundle branch block.
Am Heart J 1988;116(1 Pt 1):23–31.
[77] Menown IB, Allen J, Anderson JM, et al. Noninva-
sive assessment of reperfusion after fibrinolytic
therapy for acute myocardial infarction. Am J Car-
diol 2000;86(7):736–41.
[78] Cahyadi YH, Takekoshi N, Matsui S. Clinical effi-
cacy of PTCA and identification of restenosis: eval-
uation by serial body surface potential mapping.
Am Heart J 1991;121(4 Pt 1):1080–7.
[79] Manninen H, Takala P, Makijarvi M, et al. Re-
cording locations in multichannel magnetocardiog-
raphy and body surface potential mapping sensitive
for regional exercise-induced myocardial ischemia.
Basic Res Cardiol 2001;96(4):405–14.
[80] Montague TJ, Johnstone DE, Spencer CA, et al.
Body surface potential maps with low-level exercise
in isolated left anterior descending coronary artery
disease. Am J Cardiol 1988;61(4):273–82.
[81] Boudik F, Anger Z, Aschermann M, et al. Dipyri-
damole body surface potential mapping: noninva-
sive differentiation of syndrome X from coronary
artery disease. J Electrocardiol 2002;35(3):181–91.
[82] Kornreich F, Montague TJ, Rautaharju PM, et al.
Identification of best electrocardiographic leads for
diagnosing left ventricular hypertrophy by statisti-
cal analysis of body surface potential maps. Am
J Cardiol 1988;62(17):1285–91.
[83] Kornreich F, Montague TJ, van Herpen G, et al.
Improved prediction of left ventricular mass by re-
gression analysis of body surface potential maps.
Am J Cardiol 1990;66(4):485–92.
[84] Holt JH Jr, Barnard AC, Kramer JO Jr. Multiple
dipole electrocardiography: a comparison of elec-
trically and angiographically determined left ven-
tricular masses. Circulation 1978;57(6):1129–33.
[85] Yamada K, Toyama J, Wada M, et al. Body sur-
face isopotential mapping in Wolff-Parkinson-
White syndrome: noninvasive method to determine
the localization of the accessory atrioventricular
pathway. Am Heart J 1975;90(6):721–34.
329
[86] Tseng YZ, Hsu KL, Chiang FT, et al. The use
of body surface potential map for identifying
sites of accessory pathway in patients with Wolff-
Parkinson-White syndrome. Jpn Heart J 1998;39(4):
445–55.
[87] Liebman J, Zeno JA, Olshansky B, et al. Electro-
cardiographic body surface potential mapping in
the Wolff-Parkinson-White syndrome. Noninva-
sive determination of the ventricular insertion sites
of accessory atrioventricular connections. Circula-
tion 1991;83(3):886–901.
[88] Iwa T, Magara T. Correlation between localization
of accessory conduction pathway and body surface
maps in the Wolff-Parkinson-White syndrome. Jpn
Circ J 1981;45(10):1192–8.
[89] Dubuc M, Nadeau R, Tremblay G, et al. Pace map-
ping using body surface potential maps to guide
catheter ablation of accessory pathways in patients
with Wolff-Parkinson-White syndrome. Circula-
tion 1993;87(1):135–43.
[90] De Ambroggi L, Taccardi B, Macchi E. Body-
surface maps of heart potentials: tentative
localization of pre-excited areas in forty-two
Wolff-Parkinson-White patients. Circulation
1976;54(2):251–63.
[91] Benson DW Jr, Sterba R, Gallagher JJ, et al. Local-
ization of the site of ventricular preexcitation with
body surface maps in patients with Wolff-Parkin-
son-White syndrome. Circulation 1982;65(6):
1259–68.
[92] Carley SD, Jenkins M, Jones KM. Body surface
mapping versus the standard 12 lead ECG in the
detection of myocardial infarction amongst emer-
gency department patients: a Bayesian approach.
Resuscitation 2005;64(3):309–14.
[93] Frank E. An accurate, clinically practical system
for spatial vectorcardiography. Circulation 1954;
13(4):737–49.
[94] Dellborg M, Topol EJ, Swedberg K. Dynamic QRS
complex and ST segment vectorcardiographic
monitoring can identify vessel patency in patients
with acute myocardial infarction treated with
reperfusion therapy. Am Heart J 1991;122(4 Pt 1):
943–8.
[95] Dellborg M, Steg P, Simoons M, et al. Vectorcar-
diographic monitoring to assess early vessel pa-
tency after reperfusion therapy for acute
myocardial infarction. Eur Heart J 1995;16(1):
21–9.
[96] Dellborg M, Herlitz J, Risenfors M, et al. Elec-
trocardiographic assessment of infarct size:
comparison between QRS scoring of 12-lead elec-
trocardiography and dynamic vectorcardiography.
Int J Cardiol 1993;40(2):167–72.
[97] Willems JL, Abreu-Lima C, Arnaud P, et al. As-
sessment of the diagnostic performance of ECG
computer programs and cardiologists. In: Common
standards for quantitative electrocardiography.
330 TRÄGÅRDH
CSE 10th progress report. Leuven (Belgium): Acco;
1990. p. 197–236.
[98] Dower GE, Yakush A, Nazzal SB, et al. Deriving
the 12-lead electrocardiogram from four (EASI)
electrodes. J Electrocardiol 1988;21(Suppl):
S182–7.
[99] Feild DQ, Feldman CL, Horacek BM. Improved
EASI coefficients: their derivation, values, and per-
formance. J Electrocardiol 2002;35(Suppl):23–33.
[100] Klein M, Key-Brothers I, Feldman C. Can the vec-
torcardiographically derived EASI ECG be a suit-
able surrogate for the standard ECG in selected
circumstances? Proc Comput Cardiol 1997;5(3):
721–4.
[101] Drew BJ, Pelter MM, Wung SF, et al. Accuracy of
the EASI 12-lead electrocardiogram compared to
the standard 12-lead electrocardiogram for diag-
nosing multiple cardiac abnormalities. J Electro-
cardiol 1999;32(Suppl):38–47.
[102] Drew BJ, Adams MG, Pelter MM, et al. ST seg-
ment monitoring with a derived 12-lead electro-
cardiogram is superior to routine cardiac care
unit monitoring. Am J Crit Care 1996;5(3):
198–206.
[103] Rautaharju PM, Zhou SH, Hancock EW, et al.
Comparability of 12-lead ECGs derived from
EASI leads with standard 12-lead ECGS in the
classification of acute myocardial ischemia and
old myocardial infarction. J Electrocardiol 2002;
35(Suppl):35–9.
[104] Feldman CL, MacCallum G, Hartley LH. Com-
parison of the standard ECG with the EASIcar-
diogram for ischemia detection during exercise
et al
monitoring. Proc Comput Cardiol 1997;24:
343–5.
[105] Wehr G, Peters R, Khalife K, et al. A vector-based
5 electrode 12-lead ECG (EASI) is equivalent to the
conventional 12-lead ECG for diagnosis of myo-
cardial ischemia. J Am Coll Cardiol 2002;
39(Suppl):122A.
[106] Horacek BM, Warren JW, et al. Diagnostic accu-
racy of derived versus standard 12-lead electrocar-
diograms. J Electrocardiol 2000;33(Suppl):155–60.
[107] Sejersten M, Pahlm O, Pettersson J, et al. The rela-
tive accuracies of ECG precordial lead waveforms
derived from EASI leads and those acquired from
paramedic applied standard leads. J Electrocardiol
2003;36(3):179–85.
[108] Pahlm O, Pettersson J, Thulin A, et al. Comparison
of waveforms in conventional 12-lead ECGs and
those derived from EASI leads in children. J Elec-
trocardiol 2003;36(1):25–31.
[109] Nelwan SP, Kors JA, Meij SH, et al. Reconstruc-
tion of the 12-lead electrocardiogram from reduced
lead sets. J Electrocardiol 2004;37(1):11–8.
[110] Nelwan SP, Crater SW, Meij SH, et al. Simulta-
neous comparison of three derived 12-lead ECGs
with standard ECG at rest and during percutane-
ous coronary occlusion. J Electrocardiol 2004;
37(Suppl):171–2.
[111] Wagner GS. Marriott’s practical electrocardiogra-
phy. 10th edition. Philadelphia: Lippincott Wil-
liams & Wilkins; 2001.
[112] Anderson ST, Pahlm O, Selvester RH, et al. Pano-
ramic display of the orderly sequenced 12-lead
ECG. J Electrocardiol 1994;27(4):347–52.
 Cardiol Clin 24 (2006) 331–342

Consideration of Pitfalls in and Omissions

from the Current ECG Standards for Diagnosis
of Myocardial Ischemia/Infarction
in Patients Who Have
Acute Coronary Syndromes
Galen Wagner, MDa,*, Tobin Lim, MDa, Leonard Gettes, MDd,
Anton Gorgels, MDe, Mark Josephson, MDf, Hein Wellens, MDe,
Stanley Anderson, MB, BS, FRACPg, Rory Childers, MDh,
Peter Clemmensen, MD, PhDi, Paul Kligfield, MDk,
Peter Macfarlane, PhDb, Olle Pahlm, MD, PhDc,
Ronald Selvester, MDj
a
Division of Cardiology, Department of Medicine, Duke University Medical Center,
2400 Pratt Street, RM 0306, Durham, NC 27705, USA
b
Division of Cardiovascular and Medical Sciences, Section of Cardiology, Royal Infirmary,
Glasgow G31 2ER, Scotland, UK
c
Department of Clinical Physiology, Lund University Hospital, LunSE-221 85 Lund, Sweden
d
Division of Cardiology, The University of North Carolina at Chapel Hill, CB #7075, Bioinformatics Building,
130 Mason Farm Road, 4th Floor, Chapel Hill, NC 27599-7075, USA
e
Department of Cardiology, Cardiovascular Research Institute Maastricht,
PO Box 5800, 6202 Maastricht, The Netherlands
f
Harvard-Thorndike Electrophysiology Institute and Arrhythmia Service, Cardiology,
Harvard Medical School, Beth Israel Deaconess Medical Center,
185 Pilgrim Road, Baker 4, Boston, MA 02215, USA
g
Cabrini Hospital, 18 Bay Street, Victoria 3186, Australia
h
Section of Cardiology, University of Chicago Medical Center,
5841 S. Maryland Avenue, Chicago, IL 60637, USA
i
Rigs Hospital, The Heart Center, University of Copenhagen, Blegdamsvej 9, 2100 Copenhagen, East Denmark
j
Long Beach Memorial Hospital, 6298 East Ocean Boulevard, Long Beach, CA 90803, USA
k
Department of Medicine, Cardiac Health Center, New York-Presbyterian Hospital,
525 East 68th Street, New York, NY 10021, USA

The ECG is the key clinical test available
for the emergency determination of the patients
with acute coronary syndromes who are indeed
having acute myocardial ischemia/infarction. In
these individuals, the process of evolution from
* Corresponding author.
E-mail address: Galen.wagner@duke.edu
(G. Wagner).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.012
potentially reversible ischemia to irreversible in-
farction occurs during the minutes to hours
following coronary artery occlusion. Because the
etiology is typically thrombosis, the correct clin-
ical decision regarding reperfusion therapy is
crucial.
Reperfusion using either intravenous throm-
bolytic therapy or intracoronary intervention has
become the standard of care for the subgroup of
cardiology.theclinics.com
332 WAGNER et al

patients who have acute coronary syndrome who
The healing phase occurs during the weeks when
meet ECG criteria for ST elevation myocardial necrosis and inflammation are replaced by scar
infarction (STEMI); but is considered contra- in the infarcted portion and stunning is replaced
indicated for all others characterized by the broad by contraction in the salvaged portion [4].
term ‘‘non-STEMI’’. It has become clear, how-
The chronic phase occurs during the months
ever, that the non-STEMI subgroup with ST and years when residual scar coexists with
depression in certain ECG leads should be con- compensating myocardium.
sidered ‘‘STEMI-equivalent’’ and similarly con-
This article focuses on the pitfalls in the
sidered for reperfusion therapy. The application
existing ECG criteria for the diagnosis of the
of STEMI criteria alone results in high specificity
ischemic and infarcting phases of the ischemia/
but unacceptably low sensitivity [1].
infarction process. Other articles in this issue by
For many years, the standards for application
Atar and colleagues consider the diagnosis of
of the ECG for decision support for patients who
STEMI, and of reperfusion based on presence of
have acute coronary syndrome have been un-
resolution of ST-segment elevation. As indicated
changed [2,3]. During the past 5 years, however,
in the article on diagnosis of STEMI by Atar and
groups such as the American Heart Association
colleagues, there are current strong criteria for
(AHA), European Society of Cardiology (ESC),
both identification and localization, and ways to
and American College of Cardiology (ACC)
attain high specificity. The sensitivity of these
have initiated working groups to evolve new diag-
criteria is only about 50%, however, and it is un-
nostic standards [4]. As data become available to
determined what portion of ECG false negatives is
produce sufficient documentation of new criteria
caused by true ‘‘ECG silence’’. Because at the time
in peer-review journals, the pitfalls in current
of initial emergency evaluation it is not possible to
standards are corrected. This process, however,
determine whether the ischemic or infracting
requires the passage of time. This review emerged
phase is present, the term ‘‘ischemia/infracting
from the efforts of an AHA working group to de-
phase’’ is used. The ST-segment changes of trans-
velop new standards for clinical application of
mural ischemia may mask the Q-wave changes
electrocardiology. The general document has
of infarction, and the Q-wave changes of ischemia
been completed by Kligfield and colleagues [5],
may resolve when myocardium is salvaged by
and specific documents regarding ischemia/infarc-
reperfusion. Indeed, recent studies have docu-
tion and other aspects of electrocardiography are
mented a 13% to 17% incidence of aborted
in development. The pitfalls in the current diag-
infarction when patients receive immediate reper-
nostic standards regarding ischemia/infarction
fusion therapy triggered by the acute ECG
that have been identified by sufficiently docu-
changes of ischemia/infarction [7,8].
mented studies will be corrected. This article refers
only to the pitfalls in and omissions from current
standards for which new standards will potentially
Criteria for diagnosis of ischemia/infarction
emerge in future years.
The ECG provides diagnostic information that Patients who have acute myocardial ischemia/
complements that from other clinical methods infarction constitute a minority of the vast num-
during the sequential phases of the myocardial bers of individuals with the broad range of
ischemia/infarction process (as identified by the symptoms suggestive of acute coronary syndrome
AHA Working Group) [6]. who indeed require emergency diagnosis and
therapy. ECG standards for detection of the

The ischemic phase persists during the initial
presence, location, acuteness, severity, and extent
minutes when only potentially reversible is-
of myocardium jeopardized by acute coronary
chemia is present because the viability of the
occlusion provide support for the clinical de-
jeopardized myocardium is maintained by an-
cisions required to maximize salvage and mini-
aerobic metabolism.
mize morbidity and mortality. High ECG

The infarcting phase occurs during the hours
diagnostic accuracy is required because
when some protection is provided by collat-
eral vessels and ischemic preconditioning.
The subsequent health of the individual is so

The reperfusing phase occurs during the critically influenced by infarcted myocardium.
minutes following spontaneous or therapeutic
The typical thrombotic cause of the coronary
restoration of coronary flow. occlusion is potentially reversible.
 PITFALLS IN ECG STANDARDS OF DIAGNOSIS

Intravenous or intracoronary reperfusion
therapy has both high economic cost and
high risk of serious adverse consequences.
Acute coronary thrombosis typically produces
potentially reversible myocardial ischemia within
seconds, which then evolves to infarction within
minutes in the absence of reperfusion [9]. The
ECG changes associated with this process can
be categorized into those affecting repolarization
(ie, the ST segment and T wave) and those affect-
ing depolarization (ie, the QRS complex). Fol-
lowing acute coronary occlusion, as contrasted
with exercise-induced ischemia, the initial ECG
change is deviation of the T waves toward the in-
volved myocardial region [6]. These hyperacute
T-wave changes typically are accompanied by
or followed by similarly directed deviation of
the ST segment from the TP/PR baseline. This
ST-segment deviation is indicated by ST eleva-
tion in the ECG leads directed toward the in-
volved region and by ST depression in leads
directed away from the involved region. These
ST deviations are produced by the flow of injury
currents generated by voltage gradients across the
boundary between the ischemic and non-ischemic
myocardium during the resting and early repolar-
ization phases of the ventricular action potential
[10,11]. Questions remain about the exact loca-
tion and configuration of this boundary and the
mechanism of generation of the injury current.
The involved myocardial region is determined
by identifying the ECG lead or spatially contigu-
ous pair of leads with the maximal positive or
negative ST deviation. This method is in contrast
to that for ischemia of increased myocardial
demand, where the direction of ST-segment de-
viation on the surface ECG does not accurately
localize the involved region [12].
Pitfalls in the current diagnostic standards
The current criterion for the diagnosis of acute
myocardial ischemia/infarction is ST-segment el-
evation of at least 0.1 mV in two contiguous ECG
leads (an 0.2-mV threshold is required in precor-
dial leads V1 to V3) [3].
There are pitfalls in each of the three compo-
nents of this standard, however:
1. Two contiguous leads
2. ST segment elevation
3. A single diagnostic deviation threshold for
both genders and all age groups
333
An additional pitfall is the inconsistency of the
terms used for ECG designation of the regions of
the left ventricle involved by acute ischemia/
infarction in regard to those adopted as standards
for cardiac imaging [13]. Each of these pitfalls and
potential methods for their elimination are
considered.
Two contiguous leads
It is difficult to determine which of the ECG
leads are really contiguous (adjacent) unless they
are displayed in the order of their sequential
viewpoints of the cardiac electrical activity. Only
the six chest leads are currently displayed in their
contiguous order: V1, V2, V3, V4, V5, and V6.
The limb leads usually are displayed in two
separate groups of three: I, II, and III, and
aVR, aVL, and aVF. This display is the existing
standard despite the 30-year experience in Sweden
with the orderly (Cabrera) display of aVL, I,
aVR, II, aVF, III (Fig. 1) [14]. Indeed, this
change was endorsed in the 2000 ESC/ACC con-
sensus guidelines [4]. The orderly (Cabrera) se-
quence of frontal plane leads is contrasted with
the classic sequence in Fig. 1. It should be noted
that the classic display of the frontal plane in-
cluded lead aVR, but the orderly display replaces
this lead with lead –aVR. This lead replacement
provides six spatially contiguous leads and there-
fore five pairs of spatially contiguous leads in
the frontal as well as in the transverse plane.
Only three pairs of spatially contiguous leads are
provided by the classic limb lead display: aVL
and I, II and aVF, and aVF and III.
ST elevation
In any of the 12 standard ECG leads with their
positive poles directed toward the region of
ischemia/infarction, ST segment elevation would
be recorded, and STEMI would be diagnosed. In
leads with their negative poles directed toward the
involved region, however, ST-segment depression
would be recorded; and ‘‘non-STEMI’’ would be
diagnosed. As illustrated in Fig. 2, neither the
150 arc in the frontal plane including the positive
poles of the six orderly displayed limb leads
(Fig. 2B) nor the 150 arc in the transverse plane
including the positive poles of the six orderly dis-
played chest leads (Fig. 2C) provide views of the
basal and middle segments of either the anterosu-
perior or posterolateral left ventricular (LV) walls.
Just as lead aVR and lead –aVR are two distinct
ECG leads, each of the other 11 standard leads has
334 WAGNER et al

Fig. 1. The classic and the Cabrera orderly displays of the limb leads are contrasted. Note that lead aVR in the classic
display is replaced by lead –aVR in the Cabrera orderly display.

a reciprocal (antipodal, inverted, mirror-lake)
counterpart. Thus the 10 recording electrodes re-
quired to generate the standard 12-lead ECG
potentially provide a 24-lead ECG, because all 12
leads are bipolar. Leads I, II, and III are universally
accepted as bipolar because two physical electrodes
are used for their generation. Alternatively, the
negative poles of the augmented limb leads aVR,
aVL, and aVF are provided by the average of the
potentials recorded by the other two limb elec-
trodes, and the negative poles of the chest leads V1,
V2, V3, V4, V5, and V6 are provided by Wilson’s
central terminal that includes the potentials
recorded from all three of the limb electrodes [16].
The 12 leads in each plane can be viewed around
the two 360 perimeters as ‘‘clockface displays’’
(Fig. 3) [17,18].
The pitfall of requiring ST elevation for acute
ischemia/infarction diagnosis has profound impli-
cations in this era when reperfusion therapy is
considered to be indicated in the presence of
STEMI but contraindicated in the presence of
non-STEMI.
This ‘‘non-STEMI’’ group is large and hetero-
geneous; including four very different sub-groups
based on their ECG appearance:

Fig. 2. The ECG leads are displayed in relation to a schematic view of the heart in its typical position in the thorax as
documented by cardiac MRI [15] in the (A) transverse and (B and C) frontal planes. The chest leads are displayed in their
classical orderly sequence in A, and the limb leads in the classic nonorderly sequence in B. The limb leads are then dis-
played in the Cabrera orderly sequence in C. Note lead aVR at 150 in B and lead –aVR at þ30 in C.
 PITFALLS IN ECG STANDARDS OF DIAGNOSIS
Fig. 3. The full 24-lead ECG is presented as clock-face displays of the frontal (left) and the transverse (right) planes. The
schematic views of the heart are as in Fig. 2.
1. NORMAL - Those with essentially normal
ECGs
2. NON-SPECIFIC ST-T - Those with minimal
ST segment elevation or T wave inversion
3. CONFOUNDED - Those with ECG con-
founding factors such as ventricular hyper-
trophy or bundle branch block
4. STEMI EQUIVALENT - Those with ST de-
pression but no ECG confounding factors
Only individuals in the STEMI EQUIVA-
LENT sub-group should be considered candidates
for reperfusion therapy. Indeed, individuals in the
CONFOUNDED subgroup may have conditions
such as hypertension induced LV hypertrophy
that increase their potential for intracerebral
bleeding complications of thrombolytic therapy.
It has been shown that sensitivity for postero-
lateral infarction can be increased by recording
from nonstandard posterior chest leads V7 to V9
[19,20]. The placement of these additional elec-
trodes has not been widely accepted, however.
Further studies are required to determine if use
of these additional electrodes increases the sensi-
tivity for diagnosis of acute coronary occlusion
beyond that achieved by consideration of either
the 24-lead ECG or STEMI-equivalent ST depres-
sion in the 12-lead ECG. Indeed, patients who
have even the most proximal site of thrombotic
occlusion, the left main coronary, have wide-
spread ischemia manifested primarily by ST de-
pression, with ST elevation typically only in two
335
noncontiguous leads (limb lead aVR and chest
lead V1) [21,22].
A single diagnostic ST-deviation threshold
The pitfall of a single diagnostic ST-deviation
threshold has been addressed in the 2006 New ECG
Standards of the AHA, ESC, and ACC. Studies by
Macfarlane and colleagues [23] have been used to
establish age- and gender-specific thresholds for
the diagnosis of STEMI in the standard 12-lead
ECG. Future studies are required to establish
such thresholds for ST-segment elevation in the
negative leads or for STEMI-equivalent ST depres-
sion in the standard leads. Reference to the tables of
normal values in Macfarlane and Lawrie [24], how-
ever, reveals that more than 0.1-mV ST elevation in
leads V2 and V3 is the mean normal level in middle-
aged white males. Indeed, this is the reason that
higher STEMI thresholds are required for these
leads than the 0.1-mV threshold required for the
other standard leads. Thus, a threshold of even
less than 0.1 mV of oppositely directed ST deviation
will be required to achieve adequate sensitivity for
diagnosis of lateral wall ischemia/infarction.
Myocardial locations of ischemia/infarction
A common diagnostic pitfall is inconsistency in
the terms used to designation the locations of the
ischemia/infarction process in the LV myocardium.
A map of the coronary arteries and the LV regions
336 WAGNER et al

of ischemia/infarction resulting from their occlu-
sion is presented in Fig. 4 [25]. Like the Mercator
projection of a map of the earth, these Mercator
views provide a planar perspective of the spatial re-
lationships of the anatomy and pathology to the
basal, middle, and apical segments of the walls of
the LV. The LV is divided into four walls: the ante-
roseptal and anterosuperior walls supplied by the
left anterior descending (LAD) coronary artery,
the posterolateral wall supplied by the left circum-
flex (LCX) coronary artery, and the inferior wall
supplied from the posterior descending coronary
artery (the terminal aspect of either the right coro-
nary artery [RCA 90%] or LCX [10%]). Table 1 re-
lates the sites of coronary occlusion to the involved
segments of the LV walls.
The LAD supplies perforating branches into
the septal wall and diagonal branches into the
anterosuperior wall as it continues on its course
from the left main coronary in the groove between
the right ventricle (RV) and LV (see Fig. 4).
Occlusion of the mid to distal LAD produces is-
chemia/infarction primarily in the apical segment
of the anteroseptal wall (Fig. 5A). Occlusion prox-
imal to the first diagonal branch produces involve-
ment extending into the middle segment of that
wall and also the middle and apical segments of
the anterosuperior wall and the apical segments
of the posterolateral and inferior walls (Fig. 5B).
When the occlusion is in the first diagonal branch,
the basal and middle segments of the anterior wall
are involved (Fig. 5C). Previously the term ‘‘high
lateral’’ has been applied, but a better term would
be ‘‘anterosuperior’’ ischemia/infarction. When
the occlusion is even proximal to the first septal
perforating branch, left or right bundle-branch
block may occur.
The LCX supplies marginal branches into the
anterosuperior wall as it continues on its course
from the left main coronary in the groove between
the left atrium and ventricle (see Fig. 4). In about
90% of individuals it is nondominant and

Fig. 4. Mercator views are presented of the coronary arteries (above) and the left ventricle (below). They have been di-
vided circumferentially into quadrants to coincide with the left ventricular walls (I, anteroseptal; II, anterosuperior; III,
posterolateral; IV, inferior) and longitudinally into basal, middle, and apical segments. The RV has been removed to
provide a clear view of the interventricular septum in the entire anteroseptal quadrant and in the most anterior
aspect of the inferior quadrant. The left ventricular myocardium has been color-coded, with light blue representing
the extents of ischemia/infarction resulting from distal occlusions of each of the major coronary arteries, yellow repre-
senting the extents of involvement resulting from proximal occlusions of the LCX or RCA, and dark blue representing
the extent of involvement resulting from proximal occlusion of the LAD.
 PITFALLS IN ECG STANDARDS OF DIAGNOSIS 337

Table 1
Sites of ischemia/infarction resulting from occlusion of the major coronary arteries
Ischemia/Infarct Term LV Wall(s) Wall Segment(s) Coronary Artery Level
Anteroseptal Anteroseptal Apical LAD Mid-distal
Extensive anterior Anteroseptal Apical/middle LAD Proximal
Anterosuperior Apical/middle
Posterolateral Apical
Inferior Apical
Mid-anteriora Anterosuperior Basal/middle Diagonal or marginal Proximal
Posterolaterala Posterolateral Basal LCX Mid-distal
Extensive lateral Posterolateral Basal/middle LCX Proximal
Anterosuperior Basal
Inferolateral Inferior Basal/middle RCA Mid
Posterolateral Basal
Extensive inferior Inferior (and RV) Basal/middle RCA Proximal
Inferior Inferior Basal/middle Posterior descending Proximal
Abbreviations: LAD, left anterior descending coronary artery; LCX, left circumflex coronary artery; RCA, right
coronary artery; RV, right ventricular.
a
Represented on the 12 lead ECG as STEMI-equivalent ST depression.

terminates within that groove. LCX occlusion dis-
tal to its first marginal branch produces ischemia/
infarction primarily in the basal segment of the
posterolateral wall (Fig. 5D). This term should re-
place ‘‘posterior’’ or ‘‘postero-lateral’’ to coincide
with the term typically used when the myocar-
dium is visualized directly by clinical imaging
techniques [26,27]. Occlusion proximal to the first
marginal branch produces involvement extending
into the middle segment of the posterolateral
wall and also the basal and middle segments of
the anterosuperior wall (Fig. 5E). Occlusion in
the marginal branch involves segments of the an-
terosuperior wall similar to those described for the
diagonal branch of the LAD (see Fig. 5C).
The RCA is dominant in about 90% of
individuals because, after coursing (and supplying
RV branches) in the groove between the right
atrium and ventricle, it turns abruptly as the
posterior descending branch in the groove be-
tween the RV and LV (see Fig. 4). Occlusion of
the RCA distal to the first RV branch produces is-
chemia/infarction primarily in the basal and mid-
dle segments of the inferior wall with variable
extension into the posterolateral wall (Fig. 5F).
Occlusion proximal to the first RV branch pro-
duces involvement of that ventricle, as well as
the aspects of the LV indicated previously
(Fig. 5G). Occlusion of the posterior descending
branch of the dominant coronary produces in-
volvement limited to the basal and middle seg-
ments of the inferior wall (Fig. 5H).
Fig. 5C and D indicate that the acute mid ante-
rior ischemia/infarction resulting from diagonal or
marginal occlusion would be represented by ST de-
viation toward leads –III and –aVF and that the
acute posterolateral ischemia/infarction resulting
from LCX occlusion would be represented by ST
deviation toward leads –V1 and –V2. An alterna-
tive to considering the full 24-lead ECG is consider-
ing ST depression in the 12-lead ECG as STEMI-
equivalent. This concept may be difficult to adopt
clinically, however, because ST depression typically
is considered indicative of ischemia caused by in-
creased metabolic demand rather than ischemia/in-
farction caused by occlusion of coronary flow.
Even occlusion in the most proximal portion
(left main) of the left coronary artery produces
only STEMI-equivalent ECG changes (Fig. 5I).
Indeed, the direction of the ST-segment deviation
is the same as typically occurs with the ischemia of
increased metabolic demand during a positive
stress test. As indicated in the figure, there is usually
a large magnitude of ST depression, and, of course,
the patient is typically hemodynamically unstable.
Omissions from the current diagnostic standards
Important additions to the ECG evaluation of
patients who have acute myocardial ischemia/
infarction would be provided by formal indices
of the three key aspects of this pathologic process:
extent, acuteness, and severity. Indeed algorithms
for each of these aspects have been developed, and
literature regarding their validation has been
accumulating for many years. None of these
indices has yet been incorporated into commercial
diagnostic algorithms, however, and each is too
338 WAGNER et al

Fig. 5. (A–H) Representative 12-lead ECGs for each of the eight regions of ischemia/infarction included in Table 1. In
the order in Table 1, the infarcts are termed (A) anteroseptal, (B) extensive anterior, (C) anterosuperior, (D) posterolat-
eral, (E) anterolateral, (F) inferolateral, (G) extensive inferior, and (H) inferior. (I) More global LV involvement resulting
from main left occlusion is also included.

complex for routine manual clinical application.
Examples of presenting ECGs of patients who
have inferior (Fig. 6) and anterior (Fig. 7) STEMI
are accompanied by the calculated extent (Aldrich
score), acuteness (Anderson-Wilkins score), and
severity (Sclarovsky-Birnbaum grade).
Ischemia/infarction extent
The Aldrich score for application on the
presenting ECG for estimation of the extent of
myocardium at risk for infarction in the absence
of successful reperfusion therapy was introduced
in 1988 [28]. It is based on the slope of the rela-
tionship between the amount in millimeters of
ST segment elevation on the presenting ECG
and the Selvester QRS score on the predischarge
ECG [29] and is expressed as ‘‘% LV infarcted.’’
Fig. 6 presents the use of the Aldrich formula for
estimation of the extent of inferior wall ischemia/in-
farction based on the amount of ST elevation in
leads II, III, and aVF in two representative patients:
% myocardim at risk of infarction
¼ 3½0:6ðsum ST elevation II; III; aVFÞ þ 2:
Fig. 7 presents the use of the Aldrich formula
regarding anteroseptal and anterosuperior wall
involvement:
 PITFALLS IN ECG STANDARDS OF DIAGNOSIS 339

Fig. 6. Two presenting ECGs indicating (A,B) inferolateral ischemia/infarction of varying extent, acuteness, and
severity. The Aldrich extent score, Anderson-Wilkins acuteness score, and Sclarovsky- Birnbaum ischemia grade are in-
dicated beneath each ECG.

developed by Selvester and colleagues of the serial

% myocardim at risk of infarction
¼ 3½1:5ð# of leads with ST elevationÞ  0:4
Ischemia/infarction acuteness
The Anderson-Wilkins score for application on
the presenting ECG for estimation of the acute-
ness of the ischemia/infarction process was in-
troduced in 1995 [30]. It was based on the concept
hyperacute, acute, and subacute phases of the
acute ischemia/infarction process and is based
on the relative amounts of tall T waves and abnor-
mal Q waves in all leads with ST-segment ele-
vation [31]. The Anderson-Wilkins score is
provided as a continuous scale from 4.0 (hyper-
acute) to 1.0 (subacute) based on the comparative
hyperacute T waves versus abnormal Q waves in
each of the leads with ST-segment elevation. The

Fig. 7. Two presenting ECGs indicating (A) anteroseptal and (B) extensive anterosuperior ischemia/infarction of vary-
ing extent, acuteness, and severity. The format of presentation of the three indices is the same as in Fig. 6.
340 WAGNER
lead-specific thresholds for both the abnormally
increased T-wave amplitudes and increased Q-
wave durations are presented in the original refer-
ence [30].
The Anderson-Wilkins score has been calcu-
lated for the four representative patients in Figs. 6
and 7 using the formula:
4ð# leads with tall T yes=abnormal Q noÞ
þ3ð# leads with neither yesÞ
þ2ð# leads with tall T no=abnormal Q yesÞ
þ1ð# leads with both yesÞ
divided by the Total # leads with
ST-segment elevation
The original acuteness score was recently
modified by Heden and colleagues [32] to provide
a similarly even distribution of scores in patients
who have inferior and anterior ischemia/infarc-
tion by reducing the threshold for abnormal
Q-wave duration.
Ischemia/infarction severity
The Sclarovsky-Birnbaum score for applica-
tion on the presenting ECG for estimation of the
severity of the ischemia/infarction process was
introduced in 1990 [33]. It was based on the con-
cept that the severity of the ischemia/infarction
process is determined by the degree of myocardial
protection provided by the combination of collat-
eral vessels and ischemic preconditioning. The
Sclarovsky-Birnbaum score is expressed as grades
1 (least severe), 2 (moderately severe), and 3 (most
severe). Indeed, grade 1 of ischemia is rarely pres-
ent in patients presenting with acute myocardial
ischemia/infarction. Differentiation between
grades 2 and 3 severity requires observation of
each lead with ST elevation for the presence of
‘‘terminal QRS distortion.’’ This distortion is
characterized in leads with a terminal R wave by
a large ST-segment elevation to R-wave amplitude
ratio, and in leads with a terminal S wave by its
total disappearance. A recent study by Billgren
and colleagues [34] documents a high level of in-
terobserver agreement in calculation of the Sclar-
ovsky-Birnbaum score and provides an appendix
with calculation rules.
The Sclarovsky-Birnbaum ischemia grade is
indicated for each of the four representative
patients in Figs. 6 and 7. Indeed for each figure,
examples A were selected for these illustrations
because they have the ST elevation without QRS
et al
distortion typical of grade 2 ischemia; and examples
B were selected because they have the tombstone-
like QRS distortion typical of grade 3 ischemia.
Summary
Pitfalls have been identified in the current
standards for use of the ECG in patients present-
ing with symptoms of acute coronary syndromes
to identify those who have probable acute coro-
nary thrombosis–induced myocardial ischemia/
infarction. Indeed, there are pitfalls in every
aspect of standards for the use of this common
but inexpensive method for diagnosing this com-
mon but critical clinical problem. The acute
ischemia/infarction diagnostic criteria require
identification of two contiguous leads, even
though the classic display fails to present the
limb leads in the same orderly sequence as the
chest leads. These criteria require the presence of
ST elevation even though acute occlusion of
several branches of the coronary arteries, and
even of the main left coronary artery, can produce
only ST depression in the 12 lead ECG. These
criteria have continued to be age and gender
nonspecific until corrected in new standards
documents. There have been inconsistencies in
the terms used to designate the involved myocar-
dial regions, with retention of terms such as
‘‘posterior,’’ even though myocardial imaging
methods have documented more in-vivo correct
alternative terms.
Algorithms for identification of key aspects of
the ischemia/infarction process including extent,
acuteness, and severity that are too detailed for
routine manual application have not been in-
corporated into commercial ECGs, even though
they have been well documented in the literature
for 10 to 20 years. Additional studies are needed
to validate these indices using non-ECG reference
standards. Unfortunately historical time from
symptom onset is not a sufficiently accurate
estimate of the time of the acute coronary occlu-
sion, and formerly available clinical imaging
methods have lacked the precision to serve as
standards for ischemia/infarction extent and se-
verity. As additional studies of these indices
emerge, future working groups of the international
cardiovascular societies will have the opportunity
to consider their adoption as new standards for
ECG evaluation of individuals who have acute
coronary syndromes and who might benefit from
reperfusion before their ischemic myocardium
undergoes irreversible infarction.
 PITFALLS IN ECG STANDARDS OF DIAGNOSIS
Acknowledgment
The authors acknowledge the key contribu-
tions to this document by the AHA Working
Group for Establishing New Electrocardiographic
Standards chaired by Leonard Gettes. There has
been extensive input by the Subgroup on Acute
Ischemia/Infarction, some of whom serve as co-
authors of this article.
References
[1] Diderholm E, Andren B, Frostfeldt G, et al. ST
depression in ECG at entry indicates severe coro-
nary lesions and large benefits of an early invasive
treatment strategy in unstable coronary artery
disease; the FRISC II ECG sub study. The Fast
Revascularization during InStability in Coronary
artery disease. Eur Heart J 2002;23:41–9.
[2] Optimal electrocardiography. 10th Bethesda Confer-
ence. Am J Cardiol 1978;41:111–91.
[3] Surawicz B. Pathogenesis and clinical significance of
primary T wave abnormalities. In: RC Schlant RD,
Hurst W, editors. Advances in electrocardiography.
New York: Grune and Stratton; 1972. p. 377–422.
[4] The Joint European Society of Cardiology/Ameri-
can College of Cardiology Committee. Myocardial
infarction redefinedda consensus document of the
Joint ESC/ACC Committee for the Redefinition of
Myocardial Infarction. JACC2000:36:959–969 and
EHJ2000:21:1502–1513.
[5] Kligfield P, Gettes L, Bailey JJ, et al, and the AHA
writing group. Standards for Electrocardiography,
Part II: The ECG and its Technology. In final
AHA review.
[6] Wagner GS, Wagner NB, White R, et al. The 12 lead
ECG in acute myocardial infarctionAcute coronary
care. 2nd edition. : Mosby-Yearbook, Inc.; 1994.
[7] Lamfers EJP, Schut A, Hertzberger DP, et al.
Prehospital versus hospital fibrinolytic therapy using
automated versus cardiologist electrocardiographic
diagnosis of myocardial infarction: abortion of
myocardial infarction and unjustified fibrinolytic
therapy. Am Heart J 2004;147:509–15.
[8] Taher T, Fu Y, Wagner GS, et al. aborted myo-
cardial infarction in patients with ST segment eleva-
tiondinsights from the ASSENT 3 ECG substudy.
J Am Coll Cardiol 2004;44:38–43.
[9] Reimer KA, Lowe JE, Rasmussen MM, et al. The
wavefront phenomenon of ischemic cell death.
Circulation 1977;56(5):786–94.
[10] Samson WE, Scher AM. Mechanism of ST segment
alteration during acute myocardial injury. Circ Res
1960;8:780–7.
[11] Downar E, Janse MJ, Durrer D. The effect of acute
coronary artery occlusion on subepicardial trans-
membrane potentials in the intact porcine heart. Cir-
culation 1977;56:217–24.
341
[12] Mark DB, Hlatky MA, Lee KI, et al. Localizing
coronary artery obstruction with the exercise tread-
mill test. Ann Intern Med 1987;106:53–5.
[13] Bayés de Luna A, Cino JM, et al. Concordance of
Electrocardiographic Patterns and Healed Myocar-
dial Infarction Location Detected by Cardiovascular
Magnetic Resonance. Am J Cardiol 2006;97:443–51.
[14] White T. Ordningsföljden för extremitets-aclednin-
garna i EKG. Lakartidningen 1971;68:1352–6.
[15] Foster JE, Martin TN, Wagner GS, et al. Determi-
nation of Left Ventricular Long-Axis Orientation
using MRI: changes during the respiratory and car-
diac cycles in normal and diseased subjects. J Clini-
cal Physiology and Functional Imaging 2005;25:
286–92.
[16] Bacharova L, Selvester RH, Engblom H, et al.
Where is the central terminal located? In search of
understanding the use of the Wilson Central Termi-
nal for production of 9 of the standard 12 ECG
leads. JECG 2005;119–27.
[17] Pahlm-Webb U, Pahlm O, Sadanandan S, et al.
A new method for using the direction of ST segment
deviation to localize the site of acute coronary occlu-
sion: the 24-view standard ECG. Am J Med 2002;
113:75–8.
[18] Sadanandan S, Hochman JS, Kolodziej A, et al.
Clinical and angiographic characteristics of patients
with combined anterior and inferior ST segment
deviation on the initial ECG during acute MI. Am
Heart J 2003;146:653–61.
[19] Matetzky S, Freimark D, Chouraqui P, et al. Signif-
icance of ST segment elevations in posterior chest
leads (V7–V9) in patients with acute inferior
myocardial infarction: application for thrombolytic
therapy. J Am Coll Cardiol 1998;31:506–11.
[20] Menown IBA, Allen J, Anderson J, et al. Early
diagnosis of right ventricular or posterior infarction
associated with inferior wall left ventricular acute
myocardial infarction. Am J Cardiol 2000;85:934–8.
[21] Barrabes JA, Figueras J, Moure C, et al. Prognos-
tic significance of ST segment depression in lateral
leads 1, aVL, V5 and V6 on the admission electrocar-
diogram in patients with a first acute myocardial in-
farction without ST segment elevation. J Am Coll
Cardiol 1813;2000:35.
[22] Barrabes JA, Figueras J, Moure C, et al. Prognostic
value of lead aVR in patients with a first non-
ST-segment elevation acute myocardial infarction.
Circulation 2003;108:814.
[23] MacFarlane PW. Age, sex and the ST amplitude
in health and disease. J Electrocardiol 2001;
34(Suppl):235–41.
[24] Macfarlane PW, Lawrie TDV. Appendix I: normal
limits. In: Comprehensive electrocardiology: theory
and practice in health and disease. New York: Perga-
mon Press; 1989. p. 1442–525.
[25] Wagner NB, Wagner GS, White RD. The twelve lead
ECG and the extent of myocardium at risk of acute
infarction: cardiac anatomy and lead locations, and
342 WAGNER
the phases of serial changes during acute occlusion.
In: Califf RM, Mark DB, Wagner GS, editors. Acute
coronary care in the thrombolytic era. Chicago: Year
Book Medical Publishers; 1988. p. 31–45.
[26] Bayes de Luna A, Cino JM, Pujadas S, et al. Concor-
dance of electrocardiographic patterns and healed
myocardial infarction location detected by cardio-
vascular magnetic resonance. Am J Cardiol 2006;
97:443–51.
[27] Cino JM, Pujadas S, Carreras F, et al. Utility of
contrast-enhanced cardiovascular magnetic reso-
nance (CE-CMR) to assess how likely is an infarct
to produce a typical ECG pattern. J Cardiovasc
Magn Reson 2006;8:335–44.
[28] Aldrich HR, Wagner NB, Boswick J, et al. Use of
initial ST segment for prediction of final electrocar-
diographic size of acute myocardial infarcts. Am J
Cardiol 1988;61:749–63.
[29] Hindman NB, Schocken DD, Widmann M, et al.
Evaluation of a QRS scoring system for estimating
myocardial infarct size. V. Specificity and method
et al
of application of the complete system. Am J Cardiol
1985;55:1485–90.
[30] Wilkins ML, Pryor AD, Maynard C, et al. An elec-
trocardiographic acuteness score for quantifying the
timing of a myocardial infarction to guide decisions
regarding reperfusion therapy. Am J Card 1995;75:
617–20.
[31] Anderson ST, Wilkins M, Weaver WD, et al.
Electrocardiographic phasing of acute myocardial
infarction. J Electrocardiol 1993;25(Suppl):3–5.
[32] Heden B, Ripa R, Persson E, et al. A modified
Anderson-Wilkins ECG acuteness score for anterior
or inferior myocardial infarction. Am Heart J 2003;
146:797–803.
[33] Sclarovsky S, Mager A, Kusniec J, et al. Electrocar-
diographic classification of acute myocardial ische-
mia. Isr J Med Sc 1990;26 535–533.
[34] Billgren T, Birnbaum Y, Sgarbossa EB, et al. Detailed
definition and intraobserver agreement for the elec-
trocardiographic Sclarovsky-Birnbaum ischemia
grading system. JECG 2002;35(Suppl):201–2.
 Cardiol Clin 24 (2006) 343–365

Electrocardiographic Diagnosis of ST-elevation

Myocardial Infarction
Shaul Atar, MD, Alejandro Barbagelata, MD,
Yochai Birnbaum, MD*
Division of Cardiology, University of Texas Medical Branch, 5.106 John Sealy Annex,
301 University Boulevard, Galveston, TX 77555, USA

The ECG is the most useful and feasible
diagnostic tool for the initial evaluation, early
risk stratification, triage, and guidance of therapy
in patients who have chest pain. There is currently
a growing trend for 12-lead ECGs to be recorded
in the field by paramedics and transmitted by
cellular telephone or fax to the target emergency
department. It is conceivable that emergency
department physicians will be involved in triaging
patients in the prehospital phase to hospitals
offering primary percutaneous coronary interven-
tion (PCI), which is now recognized to be a supe-
rior reperfusion strategy than thrombolytic
therapy [1].
Patients who have ST-segment elevation or
new left bundle-branch block are usually referred
for immediate reperfusion therapy, whereas those
who do not have ST-segment elevation are being
treated initially with medications [2,3]. Patients
are diagnosed as having anterior, inferior-
posterior, or lateral myocardial infarction based
on the patterns of ST deviation, and assessment
of risk is usually based on simple crude measure-
ments of the absolute magnitude of ST-segment
deviation or the width of the QRS complexes [4].
Much more information concerning the exact
site of the infarct related lesion, prediction of final
infarct size, and estimation of prognosis can be
obtained from the admission ECG without extra
cost or time. Although some clinicians believe
that with the increased use of primary PCI in
* Corresponding author.
E-mail address: yobirnba@utmb.edu (Y. Birnbaum).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.008
patients who have ST-elevation myocardial
infarction (STEMI) this information is no longer
needed, there are many instances in which, even
with immediate coronary angiography, identifica-
tion of the infarct-related site and estimation of
the myocardial area supplied by each of the
branches distal to the coronary artery occlusion
is difficult. In some patients, more than one occlu-
sive lesion may be found, and identification of the
acutely thrombosed lesion may not always be
apparent. In other cases, total occlusion of side
branches at bifurcation of coronary arteries may
be missed during coronary angiography.
It is important to appreciate that the ECG
provides information about a totally different
aspect of pathophysiology in STEMI than does
the coronary angiogram. Coronary angiography
identifies vessel lumen anatomy, whereas the ECG
reflects the physiology of the myocardium during
acute ischemia. For this reason, it is possible to
observe severe coronary stenoses on angiography
without ECG evidence of acute ischemia. On the
other hand, it is possible to observe restored vessel
patency while the ECG continues to show signs of
ongoing ‘‘ischemia’’ or ‘‘injury pattern’’ caused by
the no-reflow phenomenon, reperfusion injury, or
myocardial damage that has already developed
before reperfusion occurs. Thus, although coro-
nary angiography remains the reference standard
for identifying the infarct-related artery, the ECG
remains the reference standard for identifying the
presence, location, and extent of acute myocardial
ischemia and injury. Moreover, with current
imaging techniques, including contrast ventricu-
lography, echocardiography, and radionuclide
cardiology.theclinics.com
344 ATAR et al

perfusion imaging, differentiation of ischemic but
still viable myocardium from necrotic myocar-
dium during the acute phase of STEMI is not
feasible, but such differentiation may be possible
with a correct interpretation of the ECG.
Nevertheless, several conditions other than
STEMI may present with ST elevation and need
immediate recognition to avoid false treatment [5].
These conditions include left ventricular hypertro-
phy with secondary repolarization abnormalities,
early repolarization pattern, Prinzmetal’s angina,
chronic left ventricular aneurysm, acute pericardi-
tis, pulmonary embolism, and the Brugada syn-
drome. Moreover, studies have pointed out that
in a healthy population, more than 90% of men
between the ages of 16 and 58 years have ST-
segment elevation of 1 to 3 mm in one or more
of the precordial leads, mainly in lead V2 [6].
The prevalence of these changes declined with
age, reaching 30% in men over the age of 76 years,
whereas in women the prevalence is only 20% and
is constant throughout the ages [7].
Some of the patients who have acute chest pain
and ST elevation may subsequently have an
increase in cardiac markers without any further
ECG changes (no ST resolution, no new Q-wave
development, and no T-wave inversion). These
patients may have non-STEMI with baseline ST
elevation, or pseudo STEMI, as shown in Fig. 1.
On the other hand, there are patients who have
transient ST elevation who have ST resolution
without an increase in the cardiac markers. Al-
though some of them may have acute pericarditis,
Prinzmetal’s angina, or aborted myocardial in-
farction [8], many simply have transient early re-
polarization or pseudo-pseudo STEMI. This last
entity has not been well characterized. The clini-
cian encountering a patient who has suggestive
symptoms and ST elevation must make rapid
therapeutic decisions concerning urgent revascu-
larization without waiting for the results of car-
diac markers. It is currently uncertain how many
patients who had pseudo or pseudo-pseudo
STEMI have been included in randomized trials
of STEMI. For example, in a recent analysis of
the Hirulog and Early Reperfusion or Occlusion
(HERO)-2 trial, 11.3% of the patients who had
ST-segment elevation who received reperfusion
therapy did not have enzymatically confirmed
myocardial infarction [9]. Increasing the threshold
for ST elevation (ie, 2 mm in the precordial leads)
may decrease the occurrence of false-positive cases
but may result in reduced sensitivity and underuse
of reperfusion therapy. Comparison with previous
ECG recordings or repeating ECG recordings for
evolution of subtle changes, along with selective
use of echocardiography, may enable the clinician
to increase the accuracy of diagnosing true

Fig. 1. ECG of a 57-year-old diabetic and hypertensive man admitted with chest pain of 3 hours’ duration. ECG shows
left ventricular hypertrophy and ST elevation in the anterolateral leads with reciprocal changes in the inferior leads. Sub-
sequently his creatine kinase muscle-brain increased to 28.9 pg/mL, and his troponin I was positive. His ECG 3 years
before admission showed the same pattern of ST elevation. Repeat ECGs during current admission did not show reso-
lution of ST elevation, T-wave inversion, or development of new Q waves. The diagnosis is pseudo STEMI.
 ECG DIAGNOSIS OF STEMI
STEMI that may benefit from urgent reperfusion
therapy.
This article concentrates on the information
that can be obtained from the admission ECG in
patients who have STEMI. In particular, it
discusses the association of various ECG patterns
of the acute phase of STEMI with estimation of
infarct size and prognosis and the correlation of
various ECG patterns with the underlying coro-
nary anatomy.
Factors that determine prognosis
in ST-elevation myocardial infarction
The immediate prognosis in patients who have
STEMI is inversely related to the amount of
myocardial reserves (total myocardial mass less
the myocardium involved in the acute STEMI
[ischemic area at risk], scarred territories caused
by previous myocardial infarction or fibrosis, and
remote ischemic myocardial segments supplied by
critically narrowed coronary arteries). Among
patients who do not have prior myocardial in-
farction or major pre-existing stenotic lesions in the
coronary arteries, prognosis is directly related to
the size of the ischemic myocardium supplied by the
culprit coronary artery distal to the occlusion. In
patients who have low myocardial reserves because
of previous myocardial infarctions or diffuse fibro-
sis, however, even a relatively small infarction may
be detrimental. Moreover, in patients who have
diffuse severe coronary artery disease, a small
myocardial infarction may interfere with the deli-
cate balance and induce remote ischemia by oblit-
erating collateral flow or creating a need for
(compensatory) augmentation of contractility in
the remote noninfarcted segments supplied by
stenosed coronary arteries. Therefore, in addition
to accurate diagnosis, there is a need for early
estimation of the size of the ischemic myocardium
at risk and myocardial reserves.
Fig. 2. The grades of ischemia in leads with baseline qR configuration (upper row) and leads with baseline rS configu-
ration (lower row).
345
The ECG may help in assessing the size of the
ischemic myocardial area at risk, may help in
differentiation between subendocardial (nontrans-
mural) and transmural ischemia, and may assist in
identifying the presence of previous infarctions
(abnormal Q waves in leads not involved in the
present infarction; for example, abnormal Q waves
in the precordial leads in a patient who has inferior
ST elevation) [10]. Furthermore, some ECG pat-
terns may indicate the presence of diffuse coronary
artery disease and remote ischemia [11–13].
On admission, part of the myocardial area at
risk (supplied by the culprit coronary artery)
might have already undergone irreversible dam-
age. The proportion of the ischemic area at risk
that has undergone irreversible necrosis depends
on the total ischemic time and on the rate of
progression of the wave front of necrosis. The rate
of progression of necrosis is highly variable and is
dependent on the presence of residual perfusion
through collateral circulation [14] or incomplete
or intermittent occlusion of the infarct related
artery [15], as well as on metabolic factors includ-
ing ischemic preconditioning [16].
ECG changes during the acute phase
of ST-elevation myocardial infarction
Shortly after occlusion of a coronary artery,
serial ECG changes are detected by leads facing
the ischemic zone, as shown in Fig. 2. First, the
T waves become tall, symmetrical, and peaked
(grade 1 ischemia); second, there is ST elevation
(grade 2 ischemia) without distortion of the termi-
nal portion of the QRS; and third, changes in the
terminal portion of the QRS complex may appear
(grade 3 ischemia) [17–19]. The changes in the
terminal portion of the QRS are explained by
prolongation of the electrical conduction in the
Purkinje fibers in the ischemic region. The delayed
conduction decreases the degree of cancellation,
346 ATAR et al

resulting in an increase in R-wave amplitude in
leads with terminal R wave and a decrease in
the S-wave amplitude in leads with terminal
S wave on the surface ECG [20,21]. The Purkinje
fibers are less sensitive to ischemia than the con-
tracting myocytes [22]. Hence, for an alteration
in the terminal portion of the QRS to occur, there
probably should be a severe and prolonged ische-
mia that would affect the Purkinje fibers [23]. In
patients who have collateral circulation, no
changes are detected in the QRS complex during
balloon angioplasty [20]. Thus, absence of distor-
tion of the terminal portion of the QRS, despite
prolonged ischemia, may be a sign for myocardial
protection (probably by persistent myocardial
flow caused by subtotal occlusion or collateral cir-
culation or by myocardial preconditioning). The
disappearance of the S waves in leads with termi-
nal S (RS configuration), mainly leads V1to V3,
can be recognized easily (Fig. 3). In contrast, the
absolute R-wave height is influenced by many
other variables. Therefore, the absolute R-wave
amplitude is not helpful in determining the sever-
ity of ischemia. Changes in the R-wave amplitude
can be detected reliably only by continuous ECG
monitoring; comparison of the admission ECG
with previous ECG recordings often is difficult be-
cause of differences in ECG instruments and dif-
ferent placement of the precordial electrodes.
Therefore, a second empiric criterion for leads
with terminal R configuration was developed.
This criterion relates the J point of the ST to the
R-wave amplitude in leads with terminal R waves
(qR configuration). As shown in Fig. 4, a ratio of
0.5 or greater indicates grade 3 ischemia [17,19].
Although the transition between the grades of is-
chemia is gradual and continuous, for practical
clinical purposes it is convenient to define grade
2 ischemia as ST elevation greater than 0.1 mV
without distortion of the terminal portion of the
QRS and grade 3 as ST elevation with distortion
of the terminal portion of the QRS (emergence

Fig. 3. (A) Grade 2 ischemia in a patient who has acute anterior STEMI. There are deep S waves in leads V1 to V3.
(B) Grade 3 ischemia in a patient who has acute anterior STEMI. There is loss of S waves in leads V1 to V3, and the
J point/R wave ratio is greater than 0.5 in leads V4 and V5.
 ECG DIAGNOSIS OF STEMI 347

Fig. 4. (A) Grade 2 ischemia in a patient who has acute inferolateral STEMI. The J point:R wave ratio is less than 0.5.
(B) Grade III ischemia in a patient who has acute inferior STEMI. There is first-degree atrioventricular block, and the J
point:R wave ratio is greater than 0.5 in leads II, III, and aVF.

of the J point O50% of the R wave in leads with
qR configuration, or disappearance of the S wave
in leads with an Rs configuration (see Figs. 3, 4)
[17,24–28]. Only later, the T waves become nega-
tive, the amplitude of the R waves decreases,
and Q waves may appear. Only a minority of
patients who have STEMI presents with grade 3
ischemia upon admission. Although the underly-
ing mechanism for this difference is still unclear,
grade 3 ischemia has large implications regarding
prognosis, as discussed later.
Diagnosis of acute ST-elevation myocardial
infarction
In a patient who has typical symptoms, the
presence of ST-segment elevation, especially when
accompanied with reciprocal changes, is highly pre-
dictive of evolving STEMI. Several investigators,
however, reported that the sensitivity of the ECG
for acute myocardial infarction may be as low as
50% [29–31]. In most of these studies only one ad-
mission ECG was analyzed. Hedges and col-
leagues [32] used the admission and a second
ECG performed 3 to 4 hours after admission
and found serial ECG changes in 15% of the pa-
tients. Continuous ECG monitoring or multiple
ECG recordings over time or during fluctuations
in the intensity of symptoms were not performed,
however. Repeated ECG recordings may improve
the ability to detect subtle ischemic changes. Fur-
thermore, as determined by independent re-
viewers, 49% of the missed acute myocardial
infarctions could have been diagnosed through
improved ECG-reading skills or by comparing
the current ECG with a previous one [30]. It
should be remembered that acute myocardial
infarction detected by elevated creatine kinase
muscle-brain (MB) or troponin levels without
ST elevation is not an indication for urgent
348 ATAR
reperfusion therapy. The only exception is new
left bundle-branch block in a patient who has
acute chest pain. Menown and colleagues [33]
studied the sensitivity and specificity of the admis-
sion ECG for diagnosing acute myocardial infarc-
tion by studying patients who had (n ¼ 1041) or
did not have (n ¼ 149) chest pain. Acute myocar-
dial infarction was defined by the presence of
chest pain of 20 minutes’ duration or longer, ele-
vation of creatine kinase two times or more the
upper laboratory normal reference level (creatine
kinase-MB activity R7% if the etiology of the to-
tal creatine kinase was equivocal), or elevation of
creatine kinase less than two times the upper lab-
oratory normal reference level accompanied by se-
rial ECG changes consistent with new myocardial
infarction (new Q waves R 0.03 seconds’ duration
or new persistent T-wave inversion in two or more
contiguous leads). The best ECG variables for the
diagnosis of acute myocardial infarction were ST
elevation greater than 0.1 mV in more than one
lateral or inferior lead or ST elevation greater
than 0.2 mV in more than one anteroseptal precor-
dial lead. These criteria correctly classified 83% of
subjects, with a sensitivity of 56% and a specificity
of 94%. Changing the degree of ST elevation sig-
nificantly modified both the sensitivity
(45%–69%) and the specificity (81%–98%). The
addition of multiple QRST variables (Q waves,
ST depression, T-wave inversion, bundle-branch
block, axes deviations, and left ventricular hyper-
trophy) increased specificity but only marginally
improved overall classification [33].
Estimation of the size of ischemic
myocardium at risk
There is no currently defined reference stan-
dard to measure the ischemic area at risk in the
acute setting. The extent of regional wall motion
abnormalities can be appreciated easily soon after
admission by two-dimensional echocardiography
or left ventriculography. With both methods,
however, a differentiation between old scars and
the acutely ischemic but viable zones is not always
possible. Because of the effect of stunning, re-
gional wall motion may persist for long periods of
time after reperfusion has occurred [34]. More-
over, differentiation of transmural from subendo-
cardial ischemia/infarction is not always possible,
because akinesis may occur when only the inner
myocardial layers are ischemic [35].
Several studies have tried to estimate the
ischemic area at risk or final infarct size by the
et al
admission ECG. In these studies, either the
number of leads with ST deviation (elevation or
depression) [36–39] or the absolute amplitude of
ST deviation [4,36,39–41] was used. The results
were conflicting, however. Arnold and Simoons
[4] have evaluated the ‘‘expected infarct size with-
out thrombolysis’’ by multivariate regression
analysis in 885 patients in the rt-PA/placebo and
rt-PA/PTCA trial conducted by the European Co-
operative Study Group and validated the findings
in 533 patients from the Intracoronary Streptoki-
nase trial of the Interuniversity Cardiology Insti-
tute of The Netherlands and in 1741 patients
from the Intravenous Streptokinase in Acute
Myocardial Infarction study; both trials con-
tained a nonthrombolyzed control group. They
defined an infarct size score that included the ab-
solute sum of the amplitude of ST deviation, the
QRS width exceeding 0.12 seconds, anterior
STEMI location, Killip class 3 or 4 on admission,
and delay from symptom onset to treatment allo-
cation and found that the expected infarct size
correlated well with the actual enzymatic infarct
size in the nonthrombolyzed patients of the latter
two studies. Limitation of infarct size by throm-
bolytic therapy was greatest in patients who had
a large expected infarct size and was absent in
patients who had a small expected infarct size.
Similarly, 1-year mortality reduction was greatest
in patients who had a large expected infarct size
without thrombolysis.
Aldrich and colleagues [36] studied 144 pa-
tients who had STEMI who did not receive
thrombolytic therapy. The best correlation be-
tween the final ECG Selvester QRS scoring system
(an estimation of infarct size) and the admission
ECG was found using the magnitude of ST eleva-
tion in leads II, III, and aVF in inferior STEMI
and the number of leads with ST elevation in an-
terior STEMI. Another study in patients who re-
ceived reperfusion therapy showed only a weak
correlation between the Aldrich score and either
the ischemic area at risk or final infarct size, as
measured by pretreatment and predischarge tech-
netium-99m (Tc-99m) sestamibi scans, respec-
tively [38]. The Aldrich formula was related
more to the collateral score than to the ischemic
area at risk or final infarct size [38]. Clemmensen
and colleagues [37] reported a good correlation
between the final Selvester score and the number
of leads with ST elevation (r ¼ 0.70) in anterior
STEMI. Neither the magnitude of ST segment el-
evation in all leads nor the number of leads with
ST elevation correlated with the final Selvester
 ECG DIAGNOSIS OF STEMI
score in inferior STEMI. Clements and colleagues
[39] also reported only a weak correlation between
myocardial area at risk (as assessed by Tc-99m
sestamibi scan) and the number of leads with ST
deviation, total ST deviation, total ST elevation,
or total ST depression. The myocardial area at
risk correlated modestly (r ¼ 0.58) with total ST
deviation in anterior STEMI and with total ST
depression normalized to the R wave (r ¼ 0.70)
in inferior STEMI. Because of large standard er-
rors (9%–15% of the left ventricle), however,
these formulas for estimation of the myocardial
area at risk cannot be used in the clinical setting
for estimation of infarct size [39]. Birnbaum and
colleagues [42] showed that among patients with
first anterior STEMI, the correlation between ei-
ther the number of leads with ST elevation or
the sum of ST elevation and the extent and sever-
ity of regional left ventricular dysfunction (both at
90 minutes after initiation of thrombolytic ther-
apy and at predischarge) was poor.
These ECG studies were based on the assump-
tion that each lead represents the same amount of
myocardium and that a similar size of ischemic
area in different locations of the left ventricle will
result in similar magnitude of ST deviation in the
same number of leads. The 12-lead ECG does not
represent all myocardial regions equally, however.
The inferior and anterior walls of the left ventricle
are well represented, but the lateral, posterior,
septal, and apical regions are relatively ECG silent
[43,44]. Moreover, ischemia in opposed regions
may attenuate or augment ST deviation. For ex-
ample, in patients who have ischemia of the high
anterolateral and inferior regions caused by prox-
imal occlusion of a dominant left circumflex coro-
nary artery (LCX), attenuation of ST deviation in
leads I, aVL, and the inferior leads may occur,
whereas subendocardial high anterolateral ische-
mia may augment ST-segment elevation in the in-
ferior leads. Posterior myocardial infarction is
commonly associated with ST depression in the
precordial leads V1 to V3 [45], whereas right ven-
tricular infarction may cause ST elevation in leads
V1 and V2 [46]. In concomitant right ventricle and
posterior myocardial infarction, the opposing
forces may neutralize each other, and therefore,
no ST deviation may occur in these leads. Because
different leads represent different areas of the
myocardium, a different coefficient probably
should be used for each lead and even for each
type of infarction. To overcome the unequal rep-
resentation of the myocardium by the different
leads, another technique has been suggested
349
[47,48]. In this technique the maximal points of
the Selvester QRS score are given to each lead
with ST elevation greater than 0.1 mV. The sum
of these initial scores is considered to represent
the percentage of the left ventricle that is ischemic.
This method was compared with thallium-201
perfusion scans in 28 patients (10 patients on ad-
mission and 18 patients on day 5 after reperfusion
therapy) [47]. A good correlation was found be-
tween this potential Selvester score and the extent
of thallium-201 perfusion defect (r ¼ 0.79; P !
.005). Birnbaum and colleagues [42] found only
a weak correlation between the maximal potential
Selvester score and the extent or severity of left
ventricular dysfunction among patients who had
first anterior STEMI and underwent left ventricu-
lography at 90 minutes after initiation of throm-
bolytic therapy and at predischarge. On the
other hand, in patients receiving streptokinase
for STEMI, Wong and colleagues [49] found
that the initial Selvester QRS score and T-wave
inversion grade were the only predictors of myo-
cardial salvage (P ! .001), with no difference
between anterior and nonanterior STEMI [49].
Another qualitative approach for predicting
final infarct size by the admission ECG based on
the grades of ischemia has been reported by
Birnbaum and colleagues [26,27,42,50]. In the
Thrombolysis in Myocardial Infarction–4
(TIMI-4) trial, patients presenting with grade 3
ischemia (n ¼ 85) on admission had a larger in-
farct as assessed by creatine kinase release over
24 hours (P ¼ .023), and a larger predischarge
Tc-99m sestamibi defect size (P ¼ .001) [26]. In
a comparison of patients with first anterior
STEMI who were assigned randomly to thrombo-
lytic therapy or conservative treatment, the final
QRS Selvester score was lowered by thrombolytic
therapy only in patients who had grade 2 ischemia
on enrollment, but not grade 3 [27]. Overall, final
QRS Selvester score was higher for patients who
had grade 3 than grade 2 ischemia on enrollment,
both in those who received and in those who did
not receive thrombolytic therapy. Among patients
with a first anterior STEMI who participated in
the Global Use of Streptokinase and t-PA for
Occluded Coronary Arteries (GUSTO)-I angio-
graphic substudy and underwent angiography
both at 90 minutes after initiation of thrombolytic
therapy and at predischarge, patients who had
grade 2 ischemia on enrollment had higher left
ventricular ejection fraction at 90 minutes than
patients who had grade 3 ischemia [42]. The differ-
ence in global left ventricular function was related
350 ATAR
mainly to the severity of regional dysfunction in
the involved segments and less to the extent of in-
volvement (size of the area at risk) [42]. On predis-
charge evaluation, the grade 3 group tended to
have lower left ventricular ejection fraction and
had significantly more chords with dysfunction
and more severe regional dysfunction than the
grade 2 group. The number of dysfunctional
chords tended to decrease from 90 minutes to pre-
discharge in the grade 2 group, whereas it tended
to increase in the grade 3 group. This finding may
reflect partial recovery from stunning at the pre-
discharge ventriculography in the grade 2 group
[42]. There was no difference in the time to ther-
apy or the success of thrombolysis between the
grade 2 and 3 groups. Thus, it seems that the dif-
ference in infarct size between the grade 2 and
grade 3 groups is explained by more severe ische-
mia and not by larger ischemic area at risk, longer
ischemia, or lower rates of successful reperfusion
[42]. Findings from the Acute Myocardial Infarc-
tion Study of Adenosine trial confirmed this hy-
pothesis [50]. In this study, patients who received
thrombolytic therapy were assigned randomly to
pretreatment with intravenous adenosine or pla-
cebo. For placebo-treated patients, the median
pretreatment Tc-99m sestamibi single-photon-
emission CT (SPECT) perfusion defect (ischemic
area at risk) did not differ significantly between
grade 2 and grade 3 patients; however, the median
infarction index (infarct size/area at risk) was
smaller in patients who had grade 2 ischemia
(66% versus 90%; P ¼ .006). Overall, infarct
size was related to baseline ischemia grade (P ¼
.0121) and was reduced by adenosine treatment
(P ¼ .045). In a recently published study, Billgren
and colleagues [51] have assessed the relation of
baseline ECG ischemia grades to area at risk
and myocardial salvage [100(area at risk
infarct
size)/area at risk] in 79 patients who underwent
primary PCI for first STEMI and had technetium
Tc-99m sestamibi SPECT before PCI and predis-
charge final infarct size. Patients were classified as
having grade 2 ischemia (ST elevation without ter-
minal QRS distortion in any of the leads; n ¼ 48),
grade 2.5 ischemia (ST elevation with terminal
QRS distortion in 1 lead, n ¼ 16), or grade 3 is-
chemia (ST elevation with terminal QRS distor-
tion in R two adjacent leads; n ¼ 15). Time to
treatment and area at risk were comparable
among groups. Myocardial salvage, as a percent-
age of the area at risk, tended to be smaller for
the grade 3 ischemia group than in the other
two groups (65%, 65%, and 45%, for grade 2,
et al
2.5, and 3, respectively; P ¼ .16). Salvage was de-
pendent on time only in the grade 3 group. The
authors conclude that patients who have grade 3
ischemia have rapid progression of necrosis over
time and less myocardial salvage, and that grade
3 ischemia is a predictor of smaller myocardial
salvage by primary PCI.
Most, but not all, studies have shown that, as
a group, patients who have inferior STEMI and
ST depression in leads V1–V3 have larger infarcts
than their counterparts who do not have ST de-
pression, as evidenced by higher peak creatine ki-
nase release; more extensive wall motion
abnormalities; larger defect size by thallium-201,
Tc-99m, and positron emission tomography; and
higher QRS scores of infarct size [13,52–57]. In
addition, ST elevation in lead V6 in patients who
have inferior STEMI also is associated with larger
myocardial infarction [58]. Others have reported
that ST elevation in the posterior leads (V7–V9)
in patients who have inferior STEMI is associated
with larger infarct size [59]. It is unclear whether
there is additive value for ST elevation in leads
V7 to V9 over V6 alone.
Many variables, such as the width of the chest
wall, the distance of the electrode from the
ischemic zone, the myocardial mass, and the
presence of ischemic preconditioning and collat-
eral circulation, have a major influence on the
absolute magnitude of ST deviation. Therefore,
currently, there is no accurate method for esti-
mating the area at risk by the admission ECG that
can be used in the individual patient, although, in
general, patients who have ST deviation (elevation
plus depression) in large number of leads or high
absolute sum of ST deviation have a larger
myocardial infarction than patients who have ST
deviation in a small number of leads or low sum of
ST deviation [4,40]. There are patients who have
a relatively large infarction who have only minor
absolute ST deviation. Infarct size underestima-
tion in these patients may lead to underuse of re-
perfusion therapy. The grades of ischemia predict
final infarct size but not the size of the ischemic
area at risk.
Differentiation between viable and necrotic
myocardium at the ischemic area at risk
Although with echocardiography old myocar-
dial scars with thinning of the ventricular wall and
dense echo reflections can be identified, none of
the direct imaging modalities (contrast ventricu-
lography and echocardiography) can differentiate
 ECG DIAGNOSIS OF STEMI
between ischemic but viable myocardium and
myocardium that has already undergone irrevers-
ible necrosis in the acute stage of infarction.
Q waves were traditionally considered as a sign
of myocardial necrosis [60]. The mechanism and
significance of Q waves that appear very early in
the course of STEMI in leads with ST elevation
are probably different, however [60–63]. Fifty-
three percent of the patients who had STEMI ad-
mitted within 1 hour of onset of symptoms had
abnormal Q waves on presentation, even before
reperfusion therapy had been initiated [61]. It
has been suggested that Q waves that appear
within 6 hours from onset of symptoms do not
signify irreversible damage and do not preclude
myocardial salvage by thrombolytic therapy [64].
Furthermore, Q waves that appear early in the
course of acute ischemia may be transient and dis-
appear later [64,65]. Several authors have found
early Q waves to be associated with larger ische-
mic zone and ultimate infarct size [61,66]. Such
Q waves have been explained by a transient loss
of electrophysiologic function caused by intense
ischemia [60,64]. In contrast, some investigators
found that Q waves develop rapidly only after re-
perfusion [62,63,67]. It has been suggested that the
presence of Q waves may be masked by the injury
current during ischemia [63], and they frequently
can be seen only after resolution of the injury cur-
rent. These changes, however, may reflect reperfu-
sion injury, interstitial edema, or hemorrhage that
later may resolve partially [67]. Ninety minutes
after thrombolytic therapy, however, TIMI flow
grade 3 is achieved less often in patients with
than without abnormal Q waves on presentation
[68]. Further studies are needed to find other
ECG markers that will assist in differentiation
Fig. 5. Admission ECG of a patient who has anterolateral STEMI, showing ST elevation with deep inverted T waves.
351
between acutely ischemic but viable from irrevers-
ible necrotic myocardium. Regardless of the un-
derlying mechanism, the presence of abnormal Q
waves in the leads with ST elevation on the admis-
sion ECG is associated with larger final infarct
size and increased in-hospital mortality [69].
Patients who have grade 3 ischemia on enroll-
ment have larger final infarct size [26,50] but not
larger initial ischemic area at risk [42,50]. In addi-
tion, patients who have grade 3 ischemia are less
likely to benefit from thrombolytic therapy than
patients who have grade 2 ischemia [27]. Distor-
tion of the terminal portion of the QRS complex
in leads with ST elevation is not a sign that irre-
versible damage had already occurred upon pre-
sentation, however, because the same ECG
pattern frequently is detected in patients who
have Prinzmetal’s angina during ischemic episodes
that are not associated with significant myocardial
damage.
Some patients who have STEMI present with
negative T waves (Fig. 5). Early inversion of the T
waves, along with resolution of ST elevation, is
a sign of reperfusion [70]; however, the signifi-
cance of negative T waves in leads with ST eleva-
tion before reperfusion therapy is initiated is
currently unclear. Wong and colleagues [68] re-
ported that 90 minutes after thrombolytic ther-
apy, TIMI flow grade 3 was seen less often in
patients who presented with ST elevation and neg-
ative T waves than in those who had positive T
waves. In a retrospective analysis of the HERO-1
study [49], the same authors found on multivariate
analysis that the T-wave inversion grade (graded
according to the depth and location of T-wave in-
version) on admission in patients with first
STEMI was the strongest predictor of less
352 ATAR
myocardial salvage (r ¼ 0.57; P ! .001). Herz and
colleagues [71] reported that among patients
treated 2 to 6 hours after onset of symptoms,
those who presented with inverted T waves in
leads with ST elevation had higher in-hospital
mortality than patients with positive T waves. In
contrast, among patients treated within the first
2 hours of onset of symptoms, patients who had
negative T waves had no hospital mortality (0/52
patients), as compared with a 5.0% mortality
rate in patients who had positive T waves (36/
726 patients; P ¼ .19) [71]. Therefore, ST eleva-
tion with negative T waves, especially if it occurs
in patients presenting more than 2 hours of onset
of symptoms, may be used as an ECG sign of
a more advanced stage of infarction or presence
of irreversible damage, associated with lesser
chance of achieving successful reperfusion, and
subsequently leading to higher mortality.
Expected rate of progression of myocardial
infarction
It currently is impossible to assess the severity
of myocardial ischemia or the expected rate of
progression of myocardial necrosis by direct
myocardial imaging. The magnitude of ST eleva-
tion reflects mainly the severity of the subepicar-
dial ischemia. The standard surface 12-lead ECG
is less sensitive to subendocardial ischemia. Sub-
endocardial ischemia may cause either ST de-
pression or no change in the ST segment. ST
depression, however, also may result from re-
ciprocal changes in leads oriented away from the
ischemic zone [72]. Augmentation of collateral
flow ameliorates the magnitude of ST deviation
during coronary balloon occlusion [73]. More-
over, ischemic preconditioning by preceding brief
ischemic episodes attenuates the magnitude of ST
deviation [73,74]. Data regarding the effects of
myocardial preconditioning or the presence of
sufficient collateral circulation on the 12-lead
ECG during acute myocardial infarction are
sparse, however. Collateral circulation reduces
the severity of the subepicardial ischemia and
hence attenuates ST elevation [38]. Indeed, Sagie
and colleagues [75] showed that in patients who
had acute anterior STEMI and who had good col-
lateral circulation, only T-wave changes, without
ST elevation (grade 1 ischemia) were observed.
Other cardiac and noncardiac variables, such as
presence of myocardial hypertrophy, the distance
of the heart from the chest wall, or the width of
the chest wall, may also affect the magnitude of
et al
ST deviation. Therefore, the absolute magnitude
of ST deviation can give only a rough estimation
of the magnitude of myocardial protection or
the severity of ischemia.
Identifying ischemia at a distance
Patients who have ST elevation in one territory
often have ST depression in other territories. The
additional ST deviation may represent ischemia in
a myocardial region other than the area of
infarction or may represent pure reciprocal
changes. There is abundant literature on the
significance of different types of ST depression
during STEMI [72]. Most of the common patterns
of remote ST depression probably represent recip-
rocal changes and not ischemia at a distance. In
anterior STEMI, ST depression in the inferior
leads is reciprocal to involvement of the basal an-
terolateral region, supplied by the first diagonal
branch and represented by ST elevation in leads
I and aVL [76,77]. In patients who have inferior
STEMI, ST depression in lead aVL is a pure recip-
rocal change and is found in almost all patients
[78], and ST depression in leads V1 to V3 probably
does not represent ischemia at a distance but
rather reciprocal changes caused by more poste-
rior, inferoseptal, apical, or lateral left ventricular
involvement [53,55,56]. In contrast, among pa-
tients who have inferior STEMI, ST depression
in leads V4 to V6 is associated with concomitant
left anterior descending (LAD) coronary artery
stenosis or three-vessel disease [11,13,79]. Thus,
presence of an atypical pattern of ST depression
and especially ST depression in leads V4 to V6
in inferior STEMI may signify ischemia at
a distance.
In special circumstances both types of ST
depression may be present. In STEMI caused by
occlusion of the first diagonal branch, in addition
to ST elevation with positive T waves in leads aVL
and V2, there usually is reciprocal ST depression
with negative T waves in the inferior leads (pure
mirror image) and a different pattern of ST de-
pression with tall, peaked T waves (subendocar-
dial ischemia) in V4 and V5; the ST segment in
lead V3 is either isoelectric or depressed [77].
Boden and colleagues [80] reported that in pa-
tients who have non-STEMI, isolated ST-segment
depression in leads V1 to V4 was more likely to
caused by posterior wall STEMI (reciprocal
changes) when it was associated with upright T
waves; it was caused by anterior subendocardial
ischemia when the T waves were negative., Porter
 ECG DIAGNOSIS OF STEMI
and colleagues [81], however, found that the
polarity of the T waves in the precordial leads
with ST depression cannot be used to differentiate
between the two etiologies of ST depression. In
many patients who have inferior STEMI and ST
depression in leads V1 to V3, the associated T
waves in these leads are negative initially (a recip-
rocal image of ST elevation with positive T waves
in leads facing the infarct zone), and only later do
the T waves become positive and the R-wave am-
plitude increase (a reciprocal image of inversion of
the T waves and development of Q waves in leads
facing the infarction).
Identification of the exact site
of the infarct-related artery
The admission ECG, by suggesting the loca-
tion of the ischemic area at risk, may assist in
identificating the exact site of coronary artery
occlusion. Because of variability in the coronary
anatomy, in some instances there may be more
than one possible explanation for a specific ECG
pattern. Moreover, because the size and exact
location of the vascular bed supplied by the
occluded artery varies considerably, occlusion in
the same site of a coronary artery in different
patients may result in a different size and location
of the ischemic area at risk and hence different
ECG changes. In addition, presence of severe pre-
existing narrowing in a nonculprit coronary artery
may cause ischemia at a distance that may alter
the classic ECG picture. Much of the work that
studied the correlation between various ECG
patterns and the site of the culprit lesion included
only patients who had single-vessel disease and
a first myocardial infarction; thus the applicability
of these criteria to the general population, and
especially to the patients who had prior STEMI or
coronary artery bypass graft, is unclear.
Acute anterior ST-elevation myocardial
infarction
Classic ECG patterns
LAD obstruction usually causes ST elevation in
the precordial leads V1 to V4 [82]. Aldrich and col-
leagues [83] reported similar findings showing the
frequency of ST elevation in patients who have
acute STEMI caused by LAD occlusion to be, in
descending order: V2, V3, V4, V5, aVL, V1, and V6
[83]. Uncommonly, ST elevation in leads V1 to V4
may be caused by proximal right coronary artery
(RCA) occlusion with concomitant right
353
ventricular infarction [46,84,85]. In Blanke and co-
workers’ [82] analysis of patients who have acute
STEMI caused by RCA occlusion, the frequency
of ST elevation in these four precordial leads was
as follows: V1, 5%; V2 and V3, 15%; and V4, 8%.
These investigators found no instances of ST eleva-
tion in leads V1 to V4 in the patients who had acute
myocardial infarction caused by LCX occlusion.
Right ventricular infarction that produces ST ele-
vation in leads V1 to V4 may be distinguished
from anterior STEMI by observing an ST elevation
in lead V1 greater than in lead V2, ST elevation in
the right precordial leads V3R and V4R, ST depres-
sion in lead V6, and ST elevation in the inferior
leads II, III, and aVF [84,86]. The magnitude of
ST elevation in lead V1 correlates better with the
magnitude of ST elevation in lead V3R than with
lead V2, suggesting that ST elevation in lead V1 re-
flects the right ventricle more than the left ventricle
[86]. The typical ECG findings in acute anterior
STEMI are presented in Box 1.
Diagnosis of anterior infarction extending
to contiguous myocardial zones
Anterosuperior myocardial zone
The high anterolateral wall at the base of the
left ventricle receives its coronary blood flow from
the first diagonal branch of the LAD, the first
obtuse marginal branch of the LCX, or, occa-
sionally, from the ramus intermedius artery [87].
The ECG lead that most directly faces this antero-
superior myocardial zone is lead aVL [87,88]. In
acute anterior STEMI, ST elevation in lead I,
and particularly in lead aVL, signifies an LAD oc-
clusion proximal to the first diagonal branch
[88,89]. In contrast, ST depression in lead aVL
during acute anterior STEMI signifies LAD oc-
clusion distal to the first diagonal branch [90]. Al-
though ST elevation in lead aVL is a very specific
sign of proximal LAD occlusion, it has a relatively
low sensitivity for this diagnosis. Sasaki and col-
leagues [91] noted that patients who have a proxi-
mal occlusion of long LAD artery that wraps
around the cardiac apex have concomitant injury
to the inferorapical and anterosuperior walls of
the left ventricle. When this happens, no ST eleva-
tion may be seen in either anterosuperior leads (ie,
I, aVL) or inferior leads (ie, II, III, aVF) because
the opposing forces cancel each other. Isolated oc-
clusion of the first diagonal branch also may result
in ST elevation in lead aVL [77,88]. The ECG can
be useful in distinguishing isolated diagonal
branch occlusion from LAD occlusion proximal
354 ATAR et al

Box 1. ECG findings in acute anterior STEMI

Precordial leads
 ST elevation is usually present in V2 to V4.
 ST elevation in V4 to.V6 without ST elevation in V1 to V3 usually is caused by LCX or distal
diagonal occlusion.
 ST elevation in leads V2 to V6 may represent LAD occlusion proximal to the first diagonal
branch.
Leads I and aVL
 ST elevation in lead I and aVL signifies
1. Occlusion of a short LAD coronary artery before the first diagonal branch (if there is ST
elevation in V2 to V4)
2. Occlusion of first diagonal branch (if associated with ST elevation in V2 and isoelectric ST
or ST depression in V3 to V6)
3. Occlusion of the first marginal branch of the LCX (if there is ST depression in V2)
 ST depression in aVL signifies LAD artery occlusion distal to the first diagonal branch.

Leads II, III, and aVF

 ST depression in the inferior leads signifies
1. Occlusion of a short LAD coronary artery before the first diagonal branch (if there is ST
elevation in V2 to V4)
2. Occlusion of the first diagonal branch (if associated with ST elevation in V2 and
isoelectric ST or ST depression in V3 to V6

 ST elevation in the inferior leads signifies occlusion of a long LAD artery (that wraps the
cardiac apex) distal to the first diagonal branch.
Lead aVR
 ST elevation in aVR signifies LAD artery occlusion proximal to the first septal branch.

Right bundle-branch block (new)

 Right bundle-branch block signifies LAD artery occlusion proximal to the first septal
branch.

to the first diagonal branch, however. Occlusion
of the diagonal branch typically results in ST ele-
vation in leads I, aVL, and V2 with ST segments in
leads V3 and V4 either isoelectric or depressed
[77,87]. In contrast, LAD occlusion proximal to
the first diagonal branch results in ST elevation
extending beyond lead V2 or V3 and occasionally,
to V4 to V6 [77,87]. In addition, when ST elevation
in leads I and aVL are caused by occlusion of the
LCX, reciprocal ST depression is usually observed
in lead V2 because the vascular bed supplied by
the LCX extends more posteriorly [25].
Several ECG criteria have been reported by
Engelen and colleagues [90] to indicate an LAD ar-
tery occlusion proximal to the first septal perforator
branch: (1) ST elevation in lead aVR (sensitivity
43%; specificity 95%; P ¼ .000); (2) right bundle
branch block (sensitivity 14%; specificity 100%;
P ¼ .004); (3) ST depression in lead V5 (sensitivity
17%; specificity 100%; P ¼ .009); and (4) ST eleva-
tion in lead V1 greater than 2.5 mm (sensitivity
12%; specificity 100%; P ¼ .011). These findings
were also supported by a recent study done by Va-
sudevan and colleagues [92] on 50 patients who had
acute anterior STEMI and underwent angiography
within 3 days of the infarct. Criteria reported to in-
dicate LAD artery occlusion distal to the first septal
perforator branch include abnormal Q waves in
leads V4 to V6 [90]. On the other hand, Birnbaum
and colleagues [93] did not find an association
between ST elevation in lead V1 and LAD artery
occlusion proximal to the first septal branch. Ben-
Gal and colleagues [94] suggested that because the
right paraseptal area is supplied by the septal
 ECG DIAGNOSIS OF STEMI
branches of the LAD, alone or together with the
conal branch originating from the RCA, ST eleva-
tion in lead V1 in anterior STEMI is caused by right
paraseptal ischemia in patients who have an ana-
tomically small and nonprotective conal branch
of the RCA.
Lateral and apical myocardial zones
Most patients (93%) who have an acute
anterior STEMI caused by LAD occlusion have
an anteroseptal pattern (ST elevation in leads V1
to V3) [82,83]. In contrast, isolated ST elevation
in leads V4 to V6, without ST elevation in leads
V1 to V3, usually is caused by an occlusion of
the LCX or distal diagonal branch. Many assume
that in patients who have extensive anterior myo-
cardial infarction (ST elevation in leads V1 to V6),
the injury extends to the distal anterolateral wall
and cardiac apex caused by a long LAD artery
or prominent diagonal branches, whereas patients
who have an anteroseptal pattern (ST elevation
confined to leads V1 to V3) have a short LAD or
large obtuse marginal branches or ramus interme-
diate branch supplying these anterolateral and
apical zones. A study by Shalev and colleagues
[95] investigated the correlation of the ECG pat-
tern of anteroseptal myocardial infarction with
the echocardiographic and angiographic findings.
They found that 48 of 52 patients (92%) who pre-
sented with ST elevation in leads V1 to V3 had an
anteroapical infarct and a normal septal motion.
The culprit narrowing was found more frequently
in the mid to distal LAD (in 85% of patients).
They conclude that the ECG pattern traditionally
termed ‘‘anteroseptal STEMI’’ should be called an
‘‘anteroapical myocardial infarction’’; the term
‘‘extensive anterior STEMI’’ should be used
when associated with diffuse ST changes involving
the anterior, lateral, and occasionally, inferior
leads. Recently, using transthoracic echocardiog-
raphy, Porter and coworkers [96] found no differ-
ence in regional wall motion abnormalities in the
lateral and apical segments in patients presenting
with first acute anterior STEMI, with or without
ST elevation in leads V5 and V6.
Inferior myocardial zone
During acute anterior STEMI, injury may
extend to the inferior wall, as evidenced by ST
elevation in leads II, III, and aVF, if the LAD artery
wraps around the cardiac apex [82,97,98]. As previ-
ously mentioned, however, anterior STEMI that is
caused by a wrapping LAD occlusion proximal to
355
the first diagonal branch does not manifest as an an-
terior and inferior injury pattern because of cancel-
lation of opposing vectors [91,98]. In theory,
occlusion of an LAD artery that supplies collateral
blood flow to an obstructed RCA or LCX may pro-
duce a similar anterior and inferior pattern; how-
ever, Tamura and colleagues [98] did not observe
such a pattern in the 12 patients they studied with
this type of ECG.
ST depression in the inferior leads II, III, and
aVF during acute anterior STEMI indicates injury
to the high anterolateral wall and does not signify
inferior wall ischemia [99,100]. Several investiga-
tors found such reciprocal ST depression in the in-
ferior leads to indicate LAD artery occlusion
proximal to the first diagonal branch [76,90]. In
patients who have a long LAD artery that wraps
around the cardiac apex, however, proximal LAD
artery occlusion may not produce reciprocal ST
depression in the inferior leads because of extension
of the infarction to the inferoapical wall [91].
Acute inferior ST-elevation myocardial
infarction
Classic ECG patterns
In inferior STEMI, the leads showing the
greatest magnitude of ST elevation, in descending
order, are leads III, aVF, and II. Most patients
who have ST elevation in these inferior leads
(80%–90%) have an occlusion of the RCA;
however, an occlusion of the LCX can produce
a similar ECG pattern [101]. In addition to ST el-
evation in the inferior leads II, III, and aVF, re-
ciprocal ST depression in lead aVL is seen in
almost all patients who have acute inferior
STEMI [78]. The ECG distinction between
RCA- and LCX-related inferior STEMI is pre-
sented in Table 1 [101–107]. ECG confirmation
of the infarct-related artery during acute inferior
STEMI may be particularly valuable when coro-
nary angiography indicates lesions in both the
RCA and LCX.
Criteria in the precordial leads
Because the right ventricular branch originates
from the RCA, criteria for right ventricular
infarction, especially ST elevation in leads V3R
and V4R, provide compelling evidence that the in-
farct-related artery in acute inferior STEMI is the
proximal RCA. Kontos and colleagues [102] re-
ported that LCX-related inferior STEMI was
356 ATAR et al

suggested by reciprocal ST depression in leads V1

Specificity (%)
and V2. When Birnbaum and colleagues [13] com-
pared patients who had inferior STEMI caused by
mid or distal RCA versus LCX occlusion, how-

93
95

76
ever, they found no difference in the frequency
of ST depression in leads V1 to V3. ST depression
is absent in leads V1 to V3 only during proximal
RCA occlusion, because the resultant right ven-
Sensitivity (%)

tricular injury pattern cancels out such reciprocal

ST depression [13]. Some investigators have re-
ported a higher frequency of ST elevation in leads
V4 to V6 in patients who had LCX-related acute
80
84

88

inferior STEMI [102,103]. Hasdai and colleagues

Left Circumflex Artery Occlusion

[105], however, found little difference in the fre-

quency of ST elevation in leads V4 to V6 between
STY V3 / ST[ III O 1.2

patients who have LCX- versus RCA-related infe-

rior STEMI. The authors state that because most
S:R ratio aVL % 3

cases of acute inferior STEMI are caused by RCA

occlusions, the positive predictive value (PPV) of
No STY aVL

ST elevation in leads V5 or V6 for LCX-related in-

Criterion

farction is only 59% [105]. Kosuge and colleagues

[106] reported that the magnitude of ST depres-
sion in lead V3 relative to the ST elevation in
Common ECG features distinguishing the culprit artery in acute inferior wall myocardial infarction

lead III (V3:III ratio) was useful in distinguishing

the culprit artery in inferior STEMI. They found
that a V3:III ratio of less than 0.5 indicated a prox-
Specificity (%)

imal RCA occlusion; a V3:III ratio of 0.5 to 1.2 in-

dicated a distal RCA occlusion; and a V3:III ratio
of more than 1.2 indicated LCX occlusion [106].
Again, these criteria are based on the fact that
87
91
93
88
94
94

with RCA occlusion the vector that causes ST de-

pression in lead V3 is masked by the vector of
right ventricular injury.
Sensitivity (%)

Criteria in the limb leads

ECG criteria in the limb leads also have been
100
91
84
76
88
80

found to be useful in distinguishing RCA and

LCX occlusion in acute inferior STEMI. For
example, greater ST elevation in lead III than in
Abbreviation: RCA, right coronary artery.

lead II has been shown to indicate RCA infarction

proximal RCA; 0.5–1.2 ¼ distal RCA

[104,108]. Additional limb lead criteria involve

careful analysis of leads I and aVL. Patients
Right Coronary Artery Occlusion

who have LCX-related STEMI less frequently

show reciprocal ST depression in lead aVL and
more often show an isoelectric or a raised ST seg-
STY V3 / ST[ III ! 0.5 ¼

ment in leads I and aVL compared with patients

who have RCA-related inferior infarction
STY; aVL O STY I
S:R ratio aVL O 3

[102,104]. Hasdai and colleagues [105] reported

ST[ V3R, V4R

that such absence of reciprocal ST depression in

ST[; III O II

lead aVL indicates injury of the anterosuperior

Criterion

base of the heart typically caused by LCX occlu-

Table 1

sion proximal to the first obtuse marginal branch.

These investigators found absence of reciprocal
 ECG DIAGNOSIS OF STEMI
ST depression in lead aVL in 86% of patients who
had proximal LCX-related inferior STEMI but in
none of the patients who had RCA- or distal
LCX-related infarctions (P ¼ .0001). An addi-
tional criterion for identifying the culprit artery
in inferior STEMI is the magnitude of ST depres-
sion in lead aVL compared with lead I [104].
Greater reciprocal ST depression in lead aVL
than in lead I suggests an RCA-related inferior
STEMI [104]. A likely explanation for this phe-
nomenon is that injury of the high posterolateral
region caused by LCX occlusion attenuates ST de-
pression in lead aVL more than in lead I, which
has a less superior orientation. A final criterion
in lead aVL to distinguish the culprit artery in in-
ferior STEMI relates to the amplitude of the re-
spective R and S waves in this lead [107]. In the
initial stages, leads facing an infarcted wall with
ST elevation tend to show QRS changes as well,
including an increase in R-wave amplitude and
a decrease in S-wave amplitude [109]. Therefore,
in inferior STEMI, the opposite pattern (that is,
decrease in R wave and increase in S wave) would
be expected in lead aVL, if there is no involvement
of the high posterolateral region (RCA infarc-
tion). In contrast, if there is concomitant involve-
ment of the high posterolateral segments (as
expected especially in proximal LCX infarction),
these reciprocal changes in the QRS may not be
apparent. Indeed, Assali and colleagues [107]
found that a decrease in R-wave amplitude and
an increase in S-wave amplitude with a S:R ratio
greater than 3 predicted RCA occlusion, whereas
a S:R ratio less than 3 predicted LCX occlusion.
A summary of the criteria to distinguish the
culprit artery in inferior STEMI is provided in
Table 1, and an example of the criteria is shown in
Fig. 6. From these criteria it is clear that the abil-
ity to differentiate RCA from LCX occlusion is
greater for proximal occlusion. Because occlusion
occurs more distally in the culprit artery, the dis-
tinctive characteristics are lost, and the ECG can-
not be expected to distinguish between right and
left posterior descending artery occlusion.
A recent study by Fiol and colleagues [110] has
suggested using the following ECG criteria in
a three-step algorithm to differentiate RCA from
LCX occlusion as the culprit artery: (1) ST
changes in lead I, (2) the ratio of ST elevation in
lead III to that in lead II, and (3) the ratio of
the sum ST depression in leads V1 to V3 divided
by the sum ST elevation in leads II, III, and
aVF. ST elevation in lead I has a 100% PPV for
LCX occlusion, whereas ST depression of 0.5 mm
357
or more has a 94% PPV for RCA occlusion.
A ratio of ST elevation in lead III to lead II
greater than 1 has a 92% PPV for RCA occlusion,
and the ratio of the sum of ST depression in leads
V1 to V3/ST elevation in II, III, aVF of 1 or less
has a 90% PPV for RCA occlusion. Application
of this sensitive algorithm suggested the location
of the culprit coronary artery (RCA versus
LCX) in 60 of 63 patients (O95%).
Diagnosis of inferior infarction extending
to contiguous myocardial zones
Right ventricular myocardial zone
Right ventricular infarction occurs almost
exclusively in the setting of inferior STEMI.
Although isolated right ventricular infarction
has been reported, it is rare and occurs most often
in patients who have right ventricular hypertro-
phy [111]. Several investigators have found that
ST elevation in lead V4R is diagnostic of right
ventricular infarction with a sensitivity and speci-
ficity well over 90% [101,112,113]. It is important
to point out that ST elevation in the right precor-
dial leads (eg, V4R) is most prominent in the early
hours of inferior STEMI and dissipates rapidly
thereafter. Hence, the window of opportunity to
diagnose right ventricular infarction by ECG is
limited, and right precordial leads should be re-
corded immediately as a patient who has ST eleva-
tion in the inferior leads presents to the emergency
room. As mentioned previously, ST elevation in
leads V1 to V3 in a patient who has inferior
STEMI is a manifestation of associated right ven-
tricular infarction caused by a proximal RCA oc-
clusion [46,84,85]. Lopez-Sendon and colleagues
[114] reported that the criterion of ST elevation
in lead V4R greater than ST elevation in any of
leads V1 to V3 is a very specific sign of right ven-
tricular infarction (specificity, 100%). This crite-
rion was less sensitive (sensitivity 78.6%) than
ST elevation in V4R alone, however [114].
A recent analysis of the GUSTO-I angio-
graphic substudy by Sadanandan and colleagues
[115], in patients who had concomitant ST eleva-
tion in the anterior (V1–V4) and inferior (II, III,
aVF) leads, revealed that the infarct-related artery
was the RCA in 59% of the cases and was the
LAD in 36%. More patients who had RCA occlu-
sion had ST elevation in lead V1 equal to or
greater than the ST elevation in lead V3 compared
with those with LAD occlusion (35% versus 12%;
P ¼ .001). Furthermore, the progression of ST
358 ATAR et al

Fig. 6. (A) Admission ECG of a patient who has acute inferior STEMI and total occlusion if the proximal right coro-
nary artery on immediate angiography. There is ST elevation in leads II, III, and aVF. ST elevation is greater in lead III
than in lead II. There is reciprocal ST depression and deep S waves in lead aVL. The magnitude of ST depression in aVL
is greater than in lead I. In addition, there is no ST depression in leads V1 to V3. (B) Admission ECG of a patient who has
acute inferior STEMI caused by occlusion of the proximal LCX on immediate angiography. There is ST elevation in
leads II, III, and aVF and ST depression in leads V1 to V3. The magnitude of ST elevation in lead III is smaller than
in lead II. Of note, there is no reciprocal ST depression in lead aVL.

elevation from lead V1 to lead V3 was significantly
greater in patients who LAD occlusion than in
those who had RCA occlusion. Thus, larger ST el-
evation in leads II, III, and aVF and absence of
progression of ST elevation from V1 to V3 differ-
entiate RCA from LAD occlusion in patients
who have combined anterior and inferior ST ele-
vation. Only 35% of the patients who had RCA
occlusion had ST elevation in lead V1 equal to
or greater than the ST elevation in V3, however.
Therefore, the absence of this criterion does not
exclude RCA occlusion.
Lateral apical myocardial zone
In patients who have inferior STEMI, ST
elevation in leads V5 and V6 is considered to indi-
cate extension of the infarct to the lateral aspect of
the cardiac apex; however, there is as yet no direct
evidence for this extension [116]. The cause of
such an extension may be occlusion of either the
LCX or RCA with a posterior descending or pos-
terolateral branch that extends to the lateral api-
cal zone [116]. Tsuka and coworkers [58] found
that ST elevation in lead V6 during inferior
STEMI was associated with a larger infarct size,
a greater frequency of major cardiac arrhythmias,
and a higher incidence of pericarditis during the
patient’s hospital course. A study by Golovchiner
and colleagues [117] assessed the correlation be-
tween ST deviation in each of the six precordial
leads and the presence of regional wall motion ab-
normalities by transthoracic echocardiography in
109 patients who had first inferior STEMI. ST el-
evation in lead V5 was associated with more fre-
quent involvement of the apical portion of the
 ECG DIAGNOSIS OF STEMI
inferior wall (P ! .02), with a specificity of 88%
and a sensitivity of 33%. Global regional wall mo-
tion abnormality score was significantly worse for
patients who had ST elevation than forpatients
who had isoelectric ST in lead V5 (P ¼ .024). ST
elevation in lead V6 was associated with regional
wall motion abnormality in the mid-posterior seg-
ment (P ! .006), with a specificity of 91% and
a sensitivity of 33%, and worse global regional
wall motion abnormality score (P ¼ .022).
Posterior myocardial zone
In patients who have inferior STEMI, ST
depression in leads V1 to V3 has been shown to
indicate a larger infarction with extension of the
injury to the posterolateral or the inferoseptal
wall [45,53–56,118–122]. Such ST depression in
these anterior leads during inferior STEMI is a
reciprocal change and does not indicate concomi-
tant LAD coronary artery disease [13,79]. It
may be seen in both RCA and LCX inferior in-
farctions [44,123]. In inferior STEMI caused by
proximal RCA occlusion with concomitant right
ventricular infarction, however, posterior wall in-
jury may be masked because the two opposed
electrical vectors may cancel each other (that is,
ST elevation in leads V1 to V3 with right ventricu-
lar infarction and reciprocal ST depression in
these same leads with concurrent posterior infarc-
tion) [122]. A more direct sign of posterior wall in-
jury is ST elevation in leads V7 to V9 [59,124–126].
Waveform amplitudes in these posterior leads are
smaller than in standard precordial leads, how-
ever. There is preliminary evidence that ST eleva-
tion of 0.5 mm should be considered a sign of
injury when analyzing the posterior leads [127].
Isolated ST elevation in leads V7 to V9 without
ST elevation in the inferior leads occurs in only
4% of patients who have acute myocardial infarc-
tion [126] and usually is caused by LCX occlusion
[124]. In patients who have acute inferior STEMI,
ST elevation in leads V7 to V9 is associated with
a higher incidence of reinfarction, heart failure,
and death [59].
Ischemia at a distance in acute inferior
ST-elevation myocardial infarction
Most of the ST depression patterns seen during
STEMI represent reciprocal changes rather than
ischemia at a distance [72]. One ECG pattern, ST
depression in leads V5 and V6 in inferior STEMI,
signifies concomitant disease of the LAD with
acute ischemia in a myocardial zone remote from
359
the infarct zone [11,13,79,128]. Patients who have
maximal ST depression in leads V4 to V6 during in-
ferior STEMI have higher morbidity and mortality
than patients without precordial ST depression or
with maximal ST depression in leads V1 to V3
[25]. Likewise, patients who have maximal ST
depression in leads V4 to V6 undergo multivessel
revascularization (multivessel PCI or coronary
artery bypass surgery) more often than do patients
who do not have such an ECG pattern [128].
ST segment elevation in lead aVR has been
shown to be a marker of severe diffuse coronary
disease in patients who have unstable angina or
non-STEMI. ST elevation in lead aVR also can be
a marker of acute left main coronary (LMCA)
occlusion. Yamaji and colleagues [129] studied
the admission 12-lead ECG in 16 consecutive pa-
tients who had acute LMCA obstruction, 46
patients who had acute LAD obstruction, and
24 patients who had acute RCA obstruction.
Lead aVR ST-segment elevation (O0.05 mV)
occurred with a significantly higher incidence
(P ! .01) in the LMCA group (88%) than in
the LAD (43%) or RCA (8%) groups. Lead
aVR ST-segment elevation was significantly
higher in the LMCA group (0.16  0.13 mV)
than in the LAD group (0.04  0.10 mV). Lead
V1 ST-segment elevation was lower in the
LMCA group (0.00  0.21 mV) than in the
LAD group (0.14  0.11 mV). The finding of
lead aVR ST-segment elevation greater than or
equal to lead V1 ST elevation distinguished the
LMCA group from the LAD group with 81%
sensitivity, 80% specificity, and 81% accuracy.
ST segment shift in lead aVR and the inferior
leads distinguished the LMCA group from the
RCA group. The authors conclude that lead
aVR ST-segment elevation with less ST-segment
elevation in lead V1 is an important predictor of
acute LMCA obstruction [129].
Discussion
The admission ECG pattern is the most in-
formative noninvasive tool for the diagnosis,
triage, and risk stratification in patients who
have STEMI. Three ECG patterns presented in
this article are especially relevant: (1) right ven-
tricular infarction accompanying acute inferior
STEMI, (2) a very proximal LAD occlusion in
anterior STEMI, and (3) patients at higher risk
(ie, those who have grade 3 ischemia, ST de-
pression in V4–V6 concomitant with inferior
STEMI, and ST elevation in lead aVR).
360 ATAR
Moreover, it is crucial to recognize cases in
which opposing ECG vectors cancel each other
and result in attenuation of the ischemic changes,
such as occlusion of a proximal LAD that wraps
the cardiac apex or a proximal dominant LCX. In
terms of the first assessment, the opportunity to
diagnose right ventricular infarction using the
ECG is greatest in the emergency department
because ST elevation in the right precordial leads
resolves quickly. The admitting physician should
make certain that all patients who have acute
inferior STEMI have a second ECG recorded with
right ventricular leads. If ST-segment elevation of
1 mm is observed in lead V4R, the diagnosis of
right ventricular infarction can be made.
Another ECG assessment of importance to
the admitting physician is the identification of
a very proximal LAD occlusion in acute anterior
STEMI. If the occlusion site is proximal to the first
diagonal branch of the LAD, a large portion of the
left ventricle is at risk of irreversible damage,
including the anteroseptal, anterosuperior, antero-
lateral, and apical regions. Such high-risk patients
should be transferred urgently to a cardiac cath-
eterization laboratory for primary PCI.
Patients who have grade 3 ischemia on the
admission ECG have higher mortality [24–
26,130,131] and reinfarction rates [130,132]. Retro-
spective analysis of the GUSTO IIB trial patients
revealed that grade 3 ischemia was associated
with higher mortality both in the primary PCI
group and in the thrombolysis group [130]. In the
grade 2 group, in-hospital mortality was similar in
the thrombolysis and angioplasty subgroups
(3.2% and 3.3%; P ¼ .941). In patients who had
grade 3 ischemia, in-hospital mortality was 6.4%
and 7.3%, respectively (P ¼ .762). These findings
were confirmed in an analysis of the Danish
multicenter randomized study on fibrinolytic
therapy versus acute coronary angioplasty in acute
myocardial infarction, DANAMI–2 [133]. The
study showed that mortality increased significantly
with symptom duration (O3 hours) in both grade 2
and grade 3 ischemia regardless of treatment
strategy. For patients presenting early (!3 hours
of symptom onset) who had grade 3 ischemia, how-
ever, those in the primary PCI group had a 5.5%
30-day mortality reduction (1.4% versus 6.9%)
compared with the thrombolysis group.
Summary
The admission ECG in patients who have
STEMI is valuable for selecting patients who are
et al
candidates for early reperfusion treatment and
also for providing information regarding the
location and extent of acute myocardial injury.
By reflecting the pathophysiology of the myocar-
dium during acute ischemia, the ECG conveys
information not provided by coronary angiogra-
phy and provides important information to guide
clinical decision making.
References
[1] Keeley EC, Boura JA, Grines CL. Primary angio-
plasty versus intravenous thrombolytic therapy
for acute myocardial infarction: a quantitative
review of 23 randomised trials. Lancet 2003;
361(9351):13–20.
[2] Antman EM, Anbe DT, Armstrong PW, et al.
ACC/AHA guidelines for the management of pa-
tients with ST-elevation myocardial infarction–
executive summary: a report of the American
College of Cardiology/American Heart Associa-
tion Task Force on Practice Guidelines (Writing
Committee to Revise the 1999 Guidelines for the
Management of Patients With Acute Myocardial
Infarction). Circulation 2004;110(5):588–636.
[3] Braunwald E, Antman EM, Beasley JW, et al.
ACC/AHA guideline update for the management
of patients with unstable angina and non-ST-
segment elevation myocardial infarction–2002:
summary article: a report of the American College
of Cardiology/American Heart Association Task
Force on Practice Guidelines (Committee on the
Management of Patients With Unstable Angina).
Circulation 2002;106(14):1893–900.
[4] Arnold AER, Simoons ML. ‘‘Expected infarct size
without thrombolysis’’, a concept that predicts im-
mediate and long-term benefit from thrombolysis
for evolving myocardial infarction. Eur Heart J
1997;18:1736–48.
[5] Wang K, Asinger RW, Marriott HJ. ST-segment
elevation in conditions other than acute myocardial
infarction. N Engl J Med 2003;349(22):2128–35.
[6] Hiss RG, Lamb LE, Allen MF. Electrocardio-
graphic findings in 67,375 asymptomatic subjects.
X. Normal values. Am J Cardiol 1960;6:200–31.
[7] Surawicz B, Parikh SR. Prevalence of male and fe-
male patterns of early ventricular repolarization in
the normal ECG of males and females from child-
hood to old age. J Am Coll Cardiol 2002;40(10):
1870–6.
[8] Dowdy L, Wagner GS, Birnbaum Y, et al. Aborted
infarction: the ultimate myocardial salvage. Am
Heart J 2004;147(3):390–4.
[9] Wong CK, French JK, Aylward PE, et al. Pa-
tients with prolonged ischemic chest pain and
presumed-new left bundle branch block have
heterogeneous outcomes depending on the pres-
ence of ST-segment changes. J Am Coll Cardiol
2005;46(1):29–38.
 ECG DIAGNOSIS OF STEMI
[10] Moshkovitz Y, Sclarovsky S, Behar S, et al. Infarct
site-related mortality in patients with recurrent
myocardial infarction. SPRINT Study Group.
Am J Med 1993;94:388–94.
[11] Strasberg B, Pinchas A, Barbash GI, et al. Im-
portance of reciprocal ST segment depression
in leads V5 and V6 as an indicator of disease
of the left anterior descending coronary artery
in acute inferior wall myocardial infarction. Br
Heart J 1990;63:339–41.
[12] Birnbaum Y, Sclarovsky S, Strasberg B. Critical
left main stenosis [letter]. Am Heart J 1994;127:
1662–3.
[13] Birnbaum Y, Wagner GS, Barbash GI, et al. Corre-
lation of angiographic findings and right (V1 to V3)
versus left (V4 to V6) precordial ST-segment de-
pression in inferior wall acute myocardial infarc-
tion. Am J Cardiol 1999;83(2):143–8.
[14] Habib GB, Heibig J, Forman SA, et al. Influ-
ence of coronary collateral vessels on myocardial
infarct size in humans. Results of phase I
Thrombolysis in Myocardial Infarction (TIMI)
Trial. Circulation 1991;83:739–46.
[15] Haider AW, Andreotti F, Hackett DR, et al.
Early spontaneous intermittent myocardial reper-
fusion during acute myocardial infarction is as-
sociated with augmented thrombogenic activity
and less myocardial damage. J Am Coll Cardiol
1995;26:662–7.
[16] Kloner RA, Yellon D. Does ischemic precondition-
ing occur in patients? J Am Coll Cardiol 1994;24(4):
1133–42.
[17] Sclarovsky S, Mager A, Kusniec J, et al. Electrocar-
diographic classification of acute myocardial ische-
mia. Isr J Med Sc 1990;26:525–33.
[18] Sclarovsky S. Electrocardiography of acute myo-
cardial ischaemic syndromes. London: Martin
Dunitz Ltd.; 1999. p. 48–72.
[19] Birnbaum Y, Wagner GS. The Initial electrocar-
diographic pattern in acute myocardial infarction:
correlation with infarct size. J Electrocardiol 1999;
32(Suppl):122–8.
[20] Spekhorst H, SippensGroenewegen A, David G,
et al. Body surface mapping during percutaneous
transluminal coronary angioplasty. QRS changes
indicating regional myocardial conduction delay.
Circulation 1990;81:840–9.
[21] Wagner N, Sevilla D, Krucoff M, et al. Transient al-
terations of the QRS complex and ST segment dur
ing percutaneous transluminal balloon angioplasty
of the left anterior descending coronary artery. Am
J Cardiol 1988;62:1038–42.
[22] DeHaan R. Differentiation of the atrioventricular
conducting system of the heart. Circulation 1961;
24:458–70.
[23] Feldman T, Chua K, Childres R. R wave of the sur-
face and intracoronary electrogram during acute
coronary artery occlusion. Am J Cardiol 1986;58:
885–90.
361
[24] Birnbaum Y, Sclarovsky S, Blum A, et al. Prognos-
tic significance of the initial electrocardiographic
pattern in a first acute anterior wall myocardial in-
farction. Chest 1993;103:1681–7.
[25] Birnbaum Y, Herz I, Sclarovsky S, et al. Prognostic
significance of the admission electrocardiogram in
acute myocardial infarction. J Am Coll Cardiol
1996;27:1128–32.
[26] Birnbaum Y, Kloner R, Sclarovsky S, et al. Distor-
tion of the terminal portion of the QRS on the ad-
mission electrocardiogram in acute myocardial
infarction and correlation with infarct size and
long term prognosis (Thrombolysis In Myocardial
Infarction 4 Trial). Am J Cardiol 1996;78:396–403.
[27] Birnbaum Y, Maynard C, Wolfe S, et al. Terminal
QRS distortion on admission is better than ST-
segment measurements in predicting final infarct
size and assessing the potential effect of thrombo-
lytic therapy in anterior wall acute myocardial
infarction. Am J Cardiol 1999;84(5):530–4.
[28] Garcia-Rubira JC, Perez-Leal I, Garcia-Martinez
JT, et al. The initial electrocardiographic pattern
is a strong predictor of outcome in acute myocar-
dial infarction. Int J Cardiol 1995;51:301–5.
[29] Rude RF, Poole KW, Muller JE, et al. Electrocar-
diographic and clinical criteria for recognition of
acute myocardial infarction based on analysis of
3,697 patients. Am J Cardiol 1983;52:936–42.
[30] Lee TH, Rouan GW, Weisberg MC, et al. Sensi-
tivity of routine clinical criteria for diagnosing
myocardial infarction within 24 hours of hospital-
ization. Ann Intern Med 1987;106(2):181–6.
[31] Gibler WB, Young GP, Hedges JR, et al. Acute
myocardial infarction in chest pain patients with
nondiagnostic ECGs: serial CK-MB sampling in
the emergency department. The Emergency Med-
icine Cardiac Research Group. Ann Emerg Med
1992;21(5):504–12.
[32] Hedges JR, Young GP, Henkel GF, et al. Serial
ECGs are less accurate than serial CK-MB re-
sults for emergency department diagnosis of myo-
cardial infarction. Ann Emerg Med 1992;21(12):
1445–50.
[33] Menown IB, Mackenzie G, Adgey AA. Optimiz-
ing the initial 12-lead electrocardiographic diag-
nosis of acute myocardial infarction. Eur Heart
J 2000;21(4):275–83.
[34] Birnbaum Y, Kloner RA. Clinical aspects of myo-
cardial stunning. Coron Artery Dis 1995;6(8):
606–12.
[35] Kaul S. There may be more to myocardial viability
than meets the eye! Circulation 1995;92:2790–3.
[36] Aldrich H, Wagner N, Boswick J, et al. Use of ini-
tial ST-segment deviation for prediction of final
electrocardiographic size of acute myocardial in-
farcts. Am J Cardiol 1988;61:749–53.
[37] Clemmensen P, Grande P, Aldrich H, et al. Evalu-
ation of formulas for estimating the final size of
acute myocardial infarcts from quantitative ST-
362 ATAR
segment elevation on the initial standard 12-lead
ECG. J Electrocardiol 1991;24:77–83.
[38] Christian T, Gibbons R, Clements I, et al. Estimates
of myocardium at risk and collateral flow in acute
myocardial infarction using electrocardiographic in-
dexes with comparison to radionuclide and angio-
graphic measures. J Am Coll Cardiol 1995;26:
388–93.
[39] Clements I, Kaufmann P, Bailey K, et al. Electro-
cardiographic prediction of myocardial area at
risk. Mayo Clin Proc 1991;66:985–90.
[40] Willems JL, Willems RJ, Willems GM, et al. Signif-
icance of initial ST segment elevation and depres-
sion for the management of thrombolytic therapy
in acute myocardial infarction. Circulation 1990;
82:1147–58.
[41] Vermeer F, Simoons ML, Bar FW, et al. Which pa-
tients benefit most from early thrombolytic therapy
with intracoronary streptokinase? Circulation
1986;74(6):1379–89.
[42] Birnbaum Y, Criger DA, Strasberg B, et al. Predic-
tion of the extent and severity of left ventricular
dysfunction in anterior acute myocardial infarction
by the admission electrocardiogram. Am Heart J
2001;141:915–24.
[43] Roberts WC, Gardin JM. Location of myocardial
infarcts: a confusion of terms and definitions. Am
J Cardiol 1978;42:868–72.
[44] Huey BL, Beller GA, Kaiser DL, et al. A compre-
hensive analysis of myocardial infarction due to
left circumflex artery occlusion: comparison with
infarction due to right coronary artery and left
anterior descending artery occlusion. J Am Coll
Cardiol 1988;12:1156–66.
[45] Sclarovsky S, Topaz O, Rechavia E, et al. Ischemic
ST segment depression in leads V2–V3 as the pre-
senting electrocardiographic feature of posterolat-
eral wall myocardial infarction. Am Heart J 1987;
113:1085–90.
[46] Geft IL, Shah PK, Rodriguez L, et al. ST elevation
in lead V1 to V5 may be caused by right coronary
artery occlusion and acute right ventricular infarc-
tion. Am J Cardiol 1984;53:991–6.
[47] Juergens CP, Fernandes C, Hasche ET, et al. Elec-
trocardiographic measurement of infarct size after
thrombolytic therapy. J Am Coll Cardiol 1996;27:
617–24.
[48] Hasche ET, Fernandes C, Freedman SB, et al. Re-
lation between ischemia time, infarct size, and left
ventricular function in humans. Circulation 1995;
92:710–9.
[49] Wong CK, French JK, Andrews J, et al. Usefulness
of the presenting electrocardiogram in predicting
myocardial salvage with thrombolytic therapy in
patients with a first acute myocardial infarction.
Eur Heart J 2002;23(5):399–404.
[50] Birnbaum Y, Mahaffey KW, Criger DA, et al.
Grade III ischemia on presentation with acute
myocardial infarction predicts rapid progression
et al
of necrosis and less myocardial salvage with throm-
bolysis. Cardiology 2002;97(3):166–74.
[51] Billgren T, Maynard C, Christian TF, et al. Grade
3 ischemia on the admission electrocardiogram pre-
dicts rapid progression of necrosis over time and
less myocardial salvage by primary angioplasty.
J Electrocardiol 2005;38(3):187–94.
[52] Shah PK, Pichler M, Berman DS, et al. Noninva-
sive identification of a high risk subset of patients
with acute inferior myocardial infarction. Am
J Cardiol 1980;46:915–21.
[53] Gibson RS, Crampton RS, Watson DD, et al. Pre-
cordial ST-segment depression during acute infe-
rior myocardial infarction: clinical, scintigraphic
and angiographic correlations. Circulation 1982;
66:732–41.
[54] Roubin GS, Shen WF, Nicholson M, Dunn RF,
et al. Anterolateral ST segment depression in acute
inferior myocardial infarction: angiographic and
clinical implications. Am Heart J 1984;107:1177–82.
[55] Ruddy TD, Yasuda T, Gold HK, et al. Anterior ST
segment depression in acute inferior myocardial in-
farction as a marker of greater inferior, apical, and
posterolateral damage. Am Heart J 1986;112:
1210–6.
[56] Ong L, Valdellon B, Coromilas J, et al. Precordial
S-T segment depression in inferior myocardial in-
farction: evaluation by quantitative thallium-201
scintigraphy and technetium-99m ventriculogra-
phy. Am J Cardiol 1983;51:734–9.
[57] Birnbaum Y, Herz I, Sclarovsky S, et al. Prognostic
significance of precordial ST segment depression on
admission electrocardiogram in patients with infe-
rior wall myocardial infarction. J Am Coll Cardiol
1996;28:313–8.
[58] Tsuka Y, Sugiura T, Hatada K, et al. Clinical char-
acteristics of ST-segment elevation in lead V6 in
patients with Q-wave acute inferior wall myocar-
dial infarction. Coron Artery Dis 1999;10(7):465–9.
[59] Matetzky S, Freimark D, Chouraqui P, et al. Sig-
nificance of ST segment elevations in posterior
chest leads (V7 to V9) in patients with acute inferior
myocardial infarction: application for thrombo-
lytic therapy. J Am Coll Cardiol 1998;31(3):506–11.
[60] Fisch C. Electrocardiography. In: Braunwald E,
editor. Heart disease. A textbook of cardiovascular
medicine. 5th edition. Philadelphia: W.B. Saun-
ders; 1997. p. 109–52.
[61] Raitt M, Maynard C, Wagner G, et al. Appearance
of abnormal Q waves early in the course of acute
myocardial infarction: implications for efficacy of
thrombolytic therapy. J Am Coll Cardiol 1995;25:
1084–8.
[62] Goldberg S, Urban P, Greenspon A, et al. Limita-
tion of infarct size with thrombolytic agents-
electrocardiographic indexes. Circulation 1983;
68(Suppl I):I-77–82.
[63] Rechavia E, Blum A, Mager A, et al. Electrocardio-
graphic Q-waves inconstancy during thrombolysis
 ECG DIAGNOSIS OF STEMI
in acute anterior wall myocardial infarction. Cardi-
ology 1992;80:392–8.
[64] Bateman TM, Czer LSC, Gray RJ, et al. Tran-
sient pathologic Q waves during acute ischemic
events: an electrocardiographic correlate of
stunned but viable myocardium. Am Heart J
1983;106:1421–6.
[65] Gross H, Rubin IL, Laufer H, et al. Transient
abnormal Q waves in the dog without myocardial
infarction. Am J Cardiol 1964;14:669–74.
[66] Bar FW, Vermeer F, de Zwaan C, et al. Value of ad-
mission electrocardiogram in predicting outcome
of thrombolytic therapy in acute myocardial infarc-
tion. A randomized trial conducted by The Nether-
lands Interuniversity Cardiology Institute. Am J
Cardiol 1987;59:6–13.
[67] Timmis G. Electrocardiographic effects of reperfu-
sion. Cardiol Clin 1987;5:427–45.
[68] Wong CK, French JK, Aylward PE, et al. Useful-
ness of the presenting electrocardiogram in predict-
ing successful reperfusion with streptokinase in
acute myocardial infarction. Am J Cardiol 1999;
83(2):164–8.
[69] Birnbaum Y, Chetrit A, Sclarovsky S, et al. Abnor-
mal Q waves on the admission electrocardiogram
of patients with first acute myocardial infarction:
prognostic implications. Clin Cardiol 1997;20(5):
477–81.
[70] Matetzky S, Barabash GI, Shahar A, et al. Early T
wave inversion after thrombolytic therapy predicts
better coronary perfusion: clinical and angio-
graphic study. J Am Coll Cardiol 1994;24(2):
378–83.
[71] Herz I, Birnbaum Y, Zlotikamien B, et al. The
prognostic implications of negative T waves in the
leads with ST segment elevation on admission in
acute myocardial infarction. Cardiology 1999;
92(2):121–7.
[72] Becker R, Alpert J. Electrocardiographic ST seg-
ment depression in coronary heart disease. Am
Heart J 1988;115:862–8.
[73] Cribier A, Korsatz L, Koning R, et al. Improved
myocardial ischemic response and enhanced col-
lateral circulation with long repetitive coronary
occlusion during angioplasty: a prospective
study. J Am Coll Cardiol 1992;20:578–86.
[74] Birnbaum Y, Hale SL, Kloner RA. Progressive de-
crease in the ST segment elevation during ischemic
preconditioning: is it related to recruitment of col-
lateral vessels? J Mol Cell Cardiol 1996;28(7):
1493–9.
[75] Sagie A, Sclarovsky S, Strasberg B, et al. Acute
anterior wall myocardial infarction presenting
with positive T waves and without ST segment
shift. Electrocardiographic features and angio-
graphic correlation. Chest 1989;95:1211–5.
[76] Birnbaum Y, Solodky A, Herz I, et al. Implication
of inferior ST segment depression in anterior acute
myocardial infarction: electrocardiographic and
363
angiographic correlation. Am Heart J 1994;127:
1467–73.
[77] Sclarovsky S, Birnbaum Y, Solodky A, et al.
Isolated mid-anterior myocardial infarction:
a special electrocardiographic sub-type of acute
myocardial infarction consisting of ST-elevation
in non-consecutive leads and two different mor-
phologic types of ST-depression. Int J Cardiol
1994;46:37–47.
[78] Birnbaum Y, Sclarovsky S, Mager A, et al. ST seg-
ment depression in aVL: a sensitive marker for
acute inferior myocardial infarction. Eur Heart J
1993;14:4–7.
[79] Hasdai D, Birnbaum Y, Porter A, et al. Maximal
precordial ST-segment depression in leads V4–V6
in patients with inferior wall acute myocardial in-
farction indicates coronary artery disease involving
the left anterior descending coronary artery system.
Int J Cardiol 1997;58:273–8.
[80] Boden WE, Kleiger RE, Gibson RS, et al. Electro-
cardiographic evolution of posterior acute myocar-
dial infarction: importance of early precordial
ST-segment depression. Am J Cardiol 1987;59(8):
782–7.
[81] Porter A, Vaturi M, Adler Y, et al. Are there dif-
ferences among patients with inferior acute myo-
cardial infarction with ST depression in leads V2
and V3 and positive versus negative T waves in
these leads on admission? Cardiology 1998;90(4):
295–8.
[82] Blanke H, Cohen M, Schlueter GU, et al. Electro-
cardiographic and coronary arteriographic correla-
tions during acute myocardial infarction. Am J
Cardiol 1984;54:249–55.
[83] Aldrich HR, Hindman NB, Hinohara T, et al.
Identification of the optimal electrocardiographic
leads for detecting acute epicardial injury in
acute myocardial infarction. Am J Cardiol
1987;59(1):20–3.
[84] Coma-Canella I, Lopez-Sendon J, Alcasena S, et al.
Electrocardiographic alterations in leads V1 to V3
in the diagnosis of right and left ventricular infarc-
tion. Am Heart J 1986;112:940–6.
[85] Porter A, Herz I, Strasberg B. Isolated right
ventricular infarction presenting as anterior
wall myocardial infarction on electrocardiogra-
phy. Clin Cardiol 1997;20:971–3.
[86] Ben-Gal T, Herz I, Solodky A, et al. Acute anterior
wall myocardial infarction entailing ST-segment
elevation in lead V1: electrocardiographic and an-
giographic correlations. Clin Cardiol 1998;21(6):
399–404.
[87] Birnbaum Y, Hasdai D, Sclarovsky S, et al. Acute
myocardial infarction entailing ST segment eleva-
tion in lead aVL: electrocardiographic differen-
tiation among occlusion of the left anterior
descending, first diagonal, and first obtuse mar-
ginal coronary arteries. Am Heart J 1996;131:
38–42.
364 ATAR
[88] Birnbaum Y, Sclarovsky S, Solodky A, et al. Pre-
diction of the level of left anterior coronary artery
obstruction during acute anterior wall myocardial
infarction by the admission electrocardiogram.
Am J Cardiol 1993;72:823–6.
[89] Arbane M, Goy JJ. Prediction of the site of total oc-
clusion in the left anterior descending coronary ar-
tery using admission electrocardiogram in anterior
wall acute myocardial infarction. Am J Cardiol
2000;85(4):487–91.
[90] Engelen DJ, Gorgels AP, Cheriex EC, et al. Value
of the electrocardiogram in localizing the occlusion
site in the left anterior descending coronary artery
in acute anterior myocardial infarction. J Am
Coll Cardiol 1999;34(2):389–95.
[91] Sasaki K, Yotsukura M, Sakata K, et al. Relation
of ST-segment changes in inferior leads during an-
terior wall acute myocardial infarction to length
and occlusion site of the left anterior descending
coronary artery. Am J Cardiol 2001;87(12):1340–5.
[92] Vasudevan K, Manjunath CN, Srinivas KH, et al.
Electrocardiographic localization of the occlusion
site in left anterior descending coronary artery in
acute anterior myocardial infarction. Indian Heart
J 2004;56(4):315–9.
[93] Birnbaum Y, Herz I, Solodky A, et al. Can we dif-
ferentiate by the admission ECG between anterior
wall acute myocardial infarction due to a left ante-
rior descending artery occlusion proximal to the or-
igin of the first septal branch and a postseptal
occlusion? American Journal of Noninvasive Car-
diology 1994;8:115–9.
[94] Ben-Gal T, Sclarovsky S, Herz I, et al. Importance
of the conal branch of the right coronary artery in
patients with acute anterior wall myocardial infarc-
tion: electrocardiographic and angiographic corre-
lation. J Am Coll Cardiol 1997;29(3):506–11.
[95] Shalev Y, Fogelman R, Oettinger M, et al. Does the
electrocardiographic pattern of ‘‘anteroseptal’’
myocardial infarction correlate with the anatomic
location of myocardial injury? Am J Cardiol
1995;75:763–6.
[96] Porter A, Wyshelesky A, Strasberg B, et al. Corre-
lation between the admission electrocardiogram
and regional wall motion abnormalities as detected
by echocardiography in anterior acute myocardial
infarction. Cardiology 2000;94(2):118–26.
[97] Sapin PM, Musselman DR, Dehmer GJ, et al. Im-
plications of inferior ST-segment elevation accom
panying anterior wall acute myocardial infarction
for the angiographic morphology of the left ante-
rior descending coronary artery morphology and
site of occlusion. Am J Cardiol 1992;69(9):860–5.
[98] Tamura A, Kataoka H, Nagase K, et al. Clinical
significance of inferior ST elevation during acute
anterior myocardial infarction. Br Heart J 1995;
74:611–4.
[99] Fletcher WO, Gibbons RJ, Clements IP. The
relationship of inferior ST depression, lateral ST
et al
elevation, and left precordial ST elevation to myo-
cardium at risk in acute anterior myocardial infarc-
tion. Am Heart J 1993;126(3 Pt 1):526–35.
[100] Haraphongse M, Tanomsup S, Jugdutt BI. Inferior
ST segment depression during acute anterior myo-
cardial infarction: clinical and angiographic corre-
lations. J Am Coll Cardiol 1984;4(3):467–76.
[101] Braat SH, Brugada P, den Dulk K, et al. Value of
lead V4R for recognition of the infarct coronary ar-
tery in acute inferior myocardial infarction. Am J
Cardiol 1984;53(11):1538–41.
[102] Kontos MC, Desai PV, Jesse RL, et al. Usefulness
of the admission electrocardiogram for identifying
the infarct-related artery in inferior wall acute
myocardial infarction. Am J Cardiol 1997;79:
182–4.
[103] Bairey CN, Shah PK, Lew AS, et al. Electrocardio-
graphic differentiation of occlusion of the left
circumflex versus the right coronary artery as
a cause of inferior acute myocardial infarction.
Am J Cardiol 1987;60:456–9.
[104] Herz I, Assali AR, Adler Y, et al. New electrocar-
diographic criteria for predicting either the right
or left circumflex artery as the culprit coronary ar-
tery in inferior wall acute myocardial infarction.
Am J Cardiol 1997;80:1343–5.
[105] Hasdai D, Birnbaum Y, Herz I, et al. ST segment
depression in lateral limb leads in inferior wall
acute myocardial infarction. Implications regard-
ing the culprit artery and the site of obstruction.
Eur Heart J 1995;16(11):1549–53.
[106] Kosuge M, Kimura K, Ishikawa T, et al. New elec-
trocardiographic criteria for predicting the site of
coronary artery occlusion in inferior wall acute
myocardial infarction. Am J Cardiol 1998;82(11):
1318–22.
[107] Assali AR, Herz I, Vaturi M, et al. Electrocardio-
graphic criteria for predicting the culprit artery in
inferior wall acute myocardial infarction. Am J
Cardiol 1999;84(1):87–9.
[108] Zimetbaum PJ, Krishnan S, Gold A, et al. Use-
fulness of ST-segment elevation in lead III ex-
ceeding that of lead II for identifying the
location of the totally occluded coronary artery
in inferior wall myocardial infarction. Am J Car-
diol 1998;81(7):918–9.
[109] Birnbaum Y, Hale SL, Kloner RA. Changes in R
wave amplitude: ECG differentiation between epi-
sodes of reocclusion and reperfusion associated
with ST-segment elevation. J Electrocardiol 1997;
30(3):211–6.
[110] Fiol M, Cygankiewicz I, Carrillo A, et al. Value of
electrocardiographic algorithm based on ‘‘ups and
downs’’ of ST in assessment of a culprit artery in
evolving inferior wall acute myocardial infarction.
Am J Cardiol 2004;94(6):709–14.
[111] Kopelman HA, Forman MB, Wilson BH, et al.
Right ventricular myocardial infarction in patients
with chronic lung disease: possible role of right
 ECG DIAGNOSIS OF STEMI
ventricular hypertrophy. J Am Coll Cardiol 1985;
5(6):1302–7.
[112] Erhardt LR, Sjogren A, Wahlberg I. Single right-
sided precordial lead in the diagnosis of right
ventricular involvement in inferior myocardial in-
farction. Am Heart J 1976;91(5):571–6.
[113] Zehender M, Kasper W, Kauder E, et al. Right ven-
tricular infarction as an independent predictor of
prognosis after acute inferior myocardial infarc-
tion. N Engl J Med 1993;328(14):981–8.
[114] Lopez-Sendon J, Coma-Canella I, Alcasena S, et al.
Electrocardiographic findings in acute right ven-
tricular infarction: sensitivity and specificity of
electrocardiographic alterations in right precordial
leads V4R, V3R, V1, V2, and V3. J Am Coll
Cardiol 1985;6(6):1273–9.
[115] Sadanandan S, Hochman JS, Kolodziej A, et al.
Clinical and angiographic characteristics of
patients with combined anterior and inferior ST-
segment elevation on the initial electrocardiogram
during acute myocardial infarction. Am Heart J
2003;146(4):653–61.
[116] Assali AR, Sclarobsky S, Herz I, et al. Comparison
of patients with inferior wall acute myocardial in-
farction with versus without ST-segment elevation
in leads V5 and V6. Am J Cardiol 1998;81:81–3.
[117] Golovchiner G, Matz I, Iakobishvili Z, et al. Corre-
lation between the electrocardiogram and regional
wall motion abnormalities as detected by echocar-
diography in first inferior acute myocardial infarc-
tion. Cardiology 2002;98(1–2):81–91.
[118] Ruddy TD, Yasuda T, Gold HK, et al. Correlations
of regional wall motion and myocardial perfusion in
patients with and without anterior precordial ST
segment depression during acute inferior myocar-
dial infarction. American Journal of Noninvasive
Cardiology 1987;1:81–7.
[119] Boden WE, Bough EW, Korr KS, et al. Inferosep-
tal myocardial infarction: another cause of precor-
dial ST-segment depression in transmural inferior
wall myocardial infarction? Am J Cardiol 1984;
54(10):1216–23.
[120] Haraphongse M, Jugdutt BI, Rossall RE. Signif-
icance of precordial ST-segment depression in
acute transmural inferior infarction: coronary
angiographic findings. Cathet Cardiovasc Diagn
1983;9(2):143–51.
[121] Hasdai D, Sclarovsky S, Solodky A, et al. Prognos-
tic significance of maximal precordial ST-segment
depression in right (V1 to V3) versus left (V4 to
V6) leads in patients with inferior wall acute myo-
cardial infarction. Am J Cardiol 1994;74:1081–4.
[122] Lew AS, Maddahi J, Shah PK, et al. Factors that
determine the direction and magnitude of precor-
dial ST-segment deviations during inferior wall
acute myocardial infarction. Am J Cardiol 1985;
55(8):883–8.
[123] Salcedo JR, Baird MG, Chambers RJ, et al. Signif-
icance of reciprocal S-T segment depression in
365
anterior precordial leads in acute inferior myocar-
dial infarction: concomitant left anterior descend-
ing coronary artery disease? Am J Cardiol 1981;
48(6):1003–8.
[124] Agarwal JB, Khaw K, Aurignac F, et al. Impor-
tance of posterior chest leads in patients with sus-
pected myocardial infarction, but nondiagnostic,
routine 12-lead electrocardiogram. Am J Cardiol
1999;83(3):323–6.
[125] Casas RE, Marriott HJ, Glancy DL. Value of leads
V7–V9 in diagnosing posterior wall acute myocar-
dial infarction and other causes of tall R waves in
V1–V2. Am J Cardiol 1997;80(4):508–9.
[126] Matetzky S, Freimark D, Feinberg MS, et al.
Acute myocardial infarction with isolated ST-
segment elevation in posterior chest leads V7–9:
‘‘hidden’’ ST-segment elevations revealing acute
posterior infarction. J Am Coll Cardiol 1999;
34(3):748–53.
[127] Wung SF, Drew BJ. New electrocardiographic
criteria for posterior wall acute myocardial is-
chemia validated by a percutaneous translumi-
nal coronary angioplasty model of acute
myocardial infarction. Am J Cardiol 2001;87
(8):970–4.
[128] Mager A, Sclarovsky S, Herz I, et al. Value of the
initial electrocardiogram in patients with inferior-
wall acute myocardial infarction for prediction of
multivessel coronary artery disease. Coron Artery
Dis 2000;11(5):415–20.
[129] Yamaji H, Iwasaki K, Kusachi S, et al. Predic-
tion of acute left main coronary artery obstruc-
tion by 12-lead electrocardiography. ST segment
elevation in lead aVR with less ST segment eleva-
tion in lead V(1). J Am Coll Cardiol 2001;38(5):
1348–54.
[130] Birnbaum Y, Goodman S, Barr A, et al. Com-
parison of primary coronary angioplasty versus
thrombolysis in patients with ST-segment eleva-
tion acute myocardial infarction and grade II
and grade III myocardial ischemia on the enroll-
ment electrocardiogram. Am J Cardiol 2001;
88(8):842–7.
[131] Lee CW, Hong M-K, Yang H-S, et al. Determi-
nants and prognostic implications of terminal
QRS complex distortion in patients treated with
primary angioplasty for acute myocardial infarc-
tion. Am J Cardiol 2001;88:210–3.
[132] Birnbaum Y, Herz I, Sclarovsky S, et al. Admission
clinical and electrocardiographic characteristics
predicting an increased risk for early reinfarction
after thrombolytic therapy. Am Heart J 1998;
135(5 Pt 1):805–12.
[133] Sejersten M, Birnbaum Y, Ripa RS, et al. Electro-
cardiographic identification of patients with ST-el-
evation acute myocardium infarction benefiting
most from primary angioplasty versus fibrinolysis:
results from the DANAMI-2 Trial. Circulation
2004;110 III-409.
 Cardiol Clin 24 (2006) 367–376

Electrocardiographic Markers
of Reperfusion in ST-elevation Myocardial Infarction
Shaul Atar, MD, Alejandro Barbagelata, MD,
Yochai Birnbaum, MD*
Division of Cardiology, University of Texas Medical Branch, 5.106 John Sealy Annex, 301 University Boulevard,
Galveston, TX 77555, USA

Reperfusion therapy with intravenous throm-
bolytic agents or percutaneous coronary interven-
tion (PCI) has emerged in the past 2 decades as an
effective means of reducing infarct size, preserving
ventricular function and topography, reducing
electrical instability, and reducing morbidity and
mortality in patients who have an acute ST-
elevation myocardial infarction (STEMI) [1,2].
Conversely, failure of reperfusion has been shown
to portend a substantial increase in morbidity and
mortality [3]. Because the outcome of patients
who fail to reperfuse with reperfusion therapy
may be improved with additional interventions
such as rescue PCI or additional pharmacologic
treatments, it becomes clinically important to rec-
ognize reperfusion or its failure at the bedside. In
contrast to experimental animal models of acute
myocardial infarction and reperfusion in which
the coronary artery is ligated for controlled pe-
riods of time, acute myocardial infarction in hu-
mans is a dynamic process with frequent repeat
episodes of coronary artery reperfusion and reoc-
clusion, both before and after the initiation of re-
perfusion therapy [4–7]. Although the extent of
myocardium involved can be estimated clinically
by physical examination (presence of heart failure,
tachycardia, hypotension, and other markers),
and with various imaging techniques (echocardi-
ography, radionuclide imaging, ventriculogra-
phy), there currently is no alternative to the
* Corresponding author.
E-mail address: yobirnba@utmb.edu (Y. Birnbaum).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.007
ECG for continuous assessment of the status of
coronary and myocardial perfusion. Coronary
angiography, Technetium-99m sestamibi single-
photon-emission CT imaging, and contrast echo-
cardiography can give only a snapshot of the
status of coronary or myocardial perfusion.
Although urgent coronary angiography can
distinguish an open from a closed culprit artery
effectively, and it remains the criterion for pa-
tency, its routine continuous application for this
purpose is seriously limited because of logistic
reasons, cost, invasive nature with attendant risk
of peri-access complications, the snapshot nature
of angiographic evaluation, and the fact that
epicardial vessel patency may exist despite lack
of nutritive flow at the level of downstream
microcirculation. For example, in numerous ani-
mal models it has been repeatedly demonstrated
that immediately after opening of the occluded
coronary artery there is a hyperemic phase,
followed later by gradual decline of myocardial
perfusion, even though the epicardial coronary
artery remains open. This phenomenon of ‘‘no
reflow’’ is currently undetected by angiograms
performed immediately after recanalization of the
infarct-related artery (IRA) [8,9]. Therefore sev-
eral investigators have evaluated a number of
noninvasive nonangiographic markers to deter-
mine the success or failure of reperfusion. Among
these techniques, ECG monitoring is most suit-
able for routine bedside application. This article
reviews the role of bedside 12-lead ECG in identi-
fying and monitoring the perfusion state of the
myocardium in STEMI.
cardiology.theclinics.com
368 ATAR et al

ECG markers of reperfusion minimal ST elevation, whereas in cases of exten-

There are four ECG markers for prediction of
the perfusion status of the ischemic myocardium:
(1) ST-segment measurements, (2) T-wave config-
uration, (3) QRS changes, and (4) reperfusion
arrhythmias.
ST resolution
Several studies showed that recanalization of
the IRA results in rapid resolution (R50%) of ST
elevation (Fig. 1) [10–14]. These results were ob-
tained from serial ECG recordings performed on
admission of the patient to the hospital and at
various time intervals after initiation of therapy.
Unfortunately, these studies were not unified
regarding the definition of ST resolution (STR)
[15] or the timing of coronary angiography and
final ECG assessment [12–14,16]. Some of the
studies [11,12,15,16] assessed a single ECG lead
with maximal ST elevation, whereas others
[10,13] have assessed the reduction in the sum of
ST elevation in all 12 leads. It seems that the latter
method is more useful in patients who have
sive ST elevation, assessing the reduction of ST
in the single lead with maximal ST elevation is
preferable [5,17].
Because reperfusion is a dynamic process, in
which the IRA may recanalize and reocclude
intermittently [7,18], serial ECG recording is lim-
ited in predicting the state of reperfusion. More-
over, one third of episodes of recurrent ST
elevation are silent [7]. Thus, unless recorded con-
tinuously, these ECG changes may be missed
completely by serial intermittent ECG recordings
or misinterpreted as signs of improvement, should
the re-elevation in ST be smaller than that in the
initial ECG recording (Fig. 2) [5]. Moreover, be-
cause some investigators have suggested that the
ECG criterion for reperfusion is 50% or greater
STR compared with maximal ST elevation at
any time-point (not necessarily the enrollment
ECG), without continuous ECG monitoring start-
ing immediately upon admission, interpretation
of STR relative to the enrollment ECG may be
misleading. A proper alternative to continuous
12-lead recording would be ECG monitoring that

Fig. 1. (A) A 48-year-old woman with 3 hours of chest pain. The admission ECG shows ST elevation in aVL and V1 to
V3 and reciprocal ST depression in the inferior leads. (B) After receiving aspirin and nitroglycerin, the patient developed
ventricular fibrillation and was defibrillated. The ECG shows an increase in S-wave amplitude in V3 to V6, ST elevation
resolution in V1 to V3, and junctional ST depression in leads V3 to V6 with tall, upright T waves. (C) Seventy minutes
later, the patient had no chest pain. Repeat ECG shows resolution of ST elevation in leads aVL and V1 to V3 and less ST
depression in the inferior leads. There is now mild ST depression in leads V4 to V6; however, T-wave amplitude in the
precordial leads has decreased. (D) On the next day, after PCI, the ECG shows isoelectric ST segments with T-wave
inversion in leads I, aVL, and V1 to V6.
 ECG MARKERS OF REPERFUSION IN STEMI 369

Fig. 2. Serial ECG tracings of a patient who had inferoposterior STEMI. (a) Before initiation of thrombolytic therapy
there is ST elevation in leads II, III, and aVF and ST depression in leads I, aVL, and V2 to V4.(b) Sixty minutes after
initiation of thrombolytic therapy, pain subsided, and there is 70% or greater STR in the inferior leads. (c) Fifteen min-
utes later, there is ST re-elevation relative to tracing B, but the ST elevation is less than 50% of the initial values. The
patient was referred to rescue PCI. (d) After PCI with stent implantation in the right coronary artery, there is 70% or
greater STR with inversion of the T waves in leads III and aVF. (Adapted from Vaturi M, Birnbaum Y. The use of the
electrocardiogram to identify epicardial coronary and tissue reperfusion in acute myocardial infarction. J Thromb-
Thrombolysis 2000;10:140; with permission.)

engages computer-assisted ST-segment analysis ‘‘reperfusion syndrome,’’ is a frequently noted

and continuous 12-lead recording (using either
the single lead or the sum of ST elevation)
[19]. Another technique, continuous vectorcardio-
graphic monitoring, assesses online both the QRS
complex vector and the ST elevation vector simul-
taneously [20,21].
Five distinct patterns of ST segment evolution
were identified by using continuous ECG record-
ing: (1) rapid STR without re-elevation, (2) rapid
STR following a delayed ST re-elevation, (3)
persistent ST elevation without STR, (4) rapid
STR followed by rapid ST re-elevation, and (5)
a delayed ST elevation peak followed by a rapid
STR and recurrent ST elevation [18]. Whereas the
first three patterns indicate the status of the infarct
related artery, the latter two patterns are less spe-
cific regarding patency, because they suggest an
unstable myocardial tissue perfusion (regardless
of the IRA patency). Krucoff and colleagues [18]
reported that absence of STR or presence of ST
re-elevation at the time of coronary angiography
predicts an occluded IRA with a sensitivity of
90% and specificity of 92%. Dellborg and col-
leagues [20,21] reported the role of changes in
ST vectors in STEMI. The sensitivity of vector
changes to predict IRA patency was 81% to
94%, and the specificity was 70% to 80%.
Additional elevation of the sum of ST of 5 mm
or more during reperfusion, also termed
phenomenon, and most studies found it to be
a marker of impaired microvascular reperfusion,
with lower coronary velocity reserve, reduced left
ventricular function, and larger infarct size
[22,23]. The significance of the reperfusion syn-
drome is still debatable, however, because others
have reported it to be a favorable prognostic
factor [24].
The Global Utilization of Streptokinase and
Tissue Plasminogen Activator for Occluded
Coronary Arteries (GUSTO-1) ECG ischemia-
monitoring substudy [25] studied 1067 patients
divided into three groups: 460 patients were mon-
itored with vector-derived 12-lead ECG, 373 with
12-lead ECG, and 288 with a three-lead Holter
system. In this study, 50% STR before either 90-
or 180-minutes’ angiography was considered to
be a sign of reperfusion. Recurrence of ST eleva-
tion was considered to be reocclusion or reische-
mia. To unify the study, only a single-lead ST
trend was considered. The addition of initial
peak ST levels, the time to 50% STR, and various
STR patterns improved the predictive accuracy of
the IRA patency. This study reinforced the notion
that 50% or greater STR within 90 minutes from
starting thrombolysis reflected patency of the
IRA. They also found that the amount of ST
elevation present during recording is a major de-
terminant of the accuracy of patency prediction.
370 ATAR
The accuracy of prediction of IRA patency corre-
lated with the degree of initial ST elevation. The
absence of STR does not accurately predict an oc-
cluded IRA, however, because approximately
50% of patients with no (!30%) STR still have
a patent IRA [26,27]. Previously, the absence of
STR despite a patent IRA had been considered
to be a false-negative sign of lake of reperfusion
by the 12-lead ECG.
Important differences exist between anterior and
inferior STEMI with regard to STR [13,14,26,28–
30]. Patients who have anterior STEMI develop
significantly less STR than those who have inferior
infarction, despite only modest differences in
epicardial blood flow, suggesting that STR is a
less accurate predictor of epicardial reperfusion
among patients who have anterior versus inferior
STEMI [13,26,28,29]. This reduced accuracy may
result from technical factors, such as the frequent
(normal) presence of J-point elevation in the ante-
rior precordial leads [31], which would serve
to decrease the extent of STR that is possible.
Additionally, anterior STEMI typically is associ-
ated with a larger infarct size and greater tissue in-
jury than inferior STEMI. As a result, different
threshold levels of STR may be appropriate for
anterior versus inferior STEMI [26]. When sensitiv-
ity analyses are performed, 70% or greater STR
seems to be the optimal threshold for patients
who have inferior myocardial infarction, whereas
50% or greater STR may be optimal for anterior
STEMI [26].
To detect STR, the recording must be started
as early as possible (preferably before the initia-
tion of thrombolytic therapy). Otherwise, the first
episode of 50% or greater ST recovery may not be
properly recorded, which may result in false
assessment of the IRA patency (Fig. 2). ST mon-
itoring then should continued, preferably for
at least 3 hours, as suggested by Schroder and
colleagues [32], who found that the prognostic
significance of incomplete STR was better at
180 minutes after the initiation of streptokinase
therapy (30-day cardiac mortality, 13.6%) than
at 90 minutes (30-day cardiac mortality, 7.3%).
Use of ST resolution for evaluation
of myocardial tissue reperfusion
Myocardial reperfusion is not a dichotomous
phenomenon. Epicardial coronary flow is graded
by the Thrombolysis in Myocardial Infarction
(TIMI) flow classification. Myocardial tissue
perfusion may be complete, partial, or absent,
et al
however, irrespective of the epicedial coronary
TIMI flow grade [30,33]. Analysis of STR may
give a better estimation of myocardial tissue per-
fusion over time [34,35]. Complete (R70%) STR
is associated with better outcome and preserva-
tion of left ventricular function than partial
(30% to 70%) or no (!30%) STR [28,32,35,36].
Thus, whereas 50% or greater resolution of ST el-
evation is a reliable indicator of patency of the
IRA, only complete (R70%) STR is an indicator
of restoration of myocardial tissue perfusion. The
same findings apply for patients who undergo suc-
cessful recanalization of the IRA by primary PCI,
in whom early resolution of ST elevation predicts
better outcome than incomplete or no resolution
of ST elevation [35,37]. Recent studies have shown
that the combination of STR with angiographic
parameters, such as the myocardial blush (MB)
and the corrected TIMI frame count, allow better
grading of microvascular reperfusion and better
prediction of cardiovascular events 30 days and
6 months after primary PCI for STEMI [38,39].
Poli and colleagues [38] combined MB and sum
of STR and identified three main groups of
patients: group 1 (n ¼ 60) had both significant
MB (grade 2 to 3) and STR (O50% versus base-
line) and had a high rate of 7-day (65%) and
6-month (95%) left ventricular functional recov-
ery. Group 2 (n ¼ 21) showed MB but persistent
ST elevation, and the prevalence of early left ven-
tricular functional recovery was low (24%) but
increased up to 86% in the late phase. Group 3
(n ¼ 28) had neither significant MB nor STR
and had poor early (18%) and late (32%) left ven-
tricular functional recovery. Thus the addition of
STR to angiographic parameters provides better
determination of myocardial reperfusion.
T-wave configuration
Early inversion of the terminal portion of the
T waves after initiation of reperfusion therapy, as
shown in Figs. 1–3, is an indicator of successful
reperfusion [40]. Moreover, a study by Corbalan
and colleagues [41], in which inversion of T waves
more than 0.5 mm below the baseline within the
first 24 hours of thrombolytic therapy in all the
leads with previous ST elevation was considered
a marker for coronary artery reperfusion, showed
that T-wave inversion was associated with the
lowest in-hospital mortality rate (odds ratio,
0.25; 95% confidence interval, 0.10–0.56) [41].
When all markers of coronary artery reperfusion
(resolution of chest pain, STR greater than 50%
 ECG MARKERS OF REPERFUSION IN STEMI 371

Fig. 3. A 73-year-old man presenting with 2 hours of chest pain. (A) The admission ECG shows ST elevation in leads V1
to V5. (B) Repeat ECG done 24 minutes later shows the same ST elevation, but now S waves appear in leads V2 to V6,
without inversion of the T waves. Immediate angiography showed proximal left anterior descending artery stenosis, with
thrombus and TIMI grade 2 flow. (C) An ECG done 2 hours and 50 minutes after successful PCI with stent insertion
with resulting TIMI grade 3 flow, showing complete (O70%) STR, QS pattern in leads V1 to V3, shortening of the
R wave in leads V3 to V6, and T-wave inversion in leads V1 to V4.

at 90 minutes, abrupt creatine kinase rise before
12 hours, and T-wave inversion) were included
in a logistic regression model, T-wave inversion
(odds ratio; 0.29; 95% confidence interval, 0.11–
0.68) and abrupt creatine kinase rise (odds ratio,
0.36; 95% confidence interval, 0.16–0.77) contin-
ued to be significantly associated with better out-
come, whereas STR was not [41].
Only a few studies have investigated the
significance of T-wave direction in leads with ST
elevation. Herz and colleagues [42] found that, be-
fore initiation of thrombolytic therapy, negative T
waves in leads with ST elevation were associated
with better prognosis in patients enrolled within
2 hours of symptoms onset, whereas in those en-
rolled 2 to 6 hours after initiation of symptoms
372 ATAR
negative T waves were associated with increased
mortality. At 90 minutes after initiation of strep-
tokinase therapy, TIMI grade 3 flow in the IRA
was more commonly seen in patients who did
not have T-wave inversion (50%) than in those
who had T-wave inversion on the pretreatment
ECG (30%; P ¼ .002) [43]. Among patients
treated within 3 hours of onset of symptoms,
TIMI grade 3 flow was seen in 62% of those with-
out versus 43% of those with T-wave inversion
(P ¼ .06). Among patients treated after 3 hours,
TIMI grade 3 flow was seen in 38% of those
who did not have T-wave inversion versus 23%
of those who had T-wave inversion (P ¼ .05)
[43]. After initiation of thrombolytic therapy,
early inversion of the T waves may be a sign of
reperfusion [41,44]. Negative T waves on the pre-
discharge ECG of patients who have anterior
STEMI, especially if associated with complete res-
olution of the ST elevation, is a sign of a relatively
small infarct size with preserved left ventricular
ejection fraction [45]. During the following
months, however, early spontaneous normaliza-
tion of the T waves in the involved leads may
be associated with better outcome and preserva-
tion of left ventricular function [46]. Therefore,
the configuration of the T waves may carry dif-
ferent meanings at different stages after STEMI.
Moreover, it is unclear whether partial inver-
sion of the terminal portion of the T waves has
the same significance as complete or giant
T-wave inversion [47]. Furthermore, the exact
underlying mechanisms and significance of vari-
ous patterns of STR and T-wave inversion have
not been studied. It is well known that the ampli-
tude of the STR is influenced mostly by epicardial
ischemia and is less influenced by the degree of
subendocardial ischemia. Thus, STR may corre-
late better with amelioration of epicardial ische-
mia caused by restoration of flow through the
IRA or by recruitment of collaterals and less with
the status of the subendocardial zones. It is possi-
ble that the configuration of the T waves is related
more to the subendocardial perfusion status. Be-
cause myocardial necrosis starts from the subendo-
cardium and expands toward the epicardium, the
configuration of the T waves after reperfusion ther-
apy may correlate better with recovery of left ven-
tricular function and prognosis [41].
QRS changes during ischemia and reperfusion
Dynamic changes in the QRS complex are
detected during reperfusion therapy for STEMI,
et al
as shown in Fig. 3. These changes have been
investigated mainly by vectorcardiography [20,21,
48]. It seems that the QRS vector changes are
less specific than the ST-vector changes for pre-
dicting reperfusion [20]. Using standard 12-lead
ECG, dynamic changes in Q-wave number, ampli-
tude, and width, R-wave amplitude and S-wave
appearance are detected. Some have reported
that early pathologic Q waves develop especially
after reperfusion [49,50]; however, others have
found these to be associated with larger ischemic
zone and ultimate necrotic area [51–53]. It has
been reported that the early appearance of Q
waves (within !6 hours of symptom onset) does
not signify irreversible damage and does not pre-
clude myocardial salvage by thrombolytic therapy
[53]; however, Q waves on admission are asso-
ciated with worse prognosis [54]. It is unclear
whether dynamic changes in Q waves early after
initiation of reperfusion therapy have additive
prognostic significance to ST monitoring and
T-wave configuration. It generally is accepted
that the loss of R waves and the appearance of
new Q waves in the days following STEMI repre-
sent myocardial necrosis [55]. During the first
48 hours of STEMI, however, a recovery of R
wave and disappearance of new Q waves can be
detected even in patients not undergoing reper-
fusion therapy [56]. This phenomenon usually is
confined to small STEMI [57].
In an open-chest rabbit model, episodes of ST
elevation caused by coronary artery occlusion
were associated with an increase in R-wave
amplitude, whereas ST-segment elevation during
reperfusion episodes was associated with a de-
crease in R-wave amplitude [58]. Absence of
S waves in leads V1 to V3 in the enrollment ECG
of patients who have anterior STEMI is associ-
ated with increased mortality, larger final infarct
size, higher rates of no reflow or no STR, and
less benefit from thrombolytic therapy [59–61].
During thrombolytic therapy, S waves in these
leads may increase or decrease in size and even
disappear (see Fig. 3). It is unclear whether
decrease in S-wave amplitude is a marker of more
severe ischemia and whether the reappearance of
an S wave is a sign of reperfusion.
Reperfusion arrhythmias
Accelerated idioventricular rhythm
Nonsustained or sustained ventricular tachy-
cardia at rates of less than or equal to 120 beats
per minute, also called accelerated idioventricular
 ECG MARKERS OF REPERFUSION IN STEMI
rhythm (AIVR), is a common arrhythmia in
patients who have STEMI. Several studies have
shown that reperfusion is accompanied by AIVR
in up to 50% of patients, especially when contin-
uous or frequent ECG monitoring is used [14,62].
Classic AIVR has been defined as a ventricular
rhythm occurring at 50 to 120 beats per minute
starting after a long pause resulting in a long cou-
pling interval. This rhythm is usually regular and
is terminated by the capture of ventricle by the si-
nus node. In a prospective study of 87 patients re-
ceiving intravenous or intracoronary thrombolysis
for STEMI, Gorgels and colleagues [63] showed
that classic AIVR occurred in 50% of patients
with reperfusion and in only 7% of patients with-
out reperfusion. AIVR is a specific but relatively
insensitive indicator of reperfusion occurring in
only in 50% of reperfused patients.
Cardioinhibitory (Bezold-Jarisch) reflex
Several investigators have shown that sudden
appearance of sinus bradycardia accompanied by
hypotension can signal reperfusion of the artery
supplying inferior wall of the myocardium, (ie, in
most instances, the right coronary artery). This
phenomenon is believed to be a type of Bezold-
Jarisch reflex provoked by stimulation of cardiac
baroreceptors with increased vagal input and
withdrawal of sympathetic tone [64]. This phe-
nomenon is observed in 23% to 65% of cases of
right coronary artery reperfusion, thus providing
corroborative evidence of reperfusion [14,65].
Signal-averaged electrocardiography
for detection of late potentials
Signal-averaged electrocardiography (SAECG)
has been used to detect late potentials as markers
of increased vulnerability for inducible and spon-
taneous ventricular arrhythmias following acute
STEMI. Several studies have demonstrated that in
patients who have acute STEMI a patent IRA is
associated with a reduced frequency of positive
late potentials compared with patients who have
persistent occlusion [66,67]. Tranchesi and col-
leagues [68] have examined the significance of
late potentials in SAECG as a marker of reperfu-
sion. In 54 patients who had acute STEMI and
an angiographically documented occlusion, a base-
line SAECG was recorded before initiation of
thrombolysis. Coronary angiogram and SAECG
were recorded again 90 minutes after thrombo-
lytic infusion. In 50% of the patients who had suc-
cessful reperfusion the late potentials disappeared
after reperfusion (from 16/35 to 8/35; P ¼ .03),
373
whereas in patients who had a closed artery there
was no change in the prevalence of positive late
potentials (8/19 before to 7/19 after attempted
but failed thrombolysis). These preliminary find-
ings, although interesting, demonstrate the limited
accuracy of late potentials and their changes fol-
lowing thrombolytic therapy for the bedside diag-
nosis of reperfusion.
Summary
At present, bedside recognition of reperfusion
in patients presenting with acute STEMI can be
accomplished best by assessment of several objec-
tive and subjective signs of termination of ische-
mia (ie, resolution of chest pain or rapid STR).
A study by Oude Ophuis and colleagues [69] of
230 patients who had STEMI suggested that the
combination of ECG and clinical markers may
better predict the patency of the IRA and the sta-
tus of myocardial reperfusion. They found that
a sudden decrease in chest pain was the most com-
mon sign of reperfusion (36%), followed by STR
of 50% or more (30%), and the development of a
terminal negative T wave (20%) in the lead with
the highest ST elevation. STR of 50% or more
and the appearance of AIVR had the highest pos-
itive predictive value for reperfusion. For TIMI
grade 3 flow, the positive predictive value of
STR was 66% and for AIVR it was 59%. The
presence of three or more noninvasive markers
of reperfusion predicted TIMI grade 3 flow accu-
rately in 80% of cases.
Because ST segments may fluctuate dramati-
cally before and during thrombolytic therapy, an
accurate determination of progressive decrease
(by R 50%) relative to the highest ST elevation
requires frequent (every 5–15 minutes) or contin-
uous monitoring of ST (in either a selected lead or
all 12 leads). Although other bedside signs such as
AIVR and Bezold-Jarisch reflex also indicate
reperfusion, their limited sensitivity restricts their
usefulness. Biochemical markers related to accel-
erated washout associated with reperfusion,
although promising, are still limited in their use-
fulness because the results are difficult to obtain in
a timely fashion. Acute coronary angiography,
although useful, is not practical, and it may turn
out not to be the reference standard for reperfu-
sion. Because the goal of reperfusion is to achieve
termination of ongoing ischemia, noninvasive
markers of ischemia termination may be a better
standard than the anatomic evidence obtained by
coronary angiography. The favorable prognostic
374 ATAR
impact of early STR in reperfusion trials supports
the clinical relevance and importance of signs of
ischemia termination.
It is necessary, however, to improve the un-
derstanding of the pathophysiologic mechanisms
leading to the ECG changes during reperfusion,
namely the significance of STR, T-wave configu-
ration, and early and terminal QRS complex
changes. Better understanding of the pathophys-
iology may help in the design of studies to
examine specific interventions (ie, intravenous
glycoprotein IIb/IIIa inhibitors, clopidogrel, ni-
trates, adenosine, and other drugs) that may be
beneficial in patients who have not reached
complete ECG signs of reperfusion (STR with
complete T-wave inversion).
References
[1] Gruppo Italiano per lo Studio della Sopravvivenza
nell’Infarto Miocardico. GISSI-2. A factorial rando-
mised trial of alteplase versus streptokinase and hep-
arin versus no heparin among 12,490 patients
with acute myocardial infarction. Lancet 1990;
336(8707):65–71.
[2] Third International Study of Infarct Survival Collab-
orative Group. ISIS-3. A randomised comparison
of streptokinase vs tissue plasminogen activator vs
anistreplase and of aspirin plus heparin vs aspirin
alone among 41,299 cases of suspected acute myocar-
dial infarction. Lancet 1992;339(8796):753–70.
[3] The GUSTO Angiographic Investigators. The ef-
fects of tissue plasminogen activator, streptokinase,
or both on coronary-artery patency, ventricular
function, and survival after acute myocardial infarc-
tion. N Engl J Med 1993;329(22):1615–22.
[4] Haider AW, Andreotti F, Hackett DR, et al. Early
spontaneous intermittent myocardial reperfusion
during acute myocardial infarction is associated
with augmented thrombogenic activity and less myo-
cardial damage. J Am Coll Cardiol 1995;26(3):662–7.
[5] Klootwijk P, Cobbaert C, Fioretti P, et al. Noninva-
sive assessment of reperfusion and reocclusion after
thrombolysis in acute myocardial infarction. Am J
Cardiol 1993;72(19):75G–84G.
[6] Krucoff MW, Croll MA, Pope JE, et al. Continuous
12-lead ST-segment recovery analysis in the TAMI
7 study. Performance of a noninvasive method for
real-time detection of failed myocardial reperfusion.
Circulation 1993;88(2):437–46.
[7] Kwon K, Freedman SB, Wilcox I, et al. The unstable
ST segment early after thrombolysis for acute infarc-
tion and its usefulness as a marker of recurrent cor-
onary occlusion. Am J Cardiol 1991;67(2):109–15.
[8] Kloner RA, Ganote CE, Jennings RB. The ‘‘no-
reflow’’ phenomenon after temporary coronary oc-
clusion in the dog. J Clin Invest 1974;54(6):1496–508.
et al
[9] Reffelmann T, Kloner RA. The ‘‘no-reflow’’ phe-
nomenon: basic science and clinical correlates. Heart
2002;87(2):162–8.
[10] Clemmensen P, Ohman EM, Sevilla DC, et al.
Changes in standard electrocardiographic ST-
segment elevation predictive of successful reper-
fusion in acute myocardial infarction. Am J Cardiol
1990;66(20):1407–11.
[11] Hogg KJ, Hornung RS, Howie CA, et al. Electrocar-
diographic prediction of coronary artery patency af-
ter thrombolytic treatment in acute myocardial
infarction: use of the ST segment as a non-invasive
marker. Br Heart J 1988;60(4):275–80.
[12] Saran RK, Been M, Furniss SS, et al. Reduction in
ST segment elevation after thrombolysis predicts
either coronary reperfusion or preservation of left
ventricular function. Br Heart J 1990;64(2):113–7.
[13] Barbash GI, Roth A, Hod H, et al. Rapid resolution
of ST elevation and prediction of clinical outcome in
patients undergoing thrombolysis with alteplase
(recombinant tissue-type plasminogen activator):
results of the Israeli Study of Early Intervention
in Myocardial Infarction. Br Heart J 1990;64(4):
241–7.
[14] Shah PK, Cercek B, Lew AS, et al. Angiographic
validation of bedside markers of reperfusion. J Am
Coll Cardiol 1993;21(1):55–61.
[15] Kircher BJ, Topol EJ, O’Neill WW, et al. Prediction
of infarct coronary artery recanalization after intra-
venous thrombolytic therapy. Am J Cardiol 1987;
59(6):513–5.
[16] Richardson SG, Morton P, Murtagh JG, et al. Rela-
tion of coronary arterial patency and left ventricular
function to electrocardiographic changes after strep-
tokinase treatment during acute myocardial infarc-
tion. Am J Cardiol 1988;61(13):961–5.
[17] Syed MA, Borzak S, Asfour A, et al. Single lead ST-
segment recovery: a simple, reliable measure of suc-
cessful fibrinolysis after acute myocardial infarction.
Am Heart J 2004;147(2):275–80.
[18] Krucoff MW, Croll MA, Pope JE, et al. Continu-
ously updated 12-lead ST-segment recovery analysis
for myocardial infarct artery patency assessment
and its correlation with multiple simultaneous early
angiographic observations. Am J Cardiol 1993;
71(2):145–51.
[19] Krucoff MW, Wagner NB, Pope JE, et al. The
portable programmable microprocessor-driven real-
time 12-lead electrocardiographic monitor: a prelim-
inary report of a new device for the noninvasive
detection of successful reperfusion or silent coronary
reocclusion. Am J Cardiol 1990;65(3):143–8.
[20] Dellborg M, Topol EJ, Swedberg K. Dynamic QRS
complex and ST segment vectorcardiographic mon-
itoring can identify vessel patency in patients with
acute myocardial infarction treated with reperfusion
therapy. Am Heart J 1991;122(4 Pt 1):943–8.
[21] Dellborg M, Steg PG, Simoons M, et al. Vectorcar-
diographic monitoring to assess early vessel patency
 ECG MARKERS OF REPERFUSION IN STEMI
after reperfusion therapy for acute myocardial in-
farction. Eur Heart J 1995;16(1):21–9.
[22] Feldman LJ, Himbert D, Juliard JM, et al. Reperfu-
sion syndrome: relationship of coronary blood flow
reserve to left ventricular function and infarct size.
J Am Coll Cardiol 2000;35(5):1162–9.
[23] Yokoshiki H, Kohya T, Tateda K, et al. Abrupt aug-
mentation of ST segment elevation associated with
successful reperfusion: a sign of diminished myocar-
dial salvage. Am Heart J 1995;130(4):698–704.
[24] Shechter M, Rabinowitz B, Beker B, et al. Addi-
tional ST segment elevation during the first hour of
thrombolytic therapy: an electrocardiographic sign
predicting a favorable clinical outcome. J Am Coll
Cardiol 1992;20(7):1460–4.
[25] Klootwijk P, Langer A, Meij S, et al. Non-invasive
prediction of reperfusion and coronary artery pa-
tency by continuous ST segment monitoring in the
GUSTO-I trial. Eur Heart J 1996;17(5):689–98.
[26] de Lemos JA, Antman EM, Giugliano RP, et al.
ST-segment resolution and infarct-related artery
patency and flow after thrombolytic therapy.
Thrombolysis in Myocardial Infarction (TIMI) 14
investigators. Am J Cardiol 2000;85(3):299–304.
[27] Zeymer U, Schroder R, Tebbe U, et al. Non-invasive
detection of early infarct vessel patency by resolu-
tion of ST-segment elevation in patients with throm-
bolysis for acute myocardial infarction; results of the
angiographic substudy of the Hirudin for Improve-
ment of Thrombolysis (HIT)-4 trial. Eur Heart J
2001;22(9):769–75.
[28] Schroder R, Dissmann R, Bruggemann T, et al. Ex-
tent of early ST segment elevation resolution: a sim-
ple but strong predictor of outcome in patients with
acute myocardial infarction. J Am Coll Cardiol
1994;24(2):384–91.
[29] Schroder R, Wegscheider K, Schroder K, et al. Ex-
tent of early ST segment elevation resolution:
a strong predictor of outcome in patients with acute
myocardial infarction and a sensitive measure to
compare thrombolytic regimens. A substudy of the
International Joint Efficacy Comparison of Throm-
bolytics (INJECT) trial. J Am Coll Cardiol 1995;
26(7):1657–64.
[30] Matetzky S, Freimark D, Chouraqui P, et al. The
distinction between coronary and myocardial reper-
fusion after thrombolytic therapy by clinical
markers of reperfusion. J Am Coll Cardiol 1998;
32(5):1326–30.
[31] Willems JL, Willems RJ, Willems GM, et al. Signifi-
cance of initial ST segment elevation and depression
for the management of thrombolytic therapy in acute
myocardial infarction. European Cooperative Study
Group for Recombinant Tissue-Type Plasminogen
Activator. Circulation 1990;82(4):1147–58.
[32] Schroder R, Zeymer U, Wegscheider K, et al. Com-
parison of the predictive value of ST segment elevation
resolution at 90 and 180 min after start of streptoki-
nase in acute myocardial infarction. A substudy of
375
the hirudin for improvement of thrombolysis
(HIT)-4 study. Eur Heart J 1999;20(21):1563–71.
[33] van ’t Hof AW, Liem A, Suryapranata H, et al. An-
giographic assessment of myocardial reperfusion in
patients treated with primary angioplasty for acute
myocardial infarction: myocardial blush grade.
Zwolle Myocardial Infarction Study Group. Circu-
lation 1998;97(23):2302–6.
[34] Santoro GM, Valenti R, Buonamici P, et al. Rela-
tion between ST-segment changes and myocardial
perfusion evaluated by myocardial contrast echocar-
diography in patients with acute myocardial infarc-
tion treated with direct angioplasty. Am J Cardiol
1998;82(8):932–7.
[35] van ’t Hof AW, Liem A, de Boer MJ, et al. Clinical
value of 12-lead electrocardiogram after successful
reperfusion therapy for acute myocardial infarction.
Zwolle Myocardial Infarction Study Group. Lancet
1997;350(9078):615–9.
[36] de Lemos JA, Braunwald E. ST segment resolution
as a tool for assessing the efficacy of reperfusion ther-
apy. J Am Coll Cardiol 2001;38(5):1283–94.
[37] Matetzky S, Novikov M, Gruberg L, et al. The sig-
nificance of persistent ST elevation versus early res-
olution of ST segment elevation after primary
PTCA. J Am Coll Cardiol 1999;34(7):1932–8.
[38] Poli A, Fetiveau R, Vandoni P, et al. Integrated
analysis of myocardial blush and ST-segment eleva-
tion recovery after successful primary angioplasty:
real-time grading of microvascular reperfusion and
prediction of early and late recovery of left ventricu-
lar function. Circulation 2002;106(3):313–8.
[39] Haager PK, Christott P, Heussen N, et al. Prediction
of clinical outcome after mechanical revasculariza-
tion in acute myocardial infarction by markers of
myocardial reperfusion. J Am Coll Cardiol 2003;
41(4):532–8.
[40] Doevendans PA, Gorgels AP, van der Zee R, et al.
Electrocardiographic diagnosis of reperfusion dur-
ing thrombolytic therapy in acute myocardial infarc-
tion. Am J Cardiol 1995;75(17):1206–10.
[41] Corbalan R, Prieto JC, Chavez E, et al. Bedside
markers of coronary artery patency and short-term
prognosis of patients with acute myocardial infarc-
tion and thrombolysis. Am Heart J 1999;138(3 Pt 1):
533–9.
[42] Herz I, Birnbaum Y, Zlotikamien B, et al. The prog-
nostic implications of negative T waves in the leads
with ST segment elevation on admission in acute
myocardial infarction. Cardiology 1999;92(2):121–7.
[43] Wong CK, French JK, Aylward PE, et al. Useful-
ness of the presenting electrocardiogram in predict-
ing successful reperfusion with streptokinase in
acute myocardial infarction. Am J Cardiol 1999;
83(2):164–8.
[44] Matetzky S, Barabash GI, Shahar A, et al. Early
T wave inversion after thrombolytic therapy predicts
better coronary perfusion: clinical and angiographic
study. J Am Coll Cardiol 1994;24(2):378–83.
376 ATAR
[45] Adler Y, Zafrir N, Ben-Gal T, et al. Relation be-
tween evolutionary ST segment and T-wave direc-
tion and electrocardiographic prediction of
myocardial infarct size and left ventricular function
among patients with anterior wall Q-wave acute
myocardial infarction who received reperfusion
therapy. Am J Cardiol 2000;85(8):927–33.
[46] Tamura A, Nagase K, Mikuriya Y, et al. Signifi-
cance of spontaneous normalization of negative
T waves in infarct-related leads during healing of
anterior wall acute myocardial infarction. Am J
Cardiol 1999;84(11):1341–4.
[47] Agetsuma H, Hirai M, Hirayama H, et al. Transient
giant negative T wave in acute anterior myocardial
infarction predicts R wave recovery and preservation
of left ventricular function. Heart 1996;75(3):229–34.
[48] Dellborg M, Riha M, Swedberg K. Dynamic QRS-
complex and ST-segment monitoring in acute myo-
cardial infarction during recombinant tissue-type
plasminogen activator therapy. The TEAHAT
Study Group. Am J Cardiol 1991;67(5):343–9.
[49] Goldberg S, Urban P, Greenspon A, et al. Limita-
tion of infarct size with thrombolytic agents–electro-
cardiographic indexes. Circulation 1983;68(2 Pt 2):
I77–82.
[50] Rechavia E, Blum A, Mager A, et al. Electrocardio-
graphic Q-waves inconstancy during thrombolysis in
acute anterior wall myocardial infarction. Cardiol-
ogy 1992;80(5–6):392–8.
[51] Raitt MH, Maynard C, Wagner GS, et al. Relation
between symptom duration before thrombolytic
therapy and final myocardial infarct size. Circula-
tion 1996;93(1):48–53.
[52] Raitt MH, Maynard C, Wagner GS, et al. Appear-
ance of abnormal Q waves early in the course of
acute myocardial infarction: implications for effi-
cacy of thrombolytic therapy. J Am Coll Cardiol
1995;25(5):1084–8.
[53] Bar FW, Vermeer F, de Zwaan C, et al. Value of ad-
mission electrocardiogram in predicting outcome of
thrombolytic therapy in acute myocardial infarc-
tion. A randomized trial conducted by The Nether-
lands Interuniversity Cardiology Institute. Am J
Cardiol 1987;59(1):6–13.
[54] Birnbaum Y, Chetrit A, Sclarovsky S, et al. Abnor-
mal Q waves on the admission electrocardiogram
of patients with first acute myocardial infarction:
prognostic implications. Clin Cardiol 1997;20(5):
477–81.
[55] Selwyn AP, Ogunro E, Shillingford JP. Loss of elec-
trically active myocardium during anterior infarc-
tion in man. Br Heart J 1977;39(11):1186–91.
[56] von Essen R, Merx W, Doerr R, et al. QRS mapping
in the evaluation of acute anterior myocardial in-
farction. Circulation 1980;62(2):266–76.
[57] Kalbfleisch JM, Shadaksharappa KS, Conrad LL,
et al. Disappearance of the Q-deflection following
myocardial infarction. Am Heart J 1968;76(2):193–8.
et al
[58] Birnbaum Y, Hale SL, Kloner RA. Changes in
R wave amplitude: ECG differentiation between
episodes of reocclusion and reperfusion associated
with ST-segment elevation. J Electrocardiol 1997;
30(3):211–6.
[59] Birnbaum Y, Maynard C, Wolfe S, et al. Terminal
QRS distortion on admission is better than ST-
segment measurements in predicting final infarct
size and assessing the potential effect of thrombolytic
therapy in anterior wall acute myocardial infarction.
Am J Cardiol 1999;84(5):530–4.
[60] Birnbaum Y, Herz I, Sclarovsky S, et al. Prognostic
significance of the admission electrocardiogram in
acute myocardial infarction. J Am Coll Cardiol
1996;27(5):1128–32.
[61] Birnbaum Y, Kloner RA, Sclarovsky S, et al.
Distortion of the terminal portion of the QRS on
the admission electrocardiogram in acute myocar-
dial infarction and correlation with infarct size and
long-term prognosis (Thrombolysis in Myocardial
Infarction 4 trial). Am J Cardiol 1996;78(4):
396–403.
[62] Cercek B, Lew AS, Laramee P, et al. Time course
and characteristics of ventricular arrhythmias after
reperfusion in acute myocardial infarction. Am J
Cardiol 1987;60(4):214–8.
[63] Gorgels AP, Vos MA, Letsch IS, et al. Usefulness of
the accelerated idioventricular rhythm as a marker
for myocardial necrosis and reperfusion during
thrombolytic therapy in acute myocardial infarc-
tion. Am J Cardiol 1988;61(4):231–5.
[64] Wei JY, Markis JE, Malagold M, et al. Cardiovascu-
lar reflexes stimulated by reperfusion of ischemic
myocardium in acute myocardial infarction. Circu-
lation 1983;67(4):796–801.
[65] Esente P, Giambartolomei A, Gensini GG, et al.
Coronary reperfusion and Bezold-Jarisch reflex
(bradycardia and hypotension). Am J Cardiol 1983;
52(3):221–4.
[66] Vatterott PJ, Hammill SC, Bailey KR, et al. Late po-
tentials on signal-averaged electrocardiograms and
patency of the infarct-related artery in survivors of
acute myocardial infarction. J Am Coll Cardiol
1991;17(2):330–7.
[67] Hong M, Peter T, Peters W, et al. Relation between
acute ventricular arrhythmias, ventricular late po-
tentials and mortality in acute myocardial infarc-
tion. Am J Cardiol 1991;68(15):1403–9.
[68] Tranchesi B Jr, Verstraete M, Van de Werf F, et al.
Usefulness of high-frequency analysis of signal-aver-
aged surface electrocardiograms in acute myocardial
infarction before and after coronary thrombolysis
for assessing coronary reperfusion. Am J Cardiol
1990;66(17):1196–8.
[69] Ophuis AJ, Bar FW, Vermeer F, et al. Angiographic
assessment of prospectively determined non-invasive
reperfusion indices in acute myocardial infarction.
Heart 2000;84(2):164–70.
 Cardiol Clin 24 (2006) 377–385
Electrocardiographic Diagnosis of Myocardial
Infarction during Left Bundle Branch Block
S. Serge Barold, MD*, Bengt Herweg, MD
Division of Cardiology, University of South Florida College of Medicine and Tampa General Hospital,
Tampa, FL 33606, USA
The diagnosis of myocardial infarction (MI) in
the presence of left bundle branch block (LBBB)
has long been considered problematic or even
almost impossible. Many proposed ECG markers
in the old literature have now been discarded.
However, the advent of reperfusion therapy has
generated greater interest in the ECG diagnosis of
acute MI (based on ST-segment abnormalities)
[1–4], although criteria for old MI (based on QRS
changes) have not been reevaluated for almost 20
years [5,6]. Furthermore, analysis of the some of
the published data is compounded by the consid-
erable interobserver variability in the interpreta-
tion of ECGs [6–8].
Acute myocardial infarction
ST-segment deviation is the only useful elec-
trocardiographic sign for the diagnosis of acute
MI in the presence of LBBB. In uncomplicated
LBBB, ECG leads with a predominantly negative
QRS complex show ST-segment elevation with
positive T waves, a pattern similar to the current
of injury observed during acute myocardial ische-
mia or MI. Studies of patients with LBBB during
either acute MI [9–11], or occlusion of a coronary
artery by an angioplasty balloon [12,13] have
shown that further ST-segment elevation occurs
in these leads. Electrocardiographic signs involv-
ing the QRS complex are not diagnostically useful
in the acute setting.
Sgarbossa and colleagues [1] studied 131 pa-
tients (enrolled in the GUSTO-1 trial) with acute
* Corresponding author.
E-mail address: ssbarold@aol.com (S.S. Barold).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.05.002
MI (documented by serum enzyme changes) and
LBBB on their baseline ECG. The following defi-
nition of LBBB was used: a QRS duration of at
least 0.125 seconds in the presence of sinus or sup-
raventricular rhythm, a QS or rS complex in lead
V1, and an R-wave peak time of at least 0.06 sec-
onds in lead I, V5, or V6 associated with the ab-
sence of a Q or q wave in the same leads.
Patients with ECGs showing intermittent LBBB
were excluded from the study. The control group
consisted of 131 patients randomly selected from
the Duke Databank for Cardiovascular Disease,
who had complete LBBB and stable, angiograph-
ically documented coronary artery disease. These
patients did not have acute chest pain at the
time of the recorded ECGs.
The maximal sensitivity with the target speci-
ficity (O90%) was achieved in the following
situations: (1) at least one lead exhibiting ST-
segment elevation R1 mm concordant with (in the
same direction as) a predominantly positive QRS
complex. (2) Discordant ST-segment elevation
5 mm with (in the opposite direction from) a pre-
dominantly negative QRS complex. (3) ST-seg-
ment depression R1 mm in V1, V2, or V3 (Figs. 1
and 2). Electrocardiographic criteria with statisti-
cal significance for the diagnosis of acute MI and
their sensitivities, specificities, and likelihood ra-
tios from the study of Sgarbossa and colleagues
are listed in Table 1. The likelihood ratios indicate
to what extent a particular criterion will increase
or decrease the probability of infarction. The
ECG criterion with the highest likelihood ratio
was ST-segment elevation of at least 1 mm in
leads with a QRS complex in the same direction.
Similarly, the absence of this criterion was associ-
ated with the lowest likelihood ratio.
cardiology.theclinics.com
378 BAROLD & HERWEG

Fig. 1. ECG meeting all three independent criteria of Sgarbossa and colleagues for the diagnosis of acute MI with
LBBB. The ECG shows at least 1-mm concordant ST-elevation in lead II, at least 1-mm ST depression in leads V2
and V3, as well as discordant ST-elevation of at least 5 mm in leads III and aVF (Reproduced from Sgarbossa EB, Pinski
SL, Barbagelata A, et al, for the GUSTO-1 investigators. Electrocardiographic diagnosis of evolving acute myocardial
infarction in the presence of left bundle branch block. N Engl J Med 1996;334:481–7; Ó 1996 Massachusetts Medical
Society. Used with permission.)

With regard to the weakest criterion (ST-
segment elevation R5 mm discordant with the
QRS), Madias [14,15] warned that this sign may
occur in clinically stable patients with LBBB
without an acute MI (6%) in the presence of
unusually large QRS complexes in V1 to V3 in
which leads the ST-segment elevations are also
large. Such patients frequently have severe left
ventricular hypertrophy or markedly dilated
hearts.

Fig. 2. Acute MI. The ECG shows sinus rhythm, and complete LBBB, and an acute anterolateral MI. There is concor-
dant ST-elevation obvious in lead aVL, and less prominent in lead I. The right precordial leads (V1–V4) show marked
discordant ST-elevation.
 ECG DIAGNOSIS OF MI DURING LBBB 379

Table 1
Results of the univariate analysis of the electrocardiographic criteria in the study of Sgarbossa and colleagues [1]
Sensitivity percent Specificity percent Positive likelihood Negative likelihood
Criterion (95% CI) (95% CI) ratio (95% CI) ratio (95% CI)
ST-segment elevation 73 (64–80) 92 (86–96) 9.54 (3.1–17.3) 0.3 (0.22–0.39)
R1 mm and concordant
with the QRS complex
ST-segment depression 25 (18–34) 96 (91–99) 6.58 (2.6–16.1) 0.78 (0.7–0.87)
R1 mm in lead
V1, V2, or V3
ST-segment elevation 31 (23–39) 92 (85–96) 3.63 (2.0–6.8) 0.75 (0.67–0.86)
R5 mm and discordant
with QRS complex
Positive T wave in lead 26 (19–34) 92 (86–96) 3.42 (0.18–6.5) 0.8 (0.72–0.9)
V5 or V6
Left-axis deviation 72 (63–79) 48 (39–57) 1.38 (1.13–9.8) 0.59 (0.25–1.39)
Abbreviation: CI, confidence interval.
Positive likelihood (LR) ratio: the percentage of acute myocardial infarction (MI) patients positive by a stated ECG
sign for diagnosis divided by the percentage of patients without MI but showing a similar positive ECG sign. LRO1
indicates an increased probability that the target disorder is present, and an LR!1 indicates a decreased probability
that the target disorder is present. A likelihood ratio of 9 means that the criterion in question is nine times as likely
to occur in acute MI as it is in a patient without an MI.

probability of an individual with acute MI having a positive sign

LRþ ¼
probability of an individual without acute MI having a positive sign

probability of an individual with acute MI having a negative sign

LR
¼
probability of an individual without acute MI having a negative sign

Reproduced from Sgarbossa EB, Pinski SL, Barbagelata A, et al, for the GUSTO-1 investigators. Electrocardio-
graphic diagnosis of evolving acute myocardial infarction in the presence of left bundle branch block. N Engl J Med
1996;334:481–7. Ó 1996 Massachusetts Medical Society. Used with permission.

Scoring system indicates a moderate-to-high probability of MI,

Sgarbossa and colleagues [1] developed an al-
gorithm where an ECG is considered positive for
MI if its score is at least three points on the basis
of three criteria: ST-segment elevation of at least 1
mm in the lead with concordant QRS complex–
a score of five points; ST-segment depression of
at least 1 mm in leads V1, V2, or V3da score of
three points; and ST-segment elevation of at least
5 mm in the lead with discordant QRS complexd
a score of two points (Table 2). The scoring sys-
tem represents the fact that ST-segment elevation
of at least 1 mm that is concordant with the QRS
complex or ST-segment depression of at least 1
mm in lead V1, V2, or V3 is a specific marker of
infarction, even when no other ECG change is ob-
served. On the other hand, the sole presence of
ST-segment elevation of at least 5 mm that is dis-
cordant with the QRS complex (with a score of 2)
and further procedures should be undertaken to
confirm the diagnosis. Sgarbossa and colleagues
[1] indicated that their algorithm based on ST-seg-
ment changes (index score of at least 3) had a sen-
sitivity of 78% and a specificity of 90% for the
diagnosis of MI in patients with LBBB.
Hirulog and Early Reperfusion or Occlusion
Trial (HERO-2)
The recently reported Hirulog and Early Re-
perfusion or Occlusion Trial (HERO-2) study [4]
involved 300 patients presenting with O30 min
of ischemic chest discomfort and presumed new-
onset LBBB according to the criteria of Sgarbossa
and colleagues [1]. Enzymatically confirmed acute
MI occurred in 80.7% of the LBBB patients.
Ninety-two patients exhibited positive ST-segment
380 BAROLD & HERWEG

Table 2
Odds ratios and scores for independent electrocardiographic criteria from Sgarbossa and colleagues [1]
Criterion Odds ratio (95% CI) Score
ST-segment elevation R1 mm and 25.2 (11.6–54.7) 5
concordant with QRS complex
ST-segment depression R1 mm in 6.0 (1.9–19.3) 3
lead V1, V2, or V3
ST-segment elevation R5 mm and 4.3 (1.8–10.6) 2
discordant with QRS complex
The odds ratio is a way of comparing whether the probability of a certain event is the same for two groups. An odds
ratio of 1 implies that the event is equally likely in both groups. An odds ratio O1 implies that the event is more likely
in the first group. An odds ratio !1 implies that the event is less likely in the first group. The table shows the ratio
of the odds of having the ECG sign in the acute myocardial infarction group relative to the odds of having the sign
in the control group. (Reproduced from Sgarbossa EB, Pinski SL, Barbagelata A, et al, for the GUSTO-1 investigators.
Electroþcardiographic diagnosis of evolving acute myocardial infarction in the presence of left bundle branch block.
N Engl J med 1996;334:481–7.)

abnormalities for the diagnosis of acute MI accord- Clinical implications of the Sgarbossa criteria
ing to the criteria of Sgarbossa and colleagues [1].
The clinical utility of the criteria and scoring
The study confirmed the findings of Sgarbossa
system of Sgarbossa and colleagues [1] have been
and colleagues [1] in terms of the following results
validated by other studies, all of which have also
(Table 3): (1) concordant ST-segment elevation
demonstrated a high specificity, but some have
R1 mm: high specificity (98.3%) but low sensitivity
shown an even lower sensitivity than the original
(33.5%). (2) ST-segment depression measuring R1
data of Sgarbossa and colleagues [1] in terms of
mm in any of the V1 to V3 leads had similarly high
the three individual ST-segment criteria and the
specificity, but only 14.1% sensitivity. Lowering
scoring algorithm [8,16–21]. As such, although
the cutoff for ST-segment changes to R0.5 mm
the criteria and the algorithm cannot be used to
for each of the criteria in 1 and 2 did not improve
rule out MI, it can help to rule it in. Patients
sensitivity. When both criteria were combined (ie,
with an acute MI and LBBB have a high mortality
concordant ST-segment elevation or lead V1 to
rate, but this is significantly related to age and co-
V3 ST-segment depression), the specificity for de-
morbidities [22–24]. Thus, these markers should
tection of enzymatically confirmed acute MI was
be used together with the clinical findings because
96.6%, and the sensitivity was 37.2%. (3) Discor-
the ECG markers alone miss acute MI in many
dant ST-segment elevation measuring R5 mm
patients who would benefit from aggressive
was neither sensitive (29.3%) nor specific (58.6%).

Table 3
Application of ST-segment criteria for the diagnosis of AMI in the 300 patients with LBBB at randomization from
Wong and colleagues [4]
n Sensitivity Specificity Positive predictive Negative predictive
(%) (%) value (%) value (%)
Concordant ST-segment 82 33.5 (27.6–39.8) 98.3 (89.5–99.9) 98.8 (92.5–99.9) 26.1 (20.6–32.6)
elevation R1 mm
Lead V1 to V3 ST-segment 35 14.1 (10.1–19.2) 98.4 (89.5–99.9) 97.1 (83.4–99.9) 21.5 (16.8–27.0)
depression R1 mm
Concordant ST-segment 92 37.2 (31.1–43.6) 96.6 (87.0–99.4) 97.8 (91.6–99.6) 26.9 (21.1–33.6)
elevation R1 mm or lead
V1 to V3 ST-segment
depression R1 mm
Abbreviations: AMI, acute myocardial infarction; LBBB, left bundle branch block. (Reproduced from Wong CK,
French JK, Aylward PE, et al, and the HERO-2 Trial Investigators. Patients with prolonged ischemic chest pain and
presumed-new left bundle branch block have heterogeneous outcomes depending on the presence of ST-segment
changes. J Am Coll cardiol 2005;46:29–38; with permission from American College of Cardiology Foundation.)
 ECG DIAGNOSIS OF MI DURING LBBB 381

Fig. 3. Anterior MI of undetermined age with double Cabrera’s sign. (A) The ECG shows sinus rhythm, complete
LBBB, and qR complexes in leads I, aVL, and V4. Note the double Cabrera’s sign in lead V4. The presence of sinus
rhythm with a normal PR interval rules out a retrograde P wave as the cause of one of the notches on the ascending
limb of the S wave. (B) Magnified ECG of leads V4 and V5.

Fig. 4. Anterior MI of undetermined age. The ECG shows sinus rhythm with first-degree block and complete LBBB.
Note the rather tall first deflection in lead V1, which is an R wave. This finding in complete LBBB is very typical of ante-
roseptal MI of undetermined age. Poor R-wave progression V1 toV6 is also consistent with anterior MI.
382 BAROLD & HERWEG

Fig. 5. ECG pattern of MI after development of complete LBBB. (A) ECG showing sinus rhythm and an extensive acute
anterior MI. (B) ECG 1 month later showing sinus rhythm, new complete LBBB, and many of the signs of anterior MI:
tall initial positive deflection (R wave) in lead V1, Cabrera’s sign in leads V2– V4, q wave in leads 1 and aVL, and poor
r-wave progression in leads V4–V6.

treatment. The published studies showing poor Old (remote) myocardial infarction
sensitivity of the ECG markers support the rec-
In uncomplicated LBBB, septal activation
ommendations of the American College of Cardi-
occurs from right to left because the left septal
ology and the American Heart Association that
mass cannot be activated via the left bundle.
all patients with LBBB irrespective of ECG fea-
Consequently, LBBB does not generate a q wave
tures and symptoms of acute MI should receive
in the lateral leads (I and V6). Lead V1 may
reperfusion therapy (angioplasty may be prefera-
show an initial r wave because of the anterior
ble to fibrinolytic therapy if there are no contrain-
component of right-to-left septal activation but
dications) [25,26].
 ECG DIAGNOSIS OF MI DURING LBBB 383

Fig. 6. Possible anterior MI unmasked by ventricular extrasystole during complete LBBB. Leads V2–V5 show ventric-
ular extrasystoles with a qR or Qr comlexes consistent with anterior MI.

leads V1 to V3 may also show QS complexes. After
crossing the ventricular septum, the activation
reaches the left ventricle, which is depolarized
via ordinary myocardium, QS complexes may be
seen in leads III and aVF. Secondary ST segment,
and T-wave abnormalities are oriented in the op-
posite direction compared with the QRS complex.
The ECG manifestations of the old MI may re-
main concealed, probably more commonly than
those of acute MI with LBBB [2].
During LBBB, an extensive anteroseptal MI
will alter the initial QRS vector, with forces point-
ing to the right because of unopposed activation
of the right ventricle. This causes (initial) q waves
in leads I, aVL, V5, and V6, producing an Qr or
qR pattern.
A number of old studies reported that the
presence of a Q wave in lead 1 was a highly specific
and relatively sensitive sign for the diagnosis of
anterior infarction in the presence of complete
LBBB [27,28]. Cabrera and Friedland [29] de-
scribed the diagnostic value of late notching of
the S wave in leads V3 to V5 (Fig. 3) in anterior in-
farction in terms of very high sensitivity and
specificity.
With regard to the QRS complex in the di-
agnosis of MI in LBBB, Wackers [6] also found
that an abnormal Q wave in leads I, aVL, or V6
(duration not stated) may be of diagnostic value
in anteroseptal MI with a sensitivity of 53% and
specificity of 91% (Figs. 4 and 5). A highly specific
criterion (100%) was the combination of an ab-
normal Q wave in V6 and an increased sharp R
wave in V1. This combination occurred only in pa-
tients with an anteroseptal MI, but the sensitivity
was low (20%). Cabrera’s sign (defined as notch-
ing of 0.05 sec in duration in the ascending limb
of the S wave in leads V3 or V4) was also useful
with a specificity of 47% for anteroseptal MI
and a specificity of 87% (see Fig. 3). Wackers [6]
also found that a number of other previously pro-
posed QRS signs were disappointing for the diag-
nosis of MI. Wackers [6] also found that so-called
primary T-wave changes (T wave in the same di-
rection as the QRS complex) carry no important
diagnostic value.
The most recent study involving the QRS
complex was published in 1989 by Hands and
colleagues [5]. They confirmed that Q waves (R30
msec) in two or more lateral leads (I, aVL, V5, and
384 BAROLD & HERWEG
V6) and R-wave regression from V1 to V4 each
had a poor sensitivity (21%) but high specificity
(100%) for the diagnosis of anterior infarction
(see Figs. 3–5). A Q wave of any size in the lateral
leads yielded a sensitivity of 29% and specificity of
91%. Pathologic studies have confirmed the pres-
ence of septal infarction in patients with LBBB
and Q waves in the lateral leads (I, aVL, V5,
and V6). Late notching of the upstroke of the S
wave (Cabrera’s sign) in at least two leads V3 to
V5 provided a sensitivity of 29% and specificity
of 91% (see Figs. 3 and 5). Other previously pro-
posed signs of MI involving the QRS complex in
LBBB were found to have poor sensitivity, speci-
ficity, and predictive value. The significance of
a tall R wave in lead V1 during LBBB as a sign
of anterior MI was not studied in the report of
Hands and colleagues (see Figs. 4 and 5) [5]. This
may be a rare but very specific sign of MI corre-
sponding to a q wave possibly in leads V7 or V8.
Ventricular extrasystoles
Ventricular extrasystoles may unmask the
pattern of an underlying MI in patients with
LBBB, but this sign is not absolutely specific
[30]. Such ventricular extrasystoles must satisfy
two conditions. (1) The configuration must be ei-
ther qR or qRs but not QS, because a QS complex
can be generated by an extrasystole originating in
an area underlying the recording electrode. (2)
The qR or qRs complex must be registered in
a lead that would ordinarily be expected to reflect
left ventricular epicardial potentials in the precor-
dial leads (Fig. 6). According to Coumel [31], who
analyzed the significance of QR complexes during
ventricular tachycardia in patients with coronary
artery disease, the QR, qR, or qRs patterns reflect
an MI, although its exact site cannot be deter-
mined. Josephson and Miller [32] disagree with
Coumel [31] because they observed qR patterns
in ventricular tachycardia with a LBBB pattern
in patients with cardiomyopathy. They empha-
sized that a QR complex could originate from
a fixed scar (infarct) or a conduction disturbance
secondary to fibrosis regardless of etiology.
References
[1] Sgarbossa EB, Pinski SL, Barbagelata A, et al, for
the GUSTO-1 investigators. Electrocardiographic
diagnosis of evolving acute myocardial infarction
in the presence of left bundle branch block. New
Engl J Med 1996;334:481–7.
[2] Sgarbossa EB. Recent advances in the electrocar-
diographic diagnosis of myocardial infarction: left
bundle branch block and pacing. Pacing Clin Elec-
trophysiol 1996;19:1370–9.
[3] Sgarbossa EB. Value of the ECG in suspected acute
myocardial infarction with left bundle branch block.
J Electrocardiol 2000;33(Suppl):87–92.
[4] Wong CK, French JK, Aylward PE, et al, and
HERO-2 Trial Investigators. Patients with pro-
longed ischemic chest pain and presumed-new left
bundle branch block have heterogeneous outcomes
depending on the presence of ST-segmentchanges.
J Am Coll Cardiol 2005;46:29–38.
[5] Hands ME, Cook EF, Stone PH, et al, and the
MILIS Study Group. Electrocardiographic diagno-
sis of myocardial infarction in the presence of com-
plete left bundle branch block. Am Heart J 1988;
116:23–32.
[6] Wackers FJ. The diagnosis of myocardial infarction
in the presence of left bundle branch block. Cardiol
Clin 1987;5:393–401.
[7] Wellens HJ. Acute myocardial infarction and left
bundle-branch blockdcan we lift the veil? N Engl
J Med 1996;334:528–9.
[8] Gula LJ, Dick A, Massel D. Diagnosing acute myo-
cardial infarction in the setting of left bundle branch
block: prevalence and observer variability from
a large community setting. Coronary Artery Dis
2003;14:387–93.
[9] Kennamer R, Prinzmetal M. Myocardial infarction
complicated by left bundle branch block. Am Heart
J 1956;51:78–90.
[10] Wackers FJ. Complete left bundle branch block: is
the diagnosis of myocardial infarction possible? Int
J Cardiol 1983;2:521–9.
[11] Sclarovsky S, Sagie A, Strasberg B, et al. Ischemic
blocks during early phase of anterior myocardial in-
farction: correlation with ST-segment shift. Clin
Cardiol 1988;11:757–62.
[12] Cannon A, Freedman SB, Bailey BP, et al. ST-seg-
ment changes during transmural myocardial ische-
mia in chronic left bundle branch block. Am J
Cardiol 1989;64:1216–7.
[13] Stark KS, Krucoff MW, Schryver B, et al. Quantifi-
cation of ST-segment changes during coronary an-
gioplasty in patients with left bundle branch block.
Am J Cardiol 1991;67:1219–22.
[14] Madias JE, Sinha A, Ashtiani R, et al. A critique
of the new ST-segment criteria for the diagnosis
of acute myocardial infarction in patients with
left bundle-branch block. Clin Cardiol 2001;24:
652–5.
[15] Madias JE, Sinha A, Agarwal H, et al. ST-segment
elevation in leads V1–V3 in patients with LBBB.
J Electrocardiol 2001;34:87–8.
[16] Li SF, Walden PL, Macrcilla O, et al. Electrocardio-
graphic diagnosis of myocardial infarction in pa-
tients with left bundle branch block. Ann Emerg
Med 2000;36:561–5.
 ECG DIAGNOSIS OF MI DURING LBBB
[17] Sokolove PE, Sgarbossa EB, Amsterdam EA, et al.
Interobserver variability in the electrocardiographic
diagnosis of acute myocardial infarction in patients
with left bundle branch block. Ann Emerg Med
2000;36:566–71.
[18] Gunnarsson G, Eriksson P, Dellborg M. ECG crite-
ria in diagnosis of acute myocardial infarction in the
presence of left bundle branch block. Int J Cardiol
2001;78:167–74.
[19] Kontos MC, McQueen RH, Jesse RL, et al. Can
myocardial infarction be rapidly identified in Emer-
gency Department patients who have left bundle
branch block? Ann Emerg Med 2001;37:431–8.
[20] Shlipak MG, Lyons WL, Go AS, et al. Should the
electrocardiogram be used to guide therapy for pa-
tients with left bundle-branch block and suspected
myocardial infarction? JAMA 1999;281:714–9.
[21] Edhouse JA, Sakr M, Angus J, et al. Suspected myo-
cardial infarction and left bundle branch block: elec-
trocardiographic indicators of acute ischaemia.
J Accid Emerg Med 1999;16:331–5.
[22] Stenestrand U, Tabrizi F, Lindback J, et al.
Comorbidity and myocardial dysfunction are the
main explanations for the higher 1-year mortality
in acute myocardial infarction with left bundle-
branch block. Circulation 2004;110(14):1896–902.
[23] Haywood LJ. Left bundle branch block in acute
myocardial infarction: benign or malignant? J Am
Coll Cardiol 2005;46:39–41.
[24] Moreno R, Garcia E, Lopez de Sa E, et al. Implica-
tions of left bundle branch block in acute myocardial
infarction treated with primary angioplasty. Am J
Cardiol 2002;90:401–3.
[25] Fibrinolytic Therapy Trialists’ (FTT) Collaborative
Group. Indications for fibrinolytic therapy in
385
suspected myocardial infarction: collaborative over-
view of early mortality and major morbidity results
from all randomised trials of more than 1000
patients. Lancet 1994;343:311–22.
[26] Antman EM, Anbe DT, Armstrong PW, et al. ACC/
AHA guidelines for the management of patients
with ST-elevation myocardial infarction; a report
of the American College of Cardiology/American
Heart Association Task Force on Practice Guide-
lines (Committee to Revise the 1999 Guidelines for
the Management of patients with acute myocardial
infarction). J Am Coll Cardiol 2004;44(3):E1–211.
[27] Besoaı́n-Santander M, Gómez-Ebensperguer G.
Electrocardiographic diagnosis of myocardial in-
farction in cases of complete left bundle branch
block. Am Heart J 1960;60:886–97.
[28] Doucet P, Walsh TJ, Massie E. A vectorcardio-
graphic and electrocardiographic study of left bun-
dle branch block with myocardial infarction. Am J
Cardiol 1966;17:171–9.
[29] Cabrera E, Friedland C. La onda de activación ven-
tricular en el bloqueo de rama izquierda con infarto:
un nuevo signo electrocardiográfico. Arch Inst Car-
diol Mex 1953;23:441–60.
[30] Dressler W. A case of myocardial infarction masked
by bundle branch block but revealed by occasional
premature ventricular beats. Am J Med Sci 1943;
206:361.
[31] Coumel P. Diagnostic significance of the QRS wave
form in patients with ventricular tachycardia. Cardi-
ol Clin 1987;5:527–40.
[32] Josephson ME, Miller JM. Endocardial and epicar-
dial recordings. Correlation of twelve-lead electro-
cardiograms at the site of origin of ventricular
tachycardia. Ann N Y Acad Sci 1990;601:128–47.
 Cardiol Clin 24 (2006) 387–399

Electrocardiographic Diagnosis of Myocardial

Infarction and Ischemia during Cardiac Pacing
S. Serge Barold, MD*, Bengt Herweg, MD, Anne B. Curtis, MD
Division of Cardiology, University of South Florida College of Medicine
and Tampa General Hospital, Tampa, FL, USA

The ECG diagnosis of myocardial infarction
(MI) and ischemia in pacemaker patients can be
challenging. Many of the criteria are insensitive,
but the diagnosis can be made in a limited number
of cases because of the high specificity of some of
the criteria.
Old myocardial infarction
Box 1 outlines the difficulties in the diagnosis
of MI, and Box 2 lists a number of signs of no value
in the diagnosis of MI. Generally, when using the
QRS complex, the sensitivity is low (25%) and the
specificity is close to 100%. One cannot determine
the age of the MI from the QRS complex.
Anterior myocardial infarction
St-qR pattern
Because the QRS complex during right ven-
tricular (RV) pacing resembles (except for the
initial forces) that of spontaneous left bundle
branch block (LBBB), many of the criteria for
the diagnosis of MI in LBBB also apply to MI
during RV pacing [1–4]. RV pacing almost invari-
ably masks a relatively small anteroseptal MI.
During RV pacing, as in LBBB, an extensive
anteroseptal MI close to the stimulating electrode
will alter the initial QRS vector, with forces
pointing to the right because of unopposed activa-
tion of the RV. This causes (initial) q waves in leads
I, aVL, V5, and V6, producing an St-qR pattern
(Fig. 1). The abnormal q wave is usually 0.03 sec-
onds or more, but a narrower one is also diagnostic.
Occasionally the St-qR complex is best seen in leads
V2 to V4, and it may even be restricted to these
leads. Finding the (initial) q wave may sometimes
require placing the leads one intercostal space
higher or perhaps lower. Ventricular fusion may
cause pseudoinfarction patterns (Fig. 2).
The sensitivity of the St-qR pattern varies from
10% to 50% according to the way data are
analyzed [5,6]. Patients who require temporary
pacing in acute MI represent a preselected group
with a large MI, so that the overall sensitivity is
substantially lower than 50% in the patient popu-
lation with implanted pacemakers. The specificity
is virtually 100%.
Late notching of the ascending S wave
(Cabrera’s sign)
As in LBBB, during RV pacing an extensive
anterior MI may produce notching of the ascending
limb of the S wave in the precordial leads usually V3
and V4dCabrera’s sign R0.03 seconds and present
in two leads (Fig. 3) [1]. The sign may occur to-
gether with the St-qR pattern in anterior MI
(see Fig. 1). The sensitivity varies from 25% to
50% according to the size of the MI, but the speci-
ficity is close to 100% if notching is properly defined
[1,5]. Interestingly, workers [7] that placed little di-
agnostic value on q waves, found a 57% sensitivity
for Cabrera’s sign (0.04-sececond notching) in the
diagnosis of extensive anterior MI. Box 3 outlines
the causes of ‘‘false’’ Cabrera’s signs and the highly
specific variants of Cabrera’s sign (Fig. 4).
Inferior myocardial infarction

* Corresponding author. The paced QRS complex is often unrevealing.

E-mail address: ssbarold@aol.com (S.S. Barold). During RV pacing in inferior MI diagnostic Qr,
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.05.003 cardiology.theclinics.com
388
Box 1. Difficulties in the diagnosis
of MI during ventricular pacing
1. Large unipolar stimuli may obscure
initial forces, cause a pseudo Q
wave and false ST segment current
of injury.
2. QS complexes are of no diagnostic
value. Only qR or Qr complexes may
be diagnostically valuable.
3. Fusion beats may cause
a pseudoinfarction pattern (qR/Qr
complex or notching of the upstroke
of the S wave).
4. Cabrera’s sign can be easily
overdiagnosed.
5. Retrograde P waves in the terminal
part of the QRS complex may mimic
Cabrera’s sign.
6. Acute MI and ischemia may be
difficult to differentiate.
7. Differentiation of acute MI
and old or indeterminate age MI
may not be possible on the basis
of abnormalities of the
ST segment.
8. Signs in the QRS complex are not
useful for the diagnosis of acute MI.
9. ST segment changes usually but not
always indicate an acute process.
10. Recording QRS signs of MI may
require different sites of the left V
leads such as a different intercostals
space.
11. Biventricular pacing can mask an
MI pattern in the QRS complex
evident during RV pacing.
12. qR or Qr complexes are common
during biventricular pacing and do
not represent an MI.
13. Cardiac memory. Repolarization
ST-T wave abnormalities (mostly T
wave inversion) in the spontaneous
rhythm may be secondary to RV
pacing per se and not related to
ischemia or non–Q wave MI.
14. QRS abnormalities have low
sensitivity (but high specificity).
15. Beware that not all the diagnostic
criteria of MI in left bundle branch
block are applicable during RV
pacing.
BAROLD et al
Box 2. QRS criteria of no value
in diagnosis of MI

QS complexes V1 to V6

RS or terminal S wave in V5 and V6

QS complexes in the inferior leads

Slight notching of R waves

Slight upward slurring of the
ascending limb of the S wave
QR, or qR complexes provide a sensitivity of 15%
and specificity of 100% (Fig. 5) [1,5]. The St-qR
pattern must not be confused with an overshoot
of the QRS complex due to overshoot of massive
ST elevation creating a diminutive terminal r wave
or ventricular fusion (see Fig. 5). Cabrera’s sign in
both leads III and aVF is very specific, but even
less sensitive than its counterpart in anterior MI
(S.S. Barold, unpublished observations).
Myocardial infarction at other sites
A posterior MI should shift the QRS forces
anteriorly and produce a dominant R wave in the
right V leads, but the diagnosis cannot be made
during RV pacing because of the many causes of
a dominant R wave in V1. An RV MI could con-
ceivably be reflected in V3R with prominent ST el-
evation. Klein and colleagues [8] suggested that
the diagnosis of RV infarction could be made
when there is prominent ST elevation in lead
V4R in the first 24 hours, but such a change
should be interpreted cautiously unless it is associ-
ated with obvious abnormalities suggestive of an
acute inferior MI.
Conflicting views on the diagnosis of myocardial
infarction of uncertain age
Kochiadakis and colleagues [9] studied ECG
patterns of ventricular pacing in 45 patients with
old MI and 26 controls (without angiographic ev-
idence of coronary artery disease) during tempo-
rary RV apical at the time of routine cardiac
catheterization (Fig. 6). In 15 of the 26 controls,
a Q wave was observed in leads I, aVL, or V6.
However, it was not specified whether the Q waves
were part of a qR (Qr) or a QS complex (their
Fig. 1E shows a QS complex). This differentiation
is important because a QS complex carries no di-
agnostic value during RV pacing in any of the
 ECG DIAGNOSIS OF MI AND ISCHEMIA DURING CARDIAC PACING 389

Fig. 1. Twelve-lead ECG showing old anteroseptal myocardial infarction during unipolar DDD pacing in a patient with
complete AV block. The ventricular stimulus does not obscure or contribute to the qR pattern in leads I, aVL, and V6.
Leads V2 to V4 show Cabrera’s sign and a variant in lead V5. The lack of an underlying rhythm because of complete AV
block excluded the presence of ventricular fusion.

standard 12 leads (QS complexes can be normal in never produces a qR complex in V5 and V6 in
leads I, II, III, aVF, V5, and V6). A well-posi- the absence of an MI. It is also possible that in
tioned lead at the RV apex rarely generates the study of Kochiadakis and colleagues [9], the
a qR complex in lead I, and in our experience pacing catheter in some of the controls might

Fig. 2. Twelve-lead ECG showing ventricular fusion related to spontaneous atrioventricular conduction. The pattern
simulates myocardial infarction during DDD pacing (atrial sensing-ventricular pacing) in a patient with sick sinus syn-
drome, relatively normal AV conduction, and no evidence of coronary artery disease. The spontaneous ECG showed
a normal QRS pattern. Note the QR complexes in leads II, III, aVF, V5, and V6.
390
Fig. 3. Twelve-lead ECG showing Cabrera’s sign during VVI pacing in a patient with an old extensive anterior myocar-
dial infarction. Note the typical notching of the S wave in leads V4 to V6. There is no qR pattern.
Box 3. Cabrera’s sign
Specific Cabrera variants

Small, narrow r wave deforming the
terminal QRS.

Series of tiny notches giving
a serrated appearance along the
ascending S wave.

Similar series of late notches on QRS
during epicardial pacing.
Notches are probably due to a gross
derangement of intraventricular
conduction.
False Cabrera’s signs

Slight notching of the ascending S
wave in V leads is normal during RV
apical pacing. It is usually confined to 1
lead, shows a sharp upward direction
on the S wave and usually <0.03
seconds; no shelflike or downward
notch typical of true Cabrera’s sign.

Ventricular fusion beats.

Early retrograde P waves deforming
the late part of the QRS complex.
BAROLD et al
have been slightly displaced away from RV apex
and produced qR ventricular complexes in leads
I and aVL (but not V6) with preservation of supe-
rior axis deviation in the frontal plane. On this ba-
sis, we cannot accept the authors’ claim of the
poor diagnostic accuracy and specificity of
Q waves in the diagnosis of MI.
Furthermore, Kochiadakis and colleagues [9]
published an ECG example of Cabrera’s sign (their
Fig. 1A), but the tracing showed unimpressive
slight slurring (with a rapid upward deflectiond
dv/dt or slope) of the ascending limb of the
S wave (see Fig. 6). In our experience, this pattern
is commonly seen during uncomplicated RV apical
pacing. A true Cabrera’s sign is more prominent,
with a markedly different dv/dt beyond the notch,
making the sign unmistakable as seen in Figs. 1
and 3. We believe that the ECG in their Fig. 1B [4]
showing Chapman’s sign (notching with minimal
slurring of the upstroke of the R wave) is also
consistent with uncomplicated RV apical pacing
(see Fig. 6).
Another group [7] has claimed that Q waves
(qR or Qr complexes were not specified) in leads
I, aVL, or V6 are not diagnostically useful, but
their conclusions are also questionable because
of problematic methodology: (1) the number
 ECG DIAGNOSIS OF MI AND ISCHEMIA DURING CARDIAC PACING 391

Fig. 4. Cabrera Variants. (A, B) There are small and narrow terminal R waves in leads V2 and V3, respectively, during
ventricular pacing. (C) Series of tiny notches representing gross derangement of intraventricular conduction during ven-
tricular pacing in a patient with an extensive anterior myocardial infarction. (From Barold SS, Falkoff MD, Ong LS,
et al. Normal and abnormal patterns of ventricular depolarization during cardiac pacing. In: Barold SS, editor. Modern
cardiac pacing. Mt Kisco [NY]: Futura; 1985; with permission.)

Fig. 5. Ventricular pacing during acute inferior wall myocardial infarction showing a qR pattern in leads II, III, and aVF
associated with ST segment elevation. The R wave in the inferior leads is substantial and, therefore, not due to an overshoot
of the QRS complex by marked ST-segment elevation. (Reproduced from Barold SS, Ong LS, Banner RL. Diagnosis of
inferior wall myocardial infarction during right ventricular right apical pacing. Chest 1976;69:232–5; with permission.)
392 BAROLD et al

Fig. 6. Criteria of Kochiadakis and colleagues for the evaluation of old myocardial infarction during ventricular pacing
[9]. (A) Notching 0.04 seconds in duration on the ascending limb of the S wave of leads V3, V4, or V5 (Cabrera’s sign).
(A is shown at the bottom in a magnified form). (B) Notching of the upstroke of the R wave in leads I, aVL, or V6 (Chap-
man’s sign). (C) Q waves O0.03 seconds in duration in leads I, aVL, or V6. (D) Notching of the first 0.04 seconds of the
QRS complex in leads II, III, and aVF. (E) Q wave O0.03 seconds in duration in leads II, III, and aVF. (From Kochia-
dakis GE, Kaleboubas MD, Igoumenidis NE, et al. Electrocardiographic diagnosis of acute myocardial infarction in the
presence of ventricular paced rhythm. PACE 2001;24:1289–90; with permission.)

of ‘‘abnormal’’ patients with Q waves only in the
two frontal plane leads and not in V6 was
not specified. (2) The protocol called for a LBBB
pattern with left axis deviation (more negative
than 30 degrees). Normal subjects might have
been included in the ‘‘abnormal’’ group because
a pacing lead somewhat away from the RV apex
can cause left-axis deviation with q waves in I
and aVL in the absence of MI.
Based on the above arguments, we believe that
the findings of Kachiadakis and colleagues [4] and
Kindwall and colleagues [5] are questionable and
probably not valid.
Acute myocardial infarction
Leads V1 to V3 sometimes show marked ST el-
evation during ventricular pacing in the absence
of myocardial ischemia or infarction [10]. The di-
agnosis of myocardial ischemia or infarction
should therefore be based on the new develop-
ment of ST elevation. Sgarbossa and colleagues
[11,12] recently reported the value of ST segment
abnormalities in the diagnosis of acute MI during
ventricular pacing and their high specificity. ST el-
evation R5 mm in predominantly negative QRS
complexes is the best marker, with a sensitivity
of 53% and specificity of 88%, and was the only
criterion of statistical significance in their study
(Figs. 7 and 8). Other less important ST changes
with high specificity include ST depression = or O
1 mm in V1, V2, and V3 (sensitivity 29%, specificity
82%), and ST elevation R1 mm in leads with a con-
cordant QRS polarity. ST depression concordant
with the QRS complex may occur in leads V3
to V6 during uncomplicated RV pacing [11,12].
Patients who present with discordant ST elevation
R5 mm have more severe coronary artery disease
than other MI patients without such ST elevation
[13,14]. Patients with an acute MI, the primary
ST changes may persist as the MI becomes old.
So-called primary T-wave abnormalities (concor-
dant) are not diagnostically useful during RV pac-
ing if they are not accompanied by primary ST
abnormalities (Fig. 9) [11].
 ECG DIAGNOSIS OF MI AND ISCHEMIA DURING CARDIAC PACING 393

Fig. 7. Twelve-lead ECG showing acute inferolateral myocardial infarction during VVI pacing. There is obvious discor-
dant ST-elevation in leads II, III, aVF, and V6 that meets the criterion of Sgarbossa and colleagues [11] for the diagnosis of
acute infarction. (From Barold SS, Falkoff MD, Ong LS, et al. Normal and abnormal patterns of ventricular depolarization
during cardiac pacing. In: Barold SS, editor. Modern cardiac pacing. Mt Kisco [NY]: Futura; 1985; with permission.)

Fig. 8. Twelve-lead ECG showing an acute anterior myocardial infarction during VVI pacing. There is marked ST-
elevation in leads V1 to V5 that meets the criterion of Sgarbossa and colleagues [11] for the diagnosis of acute infarction.
The ST-elevation drags the QRS complex upwards. Note the right superior frontal plane axis occasionally seen with
right ventricular apical pacing.
394
Fig. 9. Twelve-lead ECG during uncomplicated right ventricular apical pacing showing concordant T-wave inversion in
leads V4 to V6. So-called primary T-wave abnormalities are of no diagnostic value without accompanying ST changes.
Cardiac ischemia
Discordant ST elevation
Marked discordant ST elevation (O5 mm)
during ventricular pacing, a recently described
sign (with good specificity and moderate sensi-
tivity) for the diagnosis of myocardial infarction
[9], could also be used for the diagnosis of severe
reversible transmural myocardial ischemia as re-
cently reported in a case of anterior ischemia
(Fig. 10) [15]. Two similar cases of ischemia
with discordant ST elevation during ventricular
pacing have been published [16,17]. Both affected
the inferior wall. A report in the French litera-
ture [17] involved a temporary pacing lead in
the RV in a patient who demonstrated transient
but massive ST elevation of unspecified duration
in the inferior leads during Prinzmetal’s angina,
possibly superimposed on an inferior infarction
of undetermined age. During these ischemic epi-
sodes, the ECG documented reversible second-
degree type I (Wenckebach) atrioventricular
block and reversible type I second-degree exit
block from the pacemaker stimulus to the myo-
cardium. The latter probably occurred because
the tip of the lead was in direct contact with
BAROLD et al
the area of severe transmural ischemia. The other
case is less impressive because the patient had
a unipolar VVI system (unclear degree of over-
shoot into the ST segment) and exhibited during
chest pain of uncertain duration only about 5 mm
of additional discordant ST elevation in a Holter re-
cording with an unspecified lead [16]. Transient
massive ST elevation (O10 mm) in paced beats
and spontaneous beats in lead III was precipitated
during an ergonovine-induced spasm of a domi-
nant right coronary artery in the presence of
otherwise normal coronary arteries angiographi-
cally [16]. In this patient, the associated ST ele-
vation in spontaneously conducted beats
diminished the diagnostic value of the changes
during pacing.
Discordant ST abnormalities
ST depression in leads V1 and V2 is rarely nor-
mal, and should be considered abnormal and in-
dicative of anterior or inferior MI or ischemia.
Exercise-induced ST changes
Exercise-induced ST abnormalities are in all
likelihood nondiagnostic, as in complete LBBB.
 ECG DIAGNOSIS OF MI AND ISCHEMIA DURING CARDIAC PACING 395

Fig. 10. Diagnosis of myocardial ischemia during ventricular pacing. Three representative panels of three-channel
Holter recordings of lead V 1 on top and V5 at the bottom, together with a special pacemaker channel in the middle
displaying the pacemaker stimuli.The top control panel was recorded before chest pain. The second panel shows marked
ST-elevation (O5 mm) in V1 and to a lesser degree in V5. The bottom panel was recorded about 3.5 minutes after the
middle panel.The ST-elevation has partially resolved. (From Barold SS. Diagnosis of myocardial ischemia during ventri-
cular pacing. Pacing Clin Electrophysiol 2000;23:1060–1; with permission.)

The two cases reported by Diaz and colleagues
[18] are questionable on the basis of the criteria of
Sgarbossa and colleagues [11,12].
Cardiac memory
Abnormal depolarization causes altered re-
polarization. Cardiac memory refers to T-wave
abnormalities that manifest on resumption of
a normal ventricular activation pattern after
a period of abnormal ventricular activation,
such as ventricular pacing, transient LBBB,
ventricular arrhythmias, or Wolf-Parkinson-White
syndrome [19–22]. Pacing-induced T-wave inver-
sion is usually localized to precordial and inferior
leads. The direction of the T wave of the memory
effect in sinus rhythm is typically in the same direc-
tion as the QRS complex. In other words, the
T wave tracks the QRS vector of the abnormal im-
pulse. Thus, inhibition of a pacemaker may
396 BAROLD et al

Fig. 11. Cardiac memory effect secondary to ventricular pacing recorded in the ECG of a patient with complete heart
block from a lesion in the His bundle (confirmed by His bundle recordings). (Top) The tracing is normal except for the
rhythm. (Bottom) Chest wall stimulation was performed to inhibit a VVI pacemaker implanted several months previ-
ously. There was no clinical evidence of heart disease apart from AV block. Note the striking T-wave inversions in leads
II, III, aVF, and V3 to V6. (From Barold SS, Falkoff MD, Ong LS, et al. Electrocardiographic diagnosis of myocardial
infarction during ventricular pacing. Cardiol Clin 1987;5:403–17; with permission.)
 ECG DIAGNOSIS OF MI AND ISCHEMIA DURING CARDIAC PACING 397

allow the emergence of the spontaneous rhythm
with a diagnostic Q wave, but pacing per se may
produce prominent repolarization abnormalities
that do not represent ischemia, a non-ST eleva-
tion, or non-Q wave MI (Fig. 11) [19–22]. It
may occur even after 1 minute of RV pacing
in humans, with T-wave abnormalities visible af-
ter 20 minutes [23]. The marked repolarization
abnormalities reach a steady state in a week
with RV endocardial pacing at physiologic rates.
The repolarization abnormalities related to car-
diac memory persist when normal depolarization
is restored, and they resolve completely in
a month. The changes and their duration are
proportional to the amount of delivered ventric-
ular pacing [24]. Cardiac memory is associated
with complex biochemical abnormalities. Angio-
tensin inhibitors and AT-1 receptor blockers
attenuate the effects of short-term memory. Cal-
cium blockers reduce the impact of short-term
and long term-memory [25]. Long-term cardiac
memory involves de novo protein synthesis [26].
Differentiation of cardiac memory from ischemia
Shvilkin and colleagues [27] recently reported
that cardiac memory induced by RV pacing re-
sults in a distinctive T-vector pattern that allows
discrimination from ischemic precordial T-wave
inversions regardless of the coronary artery in-
volved. T-wave axis, polarity, and amplitude on
a 12-lead ECG during sinus rhythm were com-
pared between cardiac memory and ischemic pa-
tients (Fig. 12). The cardiac memory group
included 13 patients who were paced in the
DDD mode with a short entricular delay for
1 week after elective pacemaker implantation.
The ischemic group consisted of 47 patients with
precordial T-wave inversion identified among
228 consecutive patients undergoing percutaneous

Fig. 12. Circular histogram of frontal plane T-axes distribution in LAD, LCx, and CM groups. Solid bars indicate
LAD; hatched bars, LCx; open bars, CM. Difference in T-vector axis between CM and LAD/LCx is statistically signif-
icant (P ! 0.01). (From Shvilkin A, Ho KK, Rosen MR, et al. T-vector direction differentiates postpacing from ischemic
T-wave inversion in precordial leads. Circulation 2005;111:969–74; with permission.)
398
Fig. 13. ECG during VVI pacing showing very deep and symmetrical T-wave inversion in leads V3 to V6 in a patient
who presented with chest pain. These impressive abnormalities suggest ischemia or infarction and urgent coronary an-
giography should be considered with a view to performing angioplasty.
coronary intervention for an acute coronary syn-
drome. The combination of (1) positive T wave
in aVL, (2) positive or isoelectric T wave in lead
I, and (3) maximal precordial T wave inversion
OT-wave inversion in lead III was 92% sensitive
and 100% specific for cardiac memory, discrimi-
nating it from ischemic precordial T-wave inver-
sion regardless of the coronary artery involved.
Summary
Electrocardiographic criteria involving the
paced QRS complex are less sensitive but more
specific than primary ST abnormalities for MI
diagnosis during ventricular pacing. Although one
cannot determine with certainty the age of an MI
(hours, days, or even years), from a single ECG,
the presence of primary ST-segment abnormalities
strongly suggests the diagnosis of acute MI or
severe ischemia and need for possible emergency
revascularization. Patients with a history of chest
pain and a nondiagnostic paced ECG should
also be considered for emergency cardiac cathe-
terization with a view to performing
BAROLD et al
revascularization. A patient with the ECG shown
in Fig. 13 should certainly be a candidate for this
strategy.
References
[1] Barold SS, Falkoff MD, Ong LS, et al. Electrocar-
diographic diagnosis of myocardial infarction dur-
ing ventricular pacing. Cardiol Clin 1987;5:403–17.
[2] Hands ME, Cook EF, Stone PH, et al. Electrocar-
diographic diagnosis of myocardial infarction in
the presence of complete left bundle branch block.
Am Heart J 1988;116:23–31.
[3] Castellanos A Jr, Zoble R, Procacci PM, et al. St-qR
pattern. New sign of diagnosis of anterior myocar-
dial infarction during right ventricular pacing. Br
Heart J 1973;35:1161–5.
[4] Brandt RR, Hammil SC, Higano ST. Electrocardio-
graphic diagnosis of acute myocardial infarction
during ventricular pacing. Circulation 1998;97:
2274–5.
[5] Kafka W. ECG and VCG diagnosis of infarction in
pacemaker dependent patients [abstract]. Pacing
Clin Electrophysiol 1985;8:A-16.
[6] Kaul U, Anand IS, Bidwai PS, et al. Diagnosis of
myocardial infarction in patients during right ven-
tricular pacing. Indian J Chest Dis Allied Sci 1981;
23:68–72.
 ECG DIAGNOSIS OF MI AND ISCHEMIA DURING CARDIAC PACING
[7] Kindwall KE, Brown JP, Josephson ME. Predictive
accuracy of criteria for chronic myocardial infarc-
tion in pacing-induced left bundle branch block.
Am J Cardiol 1985;57:1255–60.
[8] Klein HO, Becker B, DiSegni E, et al. The pacing
electrogram. How important is the QRS complex
configuration? Clin Prog Electrophysiol 1986;4:
112–36.
[9] Kochiadakis GE, Kaleboubas MD, Igoumenidis NE,
et al. Electrocardiographic appearance of old myocar-
dial infarction in paced patients. Pacing Clin Electro-
physiol 2002;25:1061–5.
[10] Madias JE. The nonspecificity of ST-segment eleva-
tion O or ¼ 5.0 mm in V1–V3 in the diagnosis of
acute myocardial infarction in the presence of ven-
tricular paced rhythm. J Electrocardiol 2004;37:
135–9.
[11] Sgarbossa EB, Pinski SL, Gates KB, et al, for the
Gusto-1 Investigators. Early electrocardiographic
diagnosis of acute myocardial infarction in the pres-
ence of a ventricular paced rhythm. Am J Cardiol
1996;77:423–4.
[12] Sgarbossa EB. Recent advances in the electrocardio-
graphic diagnosis of myocardial infarction: left
bundle branch block and pacing. Pacing Clin Elec-
trophysiol 1996;19:1370–9.
[13] Caldera AE, Bryce M, Kotler M, et al. Angiographic
significance of a discordant ST-segment elevation
of R5 millimeters in patients with ventricular-
paced rhythm and acute myocardial infarction. Am
J Cardiol 2002;90:1240–3.
[14] Kochiadakis GE, Kaleboubas MD, Igoumenidis
NE, et al. Electrocardiographic diagnosis of acute
myocardial infarction in the presence of ventricular
paced rhythm. Pacing Clin Electrophysiol 2001;24:
1289–90.
[15] Barold SS. Diagnosis of myocardial ischemia during
ventricular pacing. Pacing Clin Electrophysiol 2000;
23:1060–1.
[16] Manyari DE, Klein GJ, Kostuk WJ. Electrocardio-
graphic recognition of variant angina during
399
permanent pacing. Pacing Clin Electrophysiol
1983;6:99–103.
[17] Mery D, Dagran O, Bailly E, et al. First and sec-
ond degree blocks (Wenckebach type) during right
ventricular stimulation in the course of Prinzme-
tal’s angina. Arch Mal Coeur Vaiss 1979;72:
385–90.
[18] Diaz CF, Ganim MS, Ellestad MH. Electrocar-
diographic evidence of ischemia during ventricu-
lar paced rhythms. Clin Cardiol 1996;19:520–2.
[19] Chaterjee K, Harris A, Davies G, et al. Electrocar-
diographic changes subsequent to artificial ventricu-
lar depolarization. Br Heart J 1969;31:770–9.
[20] Goldberger JJ, Kadish AH. Cardiac memory. Pac-
ing Clin Electrophysiol 1999;22:1672–9.
[21] Rosen MR. What is cardiac memory? J Cardiovasc
Electrophysiol 2000;11:1289–93.
[22] Patberg KW, Shvilkin A, Plotnikov A, et al. Cardiac
memory. Mechanisms, and clinical application.
Heart Rhythm 2005;2:1376–82.
[23] Goyal R, Syed ZA, Mukhopadhyay PS, et al.
Changes in cardiac repolarization following short
periods of ventricular pacing. J Cardiovasc Electro-
physiol 1998;9:269–80.
[24] Wecke L, Gadler F, Linde C, et al. Temporal charac-
teristics of cardiac memory in humans: vectorcardio-
graphic quantification in a model of cardiac pacing.
Heart Rhythm 2005;2:28–34.
[25] Plotnikov AN, Yu H, Geller JC, et al. Role of L-type
calcium channels in pacing-induced short-term and
long-term cardiac memory in canine heart. Circula-
tion 2003;107:2844–9.
[26] Patberg KW, Obreztchikova MN, Giardina SF,
et al. The cAMP response element binding protein
modulates expression of the transient outward cur-
rent: implications for cardiac memory. Cardiovasc
Res 2005;68:259–67.
[27] Shvilkin A, Ho KK, Rosen MR, et al. T-vector
direction differentiates postpacing from ischemic
T-wave inversion in precordial leads. Circulation
2005;111:969–74.
 Cardiol Clin 24 (2006) 401–411

Is Electrocardiography Still Useful in the Diagnosis

of Cardiac Chamber Hypertrophy and Dilatation?
Peter W. Macfarlane, DSc, FESC, FRCPa,b
a
Division of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow G12 8QQ, Scotland, UK
b
Royal Infirmary, 10 Alexandra Parade, Glasgow G31 2ER, Scotland, UK

With the ubiquitous availability of the echocar-
diograph, it is nowadays the case that a hospital
physician will seek to obtain an echocardiogram for
detailed information on ventricular function when
appropriate and in so doing obtain information on
the presence or absence of cardiac chamber hyper-
trophy or enlargement. In addition, family practi-
tioners are more easily able to refer patients to
a local hospital or cardiology practice for an
echocardiogram, although steps are having to be
taken, at least in the United Kingdom, to minimize
unnecessary referrals, particularly in patients who
have suspected heart failure, by first assessing the
ECG and measuring B-type natriuretic peptide
(BNP). In other health care systems, it is possible
that patients may be referred directly to a cardio-
logist for echocardiographic investigation as
required.
Notwithstanding, an ECG is always part of
a cardiologic work-up, and the question posed is
whether there is still value in reviewing the ECG
for evidence of changes related to chamber
enlargement when an echocardiogram can be
obtained if required.
Atrial enlargement
The P wave of the ECG is one of the smallest
components of the ECG waveform. For this
reason, accurate measurement is difficult, and
many criteria for P-wave abnormality are non-
specific as a result. It is generally accepted that
right atrial depolarization contributes to the
initial part of the P wave, whereas left atrial
E-mail address: peter.w.macfarlane@clinmed.gla.ac.uk
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.013
depolarization is responsible for the terminal
P-wave appearances. Thus, if there is any form
of right atrial abnormality, the P- wave duration
should not be increased, and the initial compo-
nent of the P wave may be increased in amplitude.
On the other hand, if there is a left atrial
abnormality, the duration of the P wave may be
lengthened, and in some leads there will be a more
obvious division of the P wave into two compo-
nents. Fig. 1 shows in schematic form how these
different changes may appear in lead II and lead
V1 of the 12-lead ECG.
There is some dispute over the terminology to
be used in cases of atrial abnormality. Left atrial
enlargement is said to arise from atrial dilatation
or pressure overload [1] or indeed from abnormal
intra-atrial conduction [2]. Thus, a term such as
‘‘left atrial abnormality’’ can be used to cover dif-
ferent forms of atrial pathology.
Waggoner and colleagues [3] reviewed ECG
criteria for left atrial enlargement and compared
findings against two-dimensional echocardio-
graph measures. They found that of 39 patients
who had false-positive ECG diagnoses of left
atrial abnormality according to echocardiograph
criteria, only 2 (5%) were free of organic heart
disease. Thus, their conclusion was that the
ECG detected left atrial abnormality rather than
left atrial enlargement. Waggoner and colleagues
[3] had used criteria such as P duration in lead
II of 120 milliseconds or longer, (negative P dura-
tion in V1)/(PR segment duration) of 1.0 or lon-
ger, and the P terminal force in V1 greater than
3 mVms. The second of these criteria is a form
of modified Macruz index, defined as P dura-
tion/PR segment (ie, end P to QRS onset) [4].
P terminal force in V1 is defined as the duration
cardiology.theclinics.com
402
Fig. 1. Atrial abnormalities. (A) If there is right atrial
enlargement, then the initial component of the P wave
is enlarged. (B) If there is left atrial enlargement, there
may be P-wave widening with an M-shaped P wave in
lead II and an increased P terminal force in V1.
of the negative (terminal) component of the P
wave times the amplitude of this component re-
ferred to baseline.
P mitrale is defined as an M-shaped P wave in
lead II with an increased P duration (Fig. 1). The
name suggests that this abnormality is found in
mitral valve disease, which is sometimes the
case, but this type of P wave also can be found
in patients who have constrictive pericarditis or
when there is an intra-atrial conduction defect.
In a study of 53 patients who had uncompli-
cated essential hypertension [5], ECG criteria of
left atrial abnormality such as P duration greater
than 0.12 milliseconds, P amplitude greater than
0.25 mV in lead II, Macruz index greater than
1.6 in lead II, and P terminal force in V1 of

4 mVms or less were assessed. (P terminal force
more negative than
4 mVms where the P termi-
nal amplitude is expressed as a negative value.)
Echocardiograph measurements of left atrial di-
mensions were obtained together with Doppler es-
timates of left ventricular (LV) filling. The Macruz
index had a sensitivity of 58.5% in detecting left
atrial abnormality and was the best of the ECG
measures. The authors concluded that ECG signs
of left atrial abnormality were related more to an
increased atrial workload, possibly secondary to
impaired ventricular filling, than to left atrial
enlargement [5].
Right atrial abnormality
Right atrial abnormality is not commonly
reported on an ECG. It can manifest as a tall P
wave in V1 with the P amplitude being greater
MACFARLANE
than an age-dependent value around 0.2 mV. It
may be found in patients who have congenital
heart disease, pulmonary hypertension, or ob-
structive airways disease.
Classically, a patient who has chronic obstruc-
tive airways disease may have so-called ‘‘P pul-
monale’’ characterized by a tall, peaked P wave of
at least 0.25-mV amplitude in lead II. There have
been few studies in patients who have this
abnormality, and most ECG criteria for right
atrial abnormality are nonspecific and somewhat
insensitive [6–8].
On occasions peaked P waves may be found in
inferior leads in the absence of any demonstrable
right atrial enlargement or dilatation, and in this
case the term ‘‘pseudo P pulmonale’’ is used. It
has been suggested that if left-sided heart disease
is present, such an ECG finding might reflect an
increase in left atrial dimensions [9].
Kaplan and colleagues [10] evaluated ECG cri-
teria for right atrial enlargement against two-di-
mensional echocardiograms in 100 patients who
had right atrial enlargement according to echocar-
diography and in 25 control patients. The most
powerful predictors of right atrial enlargement
were QRS axis greater than 90 , P amplitude in
V2 greater than 0.15 mV, and R/S greater than
1 in V1 in the absence of complete right bundle-
branch block (RBBB). The combined sensitivity
of these criteria was 49% with 100% specificity.
On the other hand, classic criteria for P pulmonale
were only 6% sensitive.
Bi-atrial enlargement
Occasionally an ECG may show P-wave ab-
normalities suggestive of both right and left atrial
enlargement, and in this case the term ‘‘bi-atrial
enlargement’’ may be used. Thus, criteria for bi-
atrial enlargement are essentially a combination
of those for right and left atrial enlargement.
Use of the signal-averaged ECG
In the last decade, there has been interest in
using the signal-averaged ECG to measure P-wave
duration and in using this parameter as an
indicator of atrial enlargement and even as
a prognostic index for the development of atrial
fibrillation. Dixen and colleagues [11] showed that
the signal-averaged P-wave duration was signifi-
cantly correlated with the left atrial diameter in
74 patients in whom a signal-averaged ECG and
echocardiogram had been recorded. The left atrial
volume, the right atrial volume, and in particular
 ECG IN THE DIAGNOSIS OF CHAMBER HYPERTROPHY
the total atrial volume were more strongly corre-
lated to the signal-averaged P-wave duration,
however.
On the other hand, Merckx and colleagues [12]
looked at correlations between the P-wave dura-
tion on the signal-averaged ECG and estimates
of atrial activation time derived from Doppler tis-
sue imaging. They showed that there was a signif-
icant correlation between the signal-averaged
ECG P-wave duration and atrial activation times
determined from the Doppler method, which was
much quicker to use than a signal-averaged ECG
when an echocardiogram was being recorded in
any event. Approximately 1 additional minute
was required. The conclusion was that this echo-
cardiographic estimate of atrial activation time
could be useful in detecting those patients prone
to develop atrial fibrillation.
In an interesting study using fractal measures,
Peters [13] showed that course F waves in patients
who had atrial fibrillation seemed to be more re-
lated to a larger left atrial size (R4.6 cm) than
were smaller fibrillatory waves.
Atrial abnormalitiesdconclusion
There are few areas where the ECG holds any
advantage over an echocardiogram in elucidating
atrial enlargement. Widened P waves as found in
intra-atrial conduction defects may, however, be
the only way of determining that there is some
form of atrial abnormality. Furthermore, the
signal-averaged ECG may be of some prognostic
value in predicting those patients at increased risk
of developing atrial fibrillation when an echocar-
diogram is not available.
Left ventricular enlargement
No ECG criteria have ever successfully sepa-
rated the different forms of LV abnormality (ie, LV
hypertrophy [LVH] caused by an increase in muscle
thickness or LV dilatation with increased LV cavity
volume). Huwez and colleagues [14] introduced
a new classification of LV geometry using M
mode ECG. This classification was based on the
use of an indexed LV mass greater than 131 g/m2
in men and 108 g/m2 in women together with
a knowledge of indexed LV volume, which was re-
garded as abnormal if it exceeded 90 mL/m2. This
classification led to four groupings based on normal
or abnormal indexed LV mass and normal or
abnormal indexed LV volume. Table 1 shows the
classifications. Extensive investigation of ECG
403
Table 1
Classification of left ventricular enlargement
Left Ventricular Massb
Left Ventricular
Volumea Normal Abnormal
Normal Normal Concentric left
ventricular
hypertrophy
Abnormal Isolated left Eccentric left
ventricular ventricular
volume hypertrophy
overload
a
Abnormal left ventricular volume is defined as vol-
ume greater than 90 mL/m2.
b
Abnormal left ventricular mass is defined as mass
greater than 131 g/m2 in men and mass greater than
108 g/m2 in women where indexing is with respect to
body surface area.
From Huwez FU, Pringle SD, Macfarlane PW. A
new classification of left ventricular geometry in patients
with cardiac disease based on M-mode echocardiogra-
phy. Am J Cardiol 1992;70:687.
appearances in a group of 202 cardiac patients
failed to elicit any ECG criteria that could separate
the different types of LV geometry, however [15].
Echocardiography itself is not without its
difficulties in estimating LV mass. On the other
hand, Rautaharju and colleagues [16] introduced
several equations for the calculation of LV mass
from the ECG. One set of equations for white
men and women is as follows:
LV mass ðmenÞ ¼ 0:026  CV þ 1:25  W
þ 34:4
LV mass ðwomenÞ ¼ 0:020  CV þ 1:12  W
þ 36:2
where LV mass equals LV mass in grams, CV
equals Cornell voltage (RaVL þ SV3) in micro-
volts [17], and W equals weight in kilograms. By
using these equations, Padmanabhan and col-
leagues [18] have shown that there are significant
heritable effects for ECG measures used in LV
mass estimation.
Influence of constitutional variables on ECG
criteria for left ventricular hypertrophy
Age and sex
It has been known for many years that normal
ECG voltages vary both with age and sex (eg, see
404 MACFARLANE
[19]). Voltages are highest in adolescence, particu-
larly in males, and decrease with increasing age,
with a leveling beyond approximately 50 years
of age. A similar trend can be seen for women, al-
though this is less marked. Significant differences
in upper limits of normal QRS voltages exist be-
tween men and women of similar age, as can be
seen in Fig. 2. These differences mean that any cri-
teria for ECG LVH and ECG right ventricular
hypertrophy (RVH) should be age and sex depen-
dent and explains why some ECG criteria for ven-
tricular hypertrophy that are not age and sex
adjusted perform in a suboptimal way.
QRS duration is approximately 7 milliseconds
longer in women than in men, but criteria that use
QRS widening as an index of LVH generally fail
to acknowledge this simple observation [19].
A fuller discussion of normal limits of ECG mea-
sures can be found elsewhere [19].
Other factors
Other factors that may affect voltage include
race. Blacks tend to have higher precordial
voltages than whites [20], and in turn it has been
shown that Chinese individuals have lower volt-
ages than whites [21–23] and hence also lower
Fig. 2. (A) The mean of SV1 þ RV5 in 725 male and 585
female adult healthy volunteers. (B) The upper normal
limits in the same data set.
voltages than blacks. Furthermore, increasing
body mass index is inversely linked with precor-
dial voltage, resulting in lower sensitivity and
higher specificity of precordial voltage criteria
for LVH in overweight individuals [24].
Selected criteria for left ventricular hypertrophy
Sokolow and Lyon index
Perhaps the best known of all ECG criteria for
LVH is that of Sokolow and Lyon introduced in
1949 [25]. These authors actually listed four crite-
ria, but the most commonly used amplitude crite-
rion is SV1 plus maximum (RV5, RV6) of 3.5 mV
or greater. It will be seen that this simple criterion
is neither age nor sex dependent and probably has
remained in use because the majority of patients
in whom LVH is likely to be found are over 50
years of age. It therefore is a very nonspecific cri-
terion in younger individuals, particularly men.
More recently, Alfakih and colleagues [26] sug-
gested new thresholds of 3.8 mV for men and
3.4 mV for women for reporting LVH, giving
a combined sensitivity of 20.3% at 95% specific-
ity. In the same study, the Cornell product (as dis-
cussed later) was 24.5% sensitive at the same
specificity of 95%. The author and colleagues’
data [27] suggest that in persons older than 50
years the upper limit of normal should be
4.6 mV for men and 3.6 mV for women. Clearly,
application of such thresholds would reduce sensi-
tivity even further. Interestingly, the same data
indicate that SV1 plus RV5 always has a higher
mean value in men and women than SV1 plus
RV6 (ie, for all age ranges and both sexes).
This could explain why the Sokolow and Lyon
criterion is often reduced to SV1 þ RV5 R
3.5 mV.
Cornell index
Two sets of criteria for LVH were published by
the group at Cornell University [17,28]. The sec-
ond of these, published in 1987, provided a simpler
set of criteria, as follows:
RaVL þ SV3 O 2:8 mV in men
RaVL þ SV3 O 2:0 mV in women
The Cornell group then introduced a voltage
times duration product known as the Cornell
product. In this case, the Cornell voltage was
multiplied by the overall QRS duration, and this
product showed improved accuracy in reporting
LVH [29,30]. Of particular importance is the fact
 ECG IN THE DIAGNOSIS OF CHAMBER HYPERTROPHY
that these ECG criteria can follow changes in
echocardiographic LV mass [31] and accurately
assess prognosis [32]. The opportunity to use
this more readily available index to track LV
mass may reduce the need for repeat echocardio-
grams. Note that in the Losartan Intervention
for Endpoint (LIFE) reduction in hypertension
study, the actual criteria used were adjusted on
the basis of recruitment experience to use a differ-
ential of 600 mV (rather than 800 mV, as given pre-
viously) between men and women; that is, in
women 600 mV was added to RaVL plus SV3 be-
fore deriving the Cornell product. The threshold
for abnormality was set at 244 mVms [32].
Secondary ST-T changes and left ventricular
hypertrophy
Although it is possible to diagnose LVH on the
basis of QRS changes only, as is evident from the
foregoing criteria, one particularly important
aspect of the ECG in assessing LVH is the
presence of so-called ‘‘secondary ST-T wave
changes,’’ which have a classic morphology as
shown in Fig. 3. Patients who have such abnor-
malities caused by LVH have been shown to
have a lengthened time interval from minimum
cavity dimension to mitral valve opening and a re-
duced rate of early diastolic wall thinning and di-
mension increase [33]. The importance of these
ECG changes is that they are well known to affect
prognosis adversely [34–36]. In fact, these changes
can add significantly to the Cornell product as
well as to the Sokolow and Lyon voltage for
the prediction of cardiovascular mortality and
Fig. 3. An ECG showing the typical features of LVH with secondary ST-T changes in the lateral leads in an 87-year-old
woman. Note the broad, deep, inverted terminal portion of the P wave in V1, left axis deviation, the very high Cornell
voltage, and the typical secondary ST-T changes, particularly in I and aVL. The QS in V2 may be caused by the LVH.
405
myocardial infarction in hypertensive patients
[36]. Furthermore, ST depression itself in V5 and
V6 has been shown to be a strong independent
predictor of LVH and increased LV mass [37].
In short, the presence of secondary ST-T changes
on the ECG is an independent predictor of
a poorer prognosis than might exist in the absence
of such a finding and shows that the ECG can
provide information that is complementary to
the echocardiogram.
In many hypertensive patients who have sec-
ondary ST-T changes, coronary artery disease is
present. Pringle and colleagues [38] in a study in-
volving 23 such patients, 20 of whom had concen-
tric LVH, found that 8 (40%) had significant
coronary artery disease at cardiac catheterization.
This finding suggests that secondary ST-T
changes in ECGs from hypertensive patients
need to be interpreted as caused by hypertrophy
with or without myocardial ischemia.
Composite criteria for left ventricular hypertrophy
Romhilt Estes criteria
The fact that ST-T changes are additive to
voltage criteria was recognized many years ago by
Romhilt and Estes [39] who introduced a point
scoring system for the ECG diagnosis of LVH.
Their criteria, although difficult to apply manu-
ally, can be used by computer programs, which
can also adapt voltage criteria for age and sex.
Table 2 shows in summary form the main criteria
used in this point score system. The Romhilt Estes
criteria also make use of the frontal plane QRS
axis, QRS duration, and intrinsicoid deflection
406
Table 2
Romhilt-Estes point score system for ECG diagnosis of
left ventricular hypertrophy*
Criteria
Any limb-lead R or S R 2.0 mV
or SV1 or SV2 R 3.0 mV
or RV5 or RV6 R 3.0 mV
ST-T is typical of LVHa
no digitalis
with digitalis
Left atrial involvement
P terminal force V1 O 4 mVms
LAD R 30
QRS duration R 90 ms
Intrinsicoid deflection V5 or V6 R 50 ms
* Five points indicates definite left ventricular hyper-
trophy; 4 points indicates probable left ventricular
hypertrophy.
a
ST-T configuration as in leads I, aVL of Fig. 3.
in V5 or V6 measured as the time from QRS onset
to the peak of the R wave.
Voltage- and interval-based criteria
More recently Salles and colleagues [40]
showed that a combination of QT-interval prolon-
gation and Cornell product resulted in increased
detection of LVH in patients who had resistant
hypertension. A prolonged QTc interval greater
than 440 milliseconds or prolonged QT dispersion
greater than 60 milliseconds together with a Cor-
nell product greater than 240 mVms was associ-
ated with a 5.3- to 9.3-fold higher chance of
having LVH compared with individuals who did
not have increased QTc or Cornell product.
Effect of echocardiographic criteria for increased
left ventricular mass on ECG criteria
Selection of echocardiographic criteria for in-
creased LV mass used as the reference standard
can undoubtedly influence the sensitivity of ECG
criteria for LVH. Cuspidi and colleagues [41]
showed in 100 untreated hypertensive patients
that the sensitivity of ECG criteria ranged from
9% to 25% depending on whether the LV mass
index was 126 g/m2 in men and 105 g/m2 in
women or 125 g/m2 in both men and women.
ECG LVH was based on either the Sokolow and
Lyon or the Cornell index being present. The first
criterion for increased echocardiographic LV
mass was indexed by height alone, whereas the
second was indexed by body surface area. Surpris-
ingly, the higher sensitivity was obtained in a crite-
rion that was not sex dependent.
MACFARLANE
Bundle-branch block and left ventricular
hypertrophy
An accurate diagnosis of LVH or LV mass in
Points
patients who have left bundle-branch block
3 (LBBB) is essentially impossible. On the other
hand, in a large series of over 1400 hearts examined
at post mortem, Havelda and colleagues [42] noted
3
that 93% of 70 hearts with ECG LBBB had LVH.
1 Thus, the presence of LBBB itself in many ways is
3 indicative of the presence of LVH with high speci-
ficity. In the presence RBBB, some ECG criteria
2 for LVH can still be applied, notably those involv-
1 ing left atrial abnormality [43,44].
1
Regression
Regression of LVH in patients receiving anti-
hypertensive therapy has been detected by assess-
ing the Cornell product after 1 year of treatment.
Those patients in whom the Cornell product
reduced in value had a higher probability of
regression of echocardiographic LVH indepen-
dent of changes in the systolic and diastolic blood
pressure [31]. Thus, the ECG can indeed also be
used to monitor LVH underlining the value of
a baseline ECG in patients with newly diagnosed
hypertension.
Prognosis of ECG left ventricular hypertrophy
More recent data on the prognostic value of
the ECG have come from the LIFE study [36] in
which the presence of secondary ST-T changes
was associated with a 1.5-fold increase in risk of
myocardial infarction or cardiovascular death
over a 5-year follow-up period. Conversely, Levy
and colleagues [45] found that a reduction in Cor-
nell voltage was linked with a lower risk of cardio-
vascular disease.
In the Heart Outcomes Prevention Evaluation
Study [46], the combined endpoint of either re-
gression of ECG LVH or prevention of progres-
sion by Sokolow Lyon voltage criteria following
ramipril-based therapy was associated with a re-
duced risk of myocardial infarction, stroke, con-
gestive heart failure, and death.
In a study of 19,434 veterans, composite
criteria were shown to be better predictors of
cardiovascular mortality than voltage-only crite-
ria [47]. In particular, a Romhilt Estes score of 5
or greater had a hazard ratio of 3.7 (95% confi-
dence interval, 3.0–4.4) after adjustment for age,
body mass index, and heart rate. This hazard ra-
tion compares with hazard ratios of 3.1 and 2.7
for Cornell voltage and product respectively.
 ECG IN THE DIAGNOSIS OF CHAMBER HYPERTROPHY
ECG in heart failure
Heart failure is regarded as unlikely to be
present in the absence of dyspnea and an abnor-
mal ECG or chest radiograph [48]. The ECG has
a high sensitivity (94%) when used in the diagno-
sis of heart failure but a low specificity (61%); it
has, however, an excellent negative predictive
value (98%) [49]. On the other hand, when a mea-
surement of plasma BNP or NTproBNP is avail-
able, a simple classification of the ECG into
normal and abnormal was found to be of little
value [50]. Future studies may clarify the role of
the ECG in avoiding unnecessary referrals for
echocardiography.
Left ventricular hypertrophydconclusion
Although clearly the echocardiogram gives
information on LV volume, wall thickness, and
LV function in particular, there is still much of
value that can be gained from the ECG in the
evaluation of LVH. The simplicity of the ap-
proach to recording the ECG, particularly in
clinical trials where patients can be followed for
several years, leads to the use of ECG abnormal-
ities as markers of risk. Indeed, in the Framing-
ham study, increased QRS voltage together with
ST-T abnormalities conferred a risk similar to
that of a previous myocardial infarction [51]. In
many situations where a simple approach to treat-
ment is required, such as in essential hypertension,
a baseline ECG is of value in monitoring the pa-
tient’s condition at least on an annual basis.
Such an investigation can be undertaken in the
community, thereby avoiding an unnecessary visit
to hospital for an echocardiogram recording.
Right ventricular hypertrophy
The ECG is notoriously inadequate in detect-
ing right ventricular enlargement or hypertrophy.
In an early study of Flowers and Horan [52], it
was shown that criteria using V1 were very specific
in detecting RVH, at 90%, but were insensitive
(2%–18%).
The vectorcardiogram is rarely used these
days, but Chou and Helm [53] described three
types of RVH patterns. Essentially these patterns
broadly translate to the 12-lead ECG as follows:
Type A: Tall R in V1, prominent S in V6
(counterclockwise inscription of the trans-
verse QRS loop)
407
Type B: R/S greater than 1 in V1 with R
greater than 0.5 mV (clockwise inscription
of the transverse plane QRS loop)
Type C: Prominent S in V5 and V6 with R/S
less than 1 in V5 (clockwise inscription of
the transverse plane QRS loop)
The first of these criteria essentially relates to
severe RVH, as in pulmonary stenosis. The
second may be associated with rheumatic heart
disease. On the other hand, type C is more linked
with chronic obstructive pulmonary disease and
sometimes mitral stenosis.
Secondary ST-T changes and right ventricular
hypertrophy
In a fashion similar to LVH, secondary ST-T
changes caused by right ventricular enlargement
may be encountered. A similar pattern of ST
depression with asymmetric T wave inversion is
typical in V1 and V2 (Fig. 4). This pattern is en-
countered most frequently in patients who have
congenital heart disease.
Other factors
An increased R/S ratio in V1 can sometimes be
caused by posterior myocardial infarction. In such
a case, the R-wave duration in V1 generally
exceeds 40 milliseconds, and the T wave is
upright. In the presence of inferoposterior myo-
cardial infarction or even lateral myocardial in-
farction, an increased R/S in V1 should not be
misinterpreted as being caused by RVH.
Right bundle-branch block and right ventricular
hypertrophy
The presence of RBBB in general terms makes
it difficult to report RVH. The combination of
known RVH with RBBB is most often seen in
congenital heart disease, but a tall R in V1 and V2
in patients who have RBBB is indeed nonspecific
for RVH.
Right ventricular hypertrophydconclusion
The ECG criteria for RVH are so poor that an
echocardiogram normally will be more valuable
than an ECG in patients who have congenital
heart disease or pulmonary disease. In pulmonary
embolism, however, the right ventricular volume
may increase, leading to ECG changes, such as
T-wave inversion in V1 to V3, which are more
readily followed on the ECG.
408
Fig. 4. An ECG showing RVH with secondary ST-T changes in a 3-year-old boy who has pulmonary valvular stenosis.
Note that V4R is used to the exclusion of V3. The upright P waves in V1 and V2 have an abnormally high amplitude for
age, as do the R waves in V1 and V2. T-wave inversion may be normal in V4R–V4 in infants and children, but the ST
depression and overall ST-T configuration in this illustration is that of a secondary ST-T change.
Biventricular hypertrophy
In general terms, criteria for biventricular
hypertrophy essentially are a combination of the
individual criteria for LVH or RVH. There have
been a number of studies linking ECGs with
anatomic biventricular hypertrophy as determined
at post mortem. The best of these showed
a sensitivity of 20%, although the specificity was
high at 94% [6].
Athlete’s heart
With the increased advocacy of participation
in sport, there has been an increase in the number
of deaths associated with a variety of athletic
activities. In one recent half-marathon in the
northeast of England (The Great North East
Run, September, 2005), four men died as a result
of participating. It would seem to be prohibitive
to consider undertaking an echocardiogram in all
asymptomatic younger individuals with a view to
detecting cardiomyopathy unless they are likely to
be serious athletes. Data from Italy indicate one
sudden death per year in 100,000 individuals aged
12 to 35 years in the general population, with an
increase to 2.3 sudden deaths per 100,000 in
those who participated in sporting activities [54].
Under these circumstances, an ECG or an echo-
cardiogram is not likely to be cost effective in
terms of the yield of abnormalities detected. The
matter of screening young athletes is topical and
controversial [55]. Some athletes may have
MACFARLANE
arrhythmogenic right ventricular dysplasia. This
condition may rarely manifest as an epsilon wave
on the ECG; additional criteria relating to differ-
ences in the sum of QRS durations in V1 þ V2 þ
V3 versus V4 þ V5 þ V6 have been suggested as be-
ing of value [56]. An MRI is more informative in
skilled hands, however. On the other hand, when
individuals are known to have heart disease, guide-
lines for advising patients wishing to take part in
competitive sport have recently been issued [57,58].
Summary
The echocardiogram undoubtedly is part of the
cardiologist’s armamentarium in the diagnosis
and elucidation of cardiac abnormalities. The
numbers of echocardiograms recorded annually
are almost certainly increasing in every hospital
and private practice, but the ECG still continues
to be the most frequently recorded noninvasive
test in medicine. Except in extreme cases, the ECG
can almost always be recorded, whereas it is
generally accepted that the echocardiogram may
be unsatisfactory in an admittedly diminishingly
small percentage of individuals referred. For
many patients, particularly those who have newly
diagnosed hypertension, a 12-lead ECG recording
may be the only test that is required as a baseline
measure. For those who have possible heart
failure, an ECG and BNP measurement may be
sufficient to obviate the need for an echocardio-
gram. On the other hand, there is no point in
denying that an echocardiogram is recorded as
 ECG IN THE DIAGNOSIS OF CHAMBER HYPERTROPHY
part of the routine investigation of many patients,
including those who have had a recent myocardial
infarction, which may of course in the first place
have been diagnosed by an ECG! Thus the two
techniques will continue to live side-by-side for the
foreseeable future.
Acknowledgments
The author wishes to thank Dr. David P
Macfarlane, of Glasgow Royal Infirmary, and
Dr. Peter Okin, of Cornell University, New York,
who provided many relevant references of interest.
References
[1] Josephson ME, Kastor JA, Morganroth J. Electro-
cardiographic left atrial enlargement. Electrophysio-
logic, echocardiographic and hemodynamic
correlates. Am J Cardiol 1977;39:967–71.
[2] Chandraratna PAN, Langevin E. On the signifi-
cance of an abnormal P-terminal force in lead VI.
Am Heart J 1978;95:267–8.
[3] Waggoner AD, Adyanthaya AV, Quinones MA,
et al. Left atrial enlargement. Echocardiographic as-
sessment of electrocardiographic criteria. Circula-
tion 1976;54:553–7.
[4] Macruz R, Perloff JK, Case RH. A method for the
electrocardiographic recognition of atrial enlarge-
ment. Circulation 1958;17:882–9.
[5] Genovesi-Ebert A, Marabotti C, Palombo C, et al.
Electrocardiographic signs of atrial overload in hy-
pertensive patients: indexes of abnormality of atrial
morphology or function. Am Heart J 1991;121:
1113–8.
[6] Murphy ML, Thenabadu PN, de Soyza N, et al.
Reevaluation of electrocardiographic criteria for
left, right and combined cardiac ventricular hyper-
trophy. Am J Cardiol 1984;53:1140–7.
[7] Cacho A, Prakash R, Sarma R, et al. Usefulness of
two-dimensional echocardiography in diagnosing
right ventricular hypertrophy. Chest 1983;84:1547.
[8] Kushner FG, Lam W, Morganroth J. Apex sector
echocardiography in evaluation of the right atrium
in patients with mitral stenosis and atrial septal de-
fect. Am J Cardiol 1978;42:7337.
[9] Chou T-C, Helm RA. The pseudo P pulmonale. Cir-
culation 1965;32:96–105.
[10] Kaplan JD, Evans GT Jr, Foster E, et al. Evaluation
of electrocardiographic criteria for right atrial en-
largement by quantitative two-dimensional echocar-
diography. J Am Coll Cardiol 1994;23:747–52.
[11] Dixen U, Joens C, Rasmussen BV, et al. Signal-aver-
aged P wave duration and the dimensions of the atria.
Ann Noninvasive Electrocardiol 2004;9:309–15.
[12] Merckx KL, De Vos CB, Palmans A, et al. Atrial ac-
tivation time determined by transthoracic Doppler
tissue imaging can be used as an estimate of the total
409
duration of atrial electrical activation. The Journal
of the American Society of Echocardiography 2005;
18:940–4.
[13] Peters RM. The fractal dimension of atrial fibrilla-
tion: a new method to predict left atrial dimension
from the surface electrocardiogram. Cardiology
1999;92:17–20.
[14] Huwez FU, Pringle SD, Macfarlane PW. A new clas-
sification of left ventricular geometry in patients
with cardiac disease based on M-mode echocardiog-
raphy. Am J Cardiol 1992;70:681–8.
[15] Huwez FU. Electrocardiography of the left ventricle
in coronary artery disease and hypertrophy [Ph D
thesis]. Glasgow (Scotland): University of Glasgow;
1990.
[16] Rautaharju PM, Parks LP, Gottdiener JS, et al.
Race- and sex-specific ECG models for left ventricu-
lar mass in older populations. Factors influencing
overestimation of left ventricular hypertrophy prev-
alence by ECG criteria in African-Americans. J Elec-
trocardiol 2000;33:205–18.
[17] Casale PN, Devereux RB, Alonso DR, et al. Im-
proved sex specific criteria of left ventricular hyper-
trophy for clinical and computer interpretation of
electrocardiograms: validation with autopsy find-
ings. Circulation 1987;75:565–72.
[18] Padmanabhan S, Connell JMC, Dominiczak AF,
et al. Heritability and genetic determinants of electro-
cardiographic measures of left ventricular massd
a two generation family study. J Hypertens 2004;
22:S180–1.
[19] Macfarlane PW, Lawrie TDV. The normal
electrocardiogram and vectorcardiogram. In:
Macfarlane PW, Lawrie TDV, editors. Comprehen-
sive electrocardiology. Oxford (UK): Pergamon
Press; 1989. p. 407–57.
[20] Rautaharju PM, Zhou SH, Calhoun HP. Ethnic dif-
ferences in ECG amplitudes North American white,
black and Hispanic men and women. J Electrocar-
diol 2004;27(Suppl):20–31.
[21] Chen CY, Chiang B, Macfarlane PW. Normal limits
of the electrocardiogram in a Chinese population.
J Electrocardiol 1989;22:1–15.
[22] Yang T-F, Macfarlane PW. Comparison of the de-
rived vectorcardiogram in apparently healthy whites
and Chinese. Chest 1994;106:1014–20.
[23] Wu J, Kors JA, Rijnbeek PR, et al. Normal limits of
the electrocardiogram in Chinese subjects. Int J Car-
diol 2003;87:37–51.
[24] Okin PM, Roman MJ, Devereux RB, et al. Electro-
cardiographic identification of left ventricular hy-
pertrophy: test performance in relation to
definition of hypertrophy and presence of obesity.
J Am Coll Cardiol 1996;27:124–31.
[25] Sokolow M, Lyon T. The ventricular complex in left
ventricular hypertrophy as obtained by unipolar pre-
cordial and limb leads. Am Heart J 1949;37:161–86.
[26] Alfakih K, Walters K, Jones T, et al. New gender-
specific partition values for ECG criteria of left
410 MACFARLANE
ventricular hypertrophy. Recalibration against car-
diac MRI. Hypertension 2004;44:175–9.
[27] Macfarlane PW, Lawrie TDV, editors. Comprehen-
sive electrocardiology, vol. 3. Oxford (UK): Perga-
mon Press; 1989.
[28] Casale PN, Devereux RB, Kligfield P, et al. Electro-
cardiographic detection of left ventricular hyper-
trophy: development and prospective validation of
improved criteria. J Am Coll Cardiol 1985;6:
572–80.
[29] Molloy TJ, Okin PM, Devereux RB, et al. Electro-
cardiographic detection of left ventricular hypertro-
phy by the simple QRS voltage-duration product.
J Am Coll Cardiol 1992;20:1180–6.
[30] Okin PM, Roman MJ, Devereux RB, et al. Electro-
cardiographic identification of increased left ventric-
ular mass by simple voltage-duration products.
J Am Coll Cardiol 1995;25:417–23.
[31] Okin PM, Devereux RB, Liu JE, et al. Regression of
electrocardiographic left ventricular hypertrophy
predicts regression of echocardiographic left ventric-
ular mass: The LIFE Study. J Human Hypertens
2004;18:403–9.
[32] Okin PM, Devereux RB, Jern S, et al. Regression of
electrocardiographic left ventricular hypertrophy
during antihypertensive treatment and prediction
of major cardiovascular events: The LIFE Study.
JAMA 2004;292:2343–9.
[33] Moore RB, Shapiro LM, Gibson DG. Relation be-
tween electrocardiographic repolarisation changes
and mechanical events in left ventricular hypertro-
phy. Br Heart J 1984;52:516–23.
[34] Verdecchia P, Schillaci G, Borgioni C, et al. Prog-
nostic value of a new electrocardiographic method
for diagnosis of left ventricular hypertrophy in es-
sential hypertension. J Am Coll Cardiol 1998;31:
383–90.
[35] Kannel WB, Gordon T, Offut D. Left ventricular hy-
pertrophy by electrocardiogram: prevalence, inci-
dence, and mortality in the Framingham Study.
Ann Intern Med 1969;71:89–105.
[36] Okin PM, Devereux RB, Nieminen MS, et al. Elec-
trocardiographic strain pattern and prediction of
cardiovascular morbidity and mortality in hyperten-
sive patients. Hypertension 2004;44:48–54.
[37] Okin PM, Devereux RB, Fabsitz RR, et al. Quanti-
tative assessment of electrocardiographic strain pre-
dicts increased left ventricular mass: The Strong
Heart Study. J Am Coll Cardiol 2002;40:1395–400.
[38] Pringle SD, Macfarlane PW, McKillop JH, et al.
Pathophysiologic assessment of left ventricular hy-
pertrophy and strain in asymptomatic patients
with essential hypertension. J Am Coll Cardiol
1989;13:1377–81.
[39] Romhilt DW, Estes EH Jr. A point-score system for
the ECG diagnosis of left ventricular hypertrophy.
Am Heart J 1968;75:752–8.
[40] Salles G, Leocadio S, Bloch K, et al. Combined QT
interval and voltage criteria improve left ventricular
hypertrophy detection in resistant hypertension.
Hypertension 2005;46:1207–12.
[41] Cuspidi C, Macca G, Sampieri L, et al. Influence of
different echocardiographic criteria for detection of
left ventricular hypertrophy on cardiovascular risk
stratification in recently diagnosed essential hyper-
tensives. J Hum Hypertens 2001;15:619–25.
[42] Havelda CJ, Sohi GS, Flowers NC, et al. The path-
ologic correlates of the electrocardiogram: complete
left bundle branch block. Circulation 1982;65:
445–51.
[43] Murphy ML, Thenabadu PN, de Soyza N, et al. Left
atrial abnormality as an electrocardiographic crite-
rion for the diagnosis of left ventricular hypertrophy
in the presence of right bundle branch block. Am J
Cardiol 1983;52:381–3.
[44] Vandenberg B, Sagar K, Romhilt D. Electro-
cardiographic criteria for the diagnosis of left ven-
tricular hypertrophy in the presence of complete
right bundle branch block. J Am Coll Cardiol
1985;5:511.
[45] Levy D, Salomon M, D’Agostino RB, et al. Prog-
nostic significance of baseline electrocardiographic
features and their serial changes in subjects with
left ventricular hypertrophy. Circulation 1994;90:
1786–93.
[46] Matthew J, Sleight P, Lonn E, et al. Reduction of
cardiovascular risk by regression of electrocardio-
graphic markers of left ventricular hypertrophy by
the angiotensin-converting enzyme inhibitor rami-
pril. Circulation 2001;104:1615–21.
[47] Hsieh BP, Pham MX, Froelicher VF. Prognostic
value of electrocardiographic criteria for left ventric-
ular hypertrophy. Am Heart J 2005;150:161–7.
[48] Struthers AD. The diagnosis of heart failure. Heart
2000;84:334–8.
[49] Zaphiriou A, Robb S, Murray-Thomas T, et al. The
diagnostic accuracy of plasma BNP and NTproBNP
in patients referred from primary care with suspected
heart failure: results of the UK natriuretic peptide
study. Eur J Heart Fail 2005;7:537–41.
[50] Kannel WB, Abbott RD. Comparison of ECG-
LVH and unrecognised myocardial infarction as
predictors of overt cardiovascular events: The
Framingham study [abstract]. Circulation 1984;
70(Suppl. II):434.
[51] Dosh SA. Diagnosis of heart failure in adults. Am
Fam Physician 2004;70:2145–52.
[52] Flowers NC, Horan LG. Subtle signs of right ven-
tricular enlargement and their relative importance.
In: Schlant RC, Hurst JW, editors. Advances in elec-
trocardiography. New York: Grune and Stratton;
1972. p. 297–308.
[53] Chou T-C, Helm R. Clinical vectorcardiography.
New York: Grune and Stratton; 1967.
[54] Corrado D, Basso C, Rizzoli G, et al. Does sport ac-
tivity enhance the risk of sudden death in adolescents
and young adults? J Am Coll Cardiol 2003;42:
1959–63.
 ECG IN THE DIAGNOSIS OF CHAMBER HYPERTROPHY 411

[55] Fuller C. Physical examinations for young athletes
[letter]. Cleve Clin J Med 2005;72:176.
[56] A multidisciplinary study of right ventricular dysplasia.
Available at: www.arvd.org. Accessed June 9, 2006.
[57] Maron BJ, Zipes DP. 36th Bethesda Conference:
Introduction Eligibility recommendations for
competitive athletes with cardiovascular abnormali-
ties. J Am Coll Cardiol 2005;45:1318–21.
[58] European Society of Cardiology consensus docu-
ment. Recommendations for competitive sports par-
ticipation in athletes with cardiovascular disease.
Eur Heart J 2005;26:1422–45.
 Cardiol Clin 24 (2006) 413–426
Electrocardiography of the Failing Heart
Vinzenz Hombach, MD
Department of Internal Medicine II, University Hospital of Ulm,
Robet-Koch Strasse 8, Ulm D-89081, Germany
Structural heart disease, electrical instability,
and increased sympathetic activity can generate
a number of specific and nonspecific ECG changes
and arrhythmias in patients with congestive heart
failure (CHF). This review describes direct alter-
ations of the P–QRS–T complex and ECG-
derived parameters in CHF, together with the
significance of cardiac arrhythmias, markers of
atrial and ventricular electrical instability, and the
parameters of sympathetic nervous system
activity.
General ECG alterations in the failing heart
Patients with CHF may display specific ECG
alterations such as Q-waves after myocardial
infarction (MI) or persistent ST-segment elevation
in MI-related leads consistent with a left ventric-
ular (LV) aneurysm. ST-segment alterations may
also occur in chronic recurrent ischemia. ECG
patterns of significant LV hypertrophy (LVH)
may indicate chronic decompensation from aortic
valvular disease, arterial hypertension, or hyper-
trophic cardiomyopathy. LVH also carries prog-
nostic significance as documented in the Heart
Outcomes Prevention Evaluation Study of 9541
patients. Electrocardiographic LVH was present
in 793 (8.3%), and of these, 19.0% sustained
a major cardiovascular event (MI, stroke, or car-
diovascular death); 15.6% died, and 6.1% de-
veloped CHF after a follow-up of 4.5 years,
compared with 15.6%, 10.8%, and 2.9%, respec-
tively, in those without ECG evidence of LVH
(P ¼ 0.0023, P ! 0.0001, and P ! 0.0001). In
multivariate analysis, ECG-LVH was an indepen-
dent predictor of cardiovascular and all-cause
E-mail address: vinzenz.hombach@uniklinik-ulm.de
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.03.002
mortality and CHF [1]. Left bundle branch block
(LBBB) in CHF patients almost always reflects
diffuse myocardial damage as a result of ischemic
or nonischemic dilated cardiomyopathy. Incom-
plete or complete right bundle branch block with
right axis deviation and P-wave-alteration
strongly suggests cor pulmonale. Finally, nonspe-
cific ECG changes may be induced by digitalis or
antiarrhythmic drugs.
The conventional ECG may be used as a first-
line diagnostic tool for CHF, as shown in the
Epidemiologia da Insufficiencia Cardiaca study
[2]. In this investigation, 6300 subjects in a general
population were screened for CHF by symptoms
or signs, chest X-ray, ECG, and echocardiogra-
phy. The diagnosis was confirmed in 551 cases.
Patients with right atrial enlargement, atrial flut-
ter or fibrillation, first degree and second degree
Mobitz type II atrio-ventricular (AV)-block,
LBBB, interstitial lung edema, and bilateral pleu-
ral effusion were more likely to be diagnosed with
CHF. An abnormal ECG had an estimated sensi-
tivity of 81% and a negative predictive value of
75%, and for an abnormal chest X-ray the num-
bers were 57% and 83%, respectively [2].
Atrial depolarization and repolarization
in the conventional and signal-averaged ECG
P-wave-duration (PWD) in the conventional or
signal-averaged ECG and P-wave-dispersion (Pd)
on the standard ECG are considered noninvasive
markers of intra-atrial conduction disturbances
that predispose to atrial fibrillation (AF). Sanders
and colleagues [3] using electroanatomic mapping,
compared electrophysiologic markers such as
right and left atrial refractory periods, conduction
times, corrected sinus node recovery time, PWD,
and conduction of the crista terminalis in 21
cardiology.theclinics.com
414 HOMBACH
patients with symptomatic CHF and 21 age-
matched controls. CHF patients demonstrated
an increase in atrial effective refractory period,
no change in the heterogeneity of refractoriness,
an increase of atrial conduction time along the
low right atrium and coronary sinus, prolongation
of the PWD, and corrected sinus node recovery
times as well as a greater number and duration
of double potentials along the crista terminalis.
CHF patients also demonstrated an increased
propensity to AF with single extrastimuli and in-
duced AF was more often sustained. Thus, simple
PWD measurement of O20 milliseconds in the
conventional ECG may be a helpful parameter
for the increased susceptibility of AF in CHF
patients.
PWD may be related to the clinical status of
CHF patients. Song and colleagues [4] measured
minimum and maximum PWD and Pd in 21 pa-
tients with decompensated CHF before and after
40 G 23 hours of diuresis with the removal of
3 G 1 L of fluid. There was a significant correla-
tion between average PWD and the amount of
removed fluid (r ¼
0.59, P ¼ 0.015) and the
average and maximum PWD decreased signifi-
cantly with diuresis (P ¼ 0.001 and 0.022, respec-
tively). Diuresis may, therefore, attenuate the
electrophysiologic atrial changes caused by fluid
overload. Beta-blockade may also influence maxi-
mum PWD and Pd according to Camsari [5] in
a study of 41 New York Heart Association
(NYHA) class III and IV patients. At the 6-month
follow-up, maximum PWD and Pd decreased sig-
nificantly with the administration of metoprolol,
a finding that may correlate with the effect of beta-
blockade in reducing the incidence of AF in CHF.
In the study of Faggiano and colleagues [6] an
increased PWD within the signal-averaged ECG
(SAECG) correlated with an increased pulmonary
capillary wedge (PCW) pressure), and an acute re-
duction of PCW pressure-reduced PWD. How-
ever, the left atrial end-systolic dimension did
not vary significantly with the PCW pressure. In
the recent study of Yamada and colleagues [7] in-
volving 75 patients without a history of parox-
ysmal AF and a LV ejection fraction (LVEF)
!40%, an abnormal P-SAECG was found in
29 of 75 patients. After a follow-up of 21 G 9
months, attacks of paroxysmal AF occurred
more frequently in patients with (32%) than in
those without an abnormal P-SAECG (2%, P ¼
0.0002). Plasma Atrial Natriuretic Peptide
(ANP)-levels were significantly higher in patients
with than in those without paroxysmal AF
attacks. Multivariate analysis identified an abnor-
mal P-SAECG (hazard ratio 19.1, P ¼ 0.0069)
and elevated ANP levels of 60 pg/mL (hazard ra-
tio 8.6, P ¼ 0.018) as important predictors of
paroxysmal AF. Yamada and colleagues [7] con-
cluded that an abnormal P-SAECG and elevated
ANP levels might be predictors of paroxysmal
AF in patients with chronic CHF. In a recent
study, Dixen and colleagues [8] evaluated the
prognostic role of prolonged signal-averaged
PWD, raised levels of natriuretic peptide in 43
consecutive patients with stable CHF. Over
a 438-day median follow-up, time to death, hospi-
talization for CHF exacerbation, and ECG-docu-
mented AF were investigated. Seventeen patients
met these endpoints. Proportional hazard regres-
sion showed that only prolonged signal averaged
PWD R49 milliseconds was associated with a in-
creased risk of meeting an early endpoint (hazard
ratio 3.94 with 95% confidence interval [CI], 1.50–
10.42, P ¼ 0.006). These workers concluded that
in patients starting with stable CHF, a prolonged
single averaged P-wave duration predicts early
death, development of AF, or hospitalization for
decompensated CHF.
Interventricular septal activationdabsence
of septal Q waves
It has been argued that the absence of septal Q
waves may be a marker of LV disease. In a recent
study on 110 elderly patients O65 years, Shamim
and colleagues [9] analyzed standard 12-lead
ECGs for the presence or absence of Q waves in
leads I, aVL, V5, and V6, and patient survival
from hospital and outpatient records for a mean
follow-up of 4 years. Septal Q waves were absent
in 71 and present in 39 patients. The overall mor-
tality rate was 47% (43 patients), and the inci-
dence of death was 49% (36 patients) in the
group with no septal Q waves and 18% (7 pa-
tients) in those with Q waves. On univariate anal-
ysis (Cox proportional hazard method) the
absence of septal Q waves was found to be a strong
marker of a poor prognosis in chronic CHF
(hazard ratio [HR] 1.40, 95% CI, 1.10–1.67, P ¼
0.003). Kaplan-Meier curves in the two groups
showed a significant difference in survival inde-
pendenly of age, NYHA functional class, peak
VO2, and QRS duration. Septal Q waves reflect
early septal activation (20 milliseconds) from left
to right across the mid-portion of the interventric-
ular septum. Such normal septal Q waves are of-
ten absent in LBBB and after left coronary
 ELECTROCARDIOGRAPHY OF THE FAILING HEART
artery occlusion. Thus, absent septal Q waves are
considered an indicator of structural heart disease
and myocardial fibrosis, and may therefore serve
as an independent predictor of poor prognosis in
elderly CHF patients [9].
QRS duration
Measurement of QRS duration may be di-
agnostically useful in systolic LV dysfunction, as
indicated by Krüger and colleagues [10] in a cohort
of 128 consecutive patients with suspected cardiac
disease. Sixty-six patients had LV systolic dys-
function judged by LVEF !50% determined
echocardiographically. The QRS duration was
longer in the group with LV dysfunction com-
pared with the 62 patients without LV systolic
dysfunction, and Brain Natriuretic Peptide (BNP)
(normal value !75 pg/mL) was higher in the group
with LV systolic dysfunction compared with the
control group. A QRS duration of O0.1, O0.11,
or O0.12 was highly specific (63%, 90%, and
98%) but less sensitive (84%, 81%, and 75%) for
predicting LV systolic dysfunction. A QRS cutoff
value of 106 milliseconds was moderately sensitive
(65%) but very specific (87%) for predicting LV
systolic dysfunction, whereas a BNP cutoff value
O84 pg/mL was highly sensitive (89%) but only
modestly specific (58%). The positive likelihood
ratio for LV systolic dysfunction of an abnormal
BNP (2.0) and QRS prolongation O100 millisec-
onds (2.3) was improved by the combination of
both criteria to 5.1. In multivariate analysis, BNP
and QRS duration were independent predictors
of LV systolic dysfunction.
QRS duration also plays a significant role
in the prognostic assessment of CHF. In a group
of 165 patients with implantable cardioverter-
defibrillators (ICD) with CHF NYHA functional
Class III the prognostic significance of a QRS
duration of R150 milliseconds (n ¼ 26, group 1)
was compared with a QRS duration %150 milli-
seconds (n ¼ 139, group 2). At the 12-month
follow-up the mortality in group 1 was signifi-
cantly higher (31.3%) compared with 9.5% in
group 2, and after 24 months the mortality rate
was 46.6% in group 1 compared with 10.2% in
group 2 (P ¼ 0.04). A QRS duration of R150
milliseconds in ICD recipients with CHF appears
to provide a simple noninvasive prognostic
index [11].
Iuliano and colleagues [12] undertook a retro-
spective analysis of 669 CHF patients according
to QRS duration !120 milliseconds or R120
415
milliseconds. They found that a prolonged QRS
was associated with a significant increase in mor-
tality (49.3% versus 34.0%, P ¼ 0.01) and sudden
death (24.8% versus 17.4%, P ¼ 0.004). LBBB
was associated with a worse survival (P ¼
0.006) but not sudden death. In patients with
LVEF !30%, QRS prolongation was associated
with a significant increase in mortality (51.6%
versus 41.1%, P ¼ 0.01) and sudden death
(28.8% versus 21.1%, P ¼ 0.02). In patients
with an LVEF of 30% to 40% QRS prolongation
was associated with a significant increase in mor-
tality (42.7% versus 23.3%, P ¼ 0.036), but not in
sudden death. After adjustment for baseline vari-
ables independent predictors of mortality in-
cluded prolongation of QRS (risk ratio 1.46,
P ¼ 0.028) and depressed LVEF (risk ratio
0.965, P ¼ 0.001). The results indicated that
QRS prolongation of R120 milliseconds was an
independent predictor of increased total mortality
and sudden death in CHF patients. Morgera and
colleagues [13] reported similar observations in
a study of 78 consecutive patients with idiopathic
dilated cardiomyopathy who were investigated
with signal averaged ECG, 24- to 48-hour ECG
monitoring and electrophysiologic study. During
a follow-up of 85 months, nine patients died: six
of cardiac death, two of CHF. The independent
predictors for death and cardiac transplantation
included an his-ventricular (HV) interval O55
milliseconds and a combination of frequent and
repetitive ventricular ectopic beats and poor LV
function. The association of a prolonged HV in-
terval with a wide (O110 milliseconds) QRS com-
plex yielded a strong index of future arrhythmic
events (ratio 4, 53, 95% CI, 1.75–13.4, P ¼
0.05). These results were confirmed by Shenkman
and colleagues [14], who selected a subgroup of
3471 CHF patients with full echocardiographic
and ECG data. The QRS duration was R120 mil-
liseconds in 20.8% of the subjects. There was a lin-
ear relationship between increased QRS duration
and decreased LVEF (P ! 0.01). A prolonged
QRS duration of 120 to 149 milliseconds was as-
sociated with an increased mortality at 60 months
(P ¼ 0.001), when adjusted for age, sex, and race
(P ¼ 0.001), and LV systolic dysfunction. A
graded increase in mortality was correlated with
ascending degrees of QRS prolongation. The
study confirmed that 20% of a general CHF pop-
ulation can be expected to have a prolonged QRS
duration within the first year of diagnosis reflect-
ing a worse prognosis, and pointing to possible
candidates for biventricular pacing.
416
In elderly patients the prognostic significance
of a prolonged QRS complex was also shown by
Shamim and colleagues [15] in 112 patients (mean
age of 73.3 G 4.4 years). During a follow-up of
12 G 5 months, 45 patients died, and QRS dura-
tion (P ¼ 0.001) and heart rate (P ¼ 0.03) at base-
line were found by Cox proportional hazard
analysis to predict adverse outcomes. Kaplan-
Meier survival analysis revealed that patients
with !5% change of QRS duration had fewer
complications than patients with a 5% to 20%
change and a O20% change was associated with
the worst prognosis. Thus, a single measurement
of prolonged QRS duration or more importantly
an increase in QRS duration over time predicts
adverse outcomes in elderly patients [15].
Can simple ECG parameters select patients
who might benefit from ICD therapy? Bloomfield
and colleagues [16] selected patients for Multicen-
ter Automatic Defibrillator Implantation Trial-II
(MADIT-II) criteria using QRS duration O120
milliseconds. Of the 177 MADIT-II- like patients,
32% had a QRS duration O120 milliseconds, and
68% had an abnormal (positive or indeterminate)
microvolt T-wave alternans test. Twenty patients
died during an average follow-up of 20 G 6
months, Patients with an abnormal T-wave alter-
nans test were compared with those with a normal
test and patients with QRS duration O120 milli-
seconds were compared with those that had
a QRS % 120 milliseconds. The odds ratios for
2-year mortality were 4.8 (P ¼ 0.020) and 1.5
(P ¼ 0.367), respectively. The actual mortality
rate was substantially lower in patients with a nor-
mal microvolt T-wave alternans test than in patients
with a narrow QRS. The results suggested that in
MADIT-II-like patients, a microvolt T-wave alter-
nans test is better than QRS duration at identifying
a high-risk group and at identifying a low-risk
group unlikely to benefit from ICD therapy.
Left bundle branch block and right axis deviation
The presence of LBBB combined with right axis
deviation (RAD) in the frontal plane is extremely
rare, and occurs in less than 1% of patients with
LBBB. According to a study of 53 patients, (3 new
patients, and 50 reviewed from the literature),
Nicolic and colleagues [17] concluded that LBBB
and RAD is an insensitive but very specific marker
of diffuse myocardial disease. In their survey, 18 pa-
tients had cardiomyopathy (15 with primary car-
diomyopathy and another 3 had alcoholic heart
HOMBACH
disease, acromegaly, and Duchenne muscular dys-
trophy, respectively). Twelve patients had arterial
hypertension, 10 had miscellaneous disorders (con-
genital or valvular disease, cardiac surgery, myo-
carditis, or traumatic systemic AV fistula), 9
‘‘aortomyocardiosclerosis,’’ and 3 had presumptive
coronary artery disease. In 13 cases, the RAD was
observed to develop de novo or was intermittent.
No clear-cut explanation for the change of axis
was proposed in terms of clinical events or heart
rate. Childers and colleagues [18] independently re-
ported another cohort of 36 patients with LBBB
and RAD from a database of 636,000 ECGs. The
majority of patients had dilated cardiomyopathy,
and in 30 of 36 RAD was episodic. The combina-
tion of LBBB and RAD in four cases was mani-
fested as a rare form of QRS aberration with
atrial premature impulses. The combination of
LBBB and RAD is interesting but rare. It primarily
reflects widespread conduction system involvement
in diffuse myocardial disease, particularly (dilated)
biventricular cardiomyopathy.
Left bundle branch block and cardiac
resynchronization
The LV normally contracts rapidly and syn-
chronously with little variation in the onset of
electrical and mechanical activation throughout
the wall. LBBB causes asynchronous (and de-
layed) electrical and mechanical LV activation.
This produces an inefficient dyssynchronous pat-
tern of LV activation with segments contracting at
different times with resultant aggravation of the
baseline LV systolic dysfunction [19–22]. Endo-
cardial catheter mapping studies have unequivo-
cally demonstrated that RV pacing mimics the
endocardial activation patterns of LBBB [23].
Chronic LBBB produces structural changes, re-
sulting in chronic systolic and diastolic LV dys-
function that contribute to reduced LV pumping
function on a long-term basis. The changes in-
clude paradoxical septal wall motion, delayed
LV lateral wall contraction, a shorter diastolic fill-
ing time (and/or overlapping systole/diastole), de-
velopment or aggravation of functional mitral
regurgitation, reduced LV pressure, increased
left atrial diameters, and distorted LV geometry.
LV conduction delay in the form of classic
LBBB or nonspecific intraventricular conduction
delay occurs in about 20% to 30% of patients
with CHF. Biventricular pacing works by reduc-
ing the degree of electromechanical asynchrony.
 ELECTROCARDIOGRAPHY OF THE FAILING HEART
The improved sequence of electrical activation (a
process known as resynchronization) translates
into beneficial acute and long-term hemodynamic
effects by virtue of producing a more coordinated
and efficient LV contraction and reduction of
functional mitral regurgitation. The clinical value
of long-term resynchronization with biventricular
stimulation has now been established in CHF
patients with LV dyssynchrony with a better re-
sponse in idiopathic cardiomyopathy than ische-
mic disease [24–30]. Growing experience with
cardiac resynchronization therapy has highlighted
the limitations of a wide QRS complex as a surro-
gate for mechanical LV dyssynchrony. About
30% to 40% of patients with a wide QRS complex
(predominantly reflecting left-sided conduction
delay) fail to exhibit mechanical LV dyssynchrony
[31]. For this reason, echocardiographic assess-
ment with quantification of LV dyssynchrony is
emerging as a better predictor of cardiac resynch-
ronization outcome than the widened QRS
complex [32].
ST-segment and T-wave alterations in acute
coronary syndromes
ST-segment and T-wave alterations are impor-
tant regarding myocardial salvage, development of
CHF, and prognosis following an ST-elevation MI
(STEMI). In a study of 432 patients with a first MI
[33] the detection of ST-segment depression in two
or more lateral (I, aVL, V5, V6) leads revealed
higher rates of death (14.3% versus 3.6%,
P ! 0.01), more severe CHF (14.3% versus 4.1%,
P ! 0.01) and angina with ECG changes (20.0%
versus 11.6%, P ¼ 0.04) than ST-segment depres-
sion in the other ECG leads. Similar findings were
reported by Mager and colleagues [34] in 238 con-
secutive patients with acute inferior MI where
patients with ST-segment depression in the left pre-
cordial leads tended to be older and had a higher in-
cidence of hypertension, previous MI (45.8 versus
20.1%, P ¼ 0.01) and CHF (21.7 versus 3.7%,
P ¼ 0.00008). Therefore, left precordial or left lat-
eral ST-segment depression in acute MI heralds
a poor outcome with higher rates of cardiac death
and CHF. In a more extended study, Barrabes
and colleagues [35] analyzed the initial ECG in
775 consecutive patients with a first MI without
ST-segment elevation in leads other than aVR or
V1 (non-STEMI). The rates of in-hospital death
in patients without (n ¼ 525) and with 0.05 to
0.1 mV (n ¼ 116) or R0.1 mV (n ¼ 134) of ST-
segment elevation in a lead VR were 1.3%, 8.6%,
417
and 19.4%, respectively (P ! 0.001). After adjust-
ment for baseline clinical predictors and for ST-
segment depression on admission, the odds ratios
for death in the last two groups (those with ST-seg-
ment elevation) were 4.2 (95% CI, 1.5–12.2) and 6.6
(95% CI, 2.5–17.6), respectively. The rates of recur-
rent ischemic events and CHF during hospital stay
also increased in a stepwise fashion among the
groups according to the degree of ST-segment ab-
normality. The results indicated that ST-segment
elevation in lead aVR provides important short-
term prognostic information in patients with a first
non-STEMI.
The ST-segment resolution pattern can also be
used to predict CHF, as was reported in 258
patients from the TIMI-7 (Thrombolysis In Myo-
cardial Infarction) and GUSTO-I (Global Utiliza-
tion of Strategies To Open Occluded Coronary
Arteries) Trials: by logistic regression analysis,
ST-segment resolution patterns were an indepen-
dent predictor of the combined outcome of mor-
tality or CHF (P ¼ 0.024), whereas TIMI-flow
grade assessment was not (P ¼ 0.693). A sub-
group of patients with TIMI-flow 0 to 1 after
thrombolysis was identified by rapid ST-segment
resolution of R50% with relatively benign out-
comes [36]. In addition, data from the GUSTO-I
study [37] revealed that patients with higher
T-waves (O98% percentile of the upper limit of
normal) had a lower 30-day mortality than those
without high T waves (5.2% versus 8.6%, P ¼
0.001) and were less likely to develop CHF
(15% versus 24%, P ! 0.001) or cardiogenic
shock (6.1% versus 8.6%, P ¼ 0.0023). The study
suggested that T-wave amplitude had prognostic
significance after controlling the data for time to
treatment in acute MI.
Table 1 provides a summary of the most signif-
icant ECG changes in CHF and their implications
for clinical diagnosis of the patients.
Cardiac repolarization in heart failure
Cardiac repolarization may be disturbed in
patients with CHF, and in principle, these alter-
ations may be measured by QT- or QTc-duration,
QT variability, QT-dispersion, and the detection
of microvolt T-wave alternans. As CHF patients
might have areas of fibrosis and scars the evalu-
ation of changes in cardiac repolarization might
be diagnostically and prognostically useful
with respect to life-threatening ventricular
tachyarrhythmias.
 418
Table 1
Outline of significant parameters of indices of increased dispersion of ventricular repolarization and disturbed sympatho-vagal balance and their clinical significance for
assessing patients with congestive heart failure
Method/Index Definition Pathophysiology Advantage Disadvantage Clinical use

HOMBACH
QT variability Beat-to-beat or Increased dispersion of Simple measured Computer algorithm Experimental
(QTV) diurnal variations repolarization Parameter (QT duration) not generally available
of QT duration (susceptibility to Criteria not standardized
reentrant ventricular
tachycardias)

QT dispersion Difference between Increased dispersion of Simple measured parameter Too many technical Obsolete
(QTd) longest and shortest repolarization problems and pitfalls
QT interval in (susceptibility to Variable and
12-lead or 3-lead reentrant ventricular Inconsistent results
(Holter) ECG tachycardias)
T wave alternans Macro-TWA: Increased dispersion of Macro: direct (eyball) Micro: complicated Experimental (near
(TWA) beat-to-beat repolarization detection from protocol clinical)
changes of T- wave (susceptibility to conventional ECG Algorithm not
polarity and/or reentrant ventricular Strong parameter of generally available
shape tachycardias) ventricular vulnerability Ventricular premature beats
Micro-TWA: beat-to- may interfere with
beat amplitude computer analysis
fluctuations of the Increased heart rate by
T wave at the exercise or emotion
microvolt level necessary for analysis
Heart rate Beat-to-beat oscillations Disturbed sympathovagal Computer algorithms Many overlapping Clinically useful
variability of relative risk-intervals balance (susceptibility widely available parameters (time/
(HRV) to life-threatening Relative simple frequency domain)
ventricular automated measurement Not valid at periods of
tachyarrhythmias) Well standardized technique physical activity, postural
and parameters changes or emotional
stimuli
Low sensitivity and positive
predictive value (PPV)
Heart rate Fluctuations of sinus Disturbed sympathovagal Simple parameter Computer algorithm not Experimental (near
turbulence rhythm cycles after a balance (susceptibility of BRS generally available clinical)
(HRT) single VPB to life-threatening Standardized Requires single
ventricular (automated) ventricular

ELECTROCARDIOGRAPHY OF THE FAILING HEART

tachyarrhythmias) measurement premature beats
Not applicable in AF
Low sensitivity and positive
predictive value (PPV)
Baroreflex Reflex heart rate response Disturbed sympathovagal Clear pathophysiologic Advanced and complicated Experimental
sensitivity (BRS) to physiologic activation balance (susceptibility concept technique
and deactivation of to life-threatening Well standardized technique Low sensitivity and positive
baroreceptors following ventricular predictive value (PPV)
IVdphenylephrine tachyarrhythmias)

419
420 HOMBACH

Ancillary parameters received an LV assist device (LVAD) for CHF.

Table 2 provides an overview of the most sig-
nificant tests and indexes related to disturbed
cardiac repolarization and sympathovagal bal-
ance, their pathophysiologic role in the initiation
of malignant ventricular tachyarrhythmias, and
clinical significance for investigating CHF
patients.
QT interval duration
In the study of Harding and colleagues [38]
significant changes of QRS- and both QT- and
QTc-duration were observed in 23 patients who
QRS duration decreased from 117 G 6 to 103 G
6 milliseconds (P ! 0.01), the absolute QT-dura-
tion increased from 359 G 6 to 378 G 8 millisec-
onds (P ! 0.05), and the QTc-interval increased
from 379 G 10 to 504 G 11 milliseconds (P ! 0.01).
None of these immediate changes were observed
among 22 patients undergoing routine coronary ar-
tery bypass grafting. The findings that QRS- and
QT-duration were significantly influenced by the
load on the failing heart and that the action poten-
tial shortened, may contribute to the improved
cellular contractile performance after sustained
LVAD support. Similar findings were noted by
Gaudron and colleagues [39] in 134 patients after

Table 2
Summary of the most significant ECG alterations and their relation to pathophysiologic and clinical factors in congestive
heart failure
Observation Implication
Normal  Left ventricular systolic dysfunction very unlikely
 Negative predictive value for CHF O90%
ST-T wave changes  Myocardial ischemia
 Arrhythmogenic right ventricular cardiomyopathy
 Hypertrophic cardiomyopathy
Left ventricular hypertrophy  Arterial hypertension
 Aortic stenosis
 Hypertrophic cardiomyopathy
 Early (subclinical) stages of dilated cardiomyopathy
Left bundle branch block  Usually associated with structural heart disease
(LBBB)  Indicates diffuse myocardial disease and fibrosis
 Prevalence: about 25% in CHF
 5-year mortality about 35%
 Only weak correlation between QRS duration and left
ventricular mechanical dyssynchrony
LBBB and right axis deviation  Prevalence: extremely rare (!1% of LBBB)
 Relatively specific for diffuse disease like dilated congestive
cardiomyopathy (DCM)
LBBB with left axis deviation  Most common in diffuse structural heart disease
Right bundle branch block  No clear association with heart failure except for cor pulmonale
(RBBB)  Prevalence only about 6% in CHF
Small QRS complexes  Pericardial effusion
 Hypothyroidism
 Amyloidosis
Is there a typical ECG pattern  left ventricular hypertrophy in percordial leads with small voltage
of CHF? in frontal leads (HT, AS)
 left bundle branch block with left axis deviation atrial fibrillation
(DCM, CHD)
 LBBB with right axis deviation (DCM)
What ECG markers best  Q waves in precordial leads or LBBB in patients with known CHD
correlate with low LVEF?  Left atrial ‘‘enlargement’’ (P-wave) and/or LVH in (decompensated)
diseases with pressure overload
Abbreviations: As, aortic stenosis; CHD, coronary heart cartery disease; CHF, congestive heart failure; DCM, dilated
cardiomyopathy; HT, systemic (arterial) hypertension; LVH, left ventricular hypertrophy; LBBB, left bundle branch
block.
 ELECTROCARDIOGRAPHY OF THE FAILING HEART
acute MI, in whom LV dilatation was closely re-
lated to the end-diastolic LV volume index,
Lown-ventricular premature beat (VPB) score
(r ¼ 0.98, P ! 0.01), and the QTc-prolongation
(r ¼ 0.998, P ! 0.01), respectively. Patients with
progressive LV remodeling reflected by LV dilata-
tion on echocardiography, and QTc prolongation
may be more susceptible to enhanced electrical in-
stability. LV remodeling after MI might serve as
a link between dysfunction, electrical instability,
and sudden cardiac death.
In a recent study Vrtovec and colleagues [40]
measured corrected QT interval using Bazett for-
mula from standard 12-lead ECGs in 241 patients
with CHF and elevated BNP levels O400 pg/mL.
QTc intervals were prolonged in 122 (52%) and
normal in 119 (49%) patients, but BNP levels
were comparable in both groups. After 6-month
follow-up 46 patients died, 9 underwent trans-
plantation, and another 17 underwent LV assist
device implantation. The deaths were attributed
to pump failure (n ¼ 24), sudden cardiac death
(n ¼ 18), and noncardiac causes (n ¼ 4). Ka-
plan-Meier survival rates were three times higher
in the normal QTc group than in the prolonged
QTc group, and on multivariate analysis pro-
longed QTc interval was an independent predictor
of all-cause death (P ¼ 0.0001), cardiac death
(P ¼ 0.0006), sudden cardiac death (P ¼ 0.004),
and pump failure death (P ¼ 0.0006). A pro-
longed QTc interval is thus a strong, independent
predictor of adverse outcome in CHF patients
with BNP levels O400 pg/mL.
As a more recent parameter of increased
dispersion of repolarization, QT variability was
studied by Berger and colleagues [41]. In 83 pa-
tients with ischemic and nonischemic dilated car-
diomyopathy and 60 control subjects QT
variability index was calculated in each subject
as the logarithm of the ratio of normalized QT
variance to heart rate variance, and the coherence
between heart rate and QT interval fluctuations
was determined by spectral analysis. Patients
with dilated cardiomyopathy had greater QT var-
iance than control subjects (60.4 G 63.1 versus
25.7 G 24.8 ms2, P ! 0.0001) despite reduced heart
rate variance (6.7 G 7.8 versus 10.5 G 10.4 bpm2,
P ¼ 0.01). The QT variability index was higher in
patients with dilated cardiomyopathy than in con-
trol subjects with a high degree of significance,
and QT variability index did not correlate with
LVEF but did depend on NYHY functional class.
QT variability index did not differ between pa-
tients with ischemic and those with nonischemic
421
origin. Coherence between heart rate and QT inter-
val fluctuations at physiologic frequencies was
lower in patients with dilated cardiomyopathy
compared with those without dilated cardiomyop-
athy and QT interval variability increased with
worsening of functional class but was independent
of LVEF, indicating that dilated cardiomyopathy
leads to temporal lability in ventricular
repolarization.
QT dispersion
QT dispersion (QTd) is simply measured as the
difference between the longest and shortest QT
interval within the normal 12-lead ECG. QTd is
often increased in patients with idiopathic dilated
cardiomyopathy and CHF, but it does not neces-
sarily predict an increased risk of arrhythmic
death [42,43]. Pinsky and colleagues [44] studied
108 consecutive patients referred for heart trans-
plantation, 80 of whom were placed on the wait-
ing list. During 25 months of follow-up
comparison of QTd R140 milliseconds versus
%140 milliseconds, and a QTc of R9% of normal
predicted a 4.1 increase of the risk of cardiac
death. In a recent study of Brooksby and col-
leagues [45] in 554 ambulatory outpatients with
CHF, the heart rate corrected QT-dispersion
and maximum QT-interval were significant uni-
variate predictors of all-cause mortality during
a follow-up period of 471 G 168 days (P ¼
0.026, and !0.0001, respectively), and of sudden
death and progressive heart failure death. These
indices were not related to outcome in the multi-
variate analysis, in which independent predictors
of all-cause mortality consisted of cardiothoracic
ratio (P ¼ 0.0003), echocardiographically derived
end-diastolic dimension (P ¼ 0.007), and ven-
tricular couplets on 24-hour ECG monitoring
(P ¼ 0.015).
These results could not be reproduced in the
Evaluation of Losartan In the Elderly trial (Los-
artan Heart Failure Survival Study) [46], in which
986 high-uality ECGs were retained and analyzed
for heart rate, QRS duration, maximum QT and
JT intervals,and QT and JT dispersion from
a pool of 3118 ECGs of 1804 patients. During
a follow-up of 540 G 153 days there were 140
(14%) deaths from all causes, including 119 car-
diac (12%) and 59 (6%) sudden cardiac deaths.
The mean heart rate was significantly faster in
nonsurvivors than in survivors (77 G 16 versus
74 G 14 bpm, P ¼ 0.0006) and in patients who
died of cardiac causes (76 G 16 bpm, P ¼ 0.04
422
versus survivors). Mean QRS duration was signif-
icantly longer in nonsurvivors (107 G 25 millisec-
onds), and in the subgroups who died of cardiac
(107 G 24 milliseconds) or sudden death (112 G
23 milliseconds) than in survivors (99 G 24 milli-
seconds, P ! 0.01 for all). The maximum QTc and
QT intervals were similar for nonsurvivors, re-
gardless of cause of death and in survivors. There
were no significant differences in QTd or JTd (JT
dispersion from the end of the QRS to the end of
the T wave) between patients with any mode of
death and survivors (P ! 0.1 for all). Neither los-
artan nor captopril significantly modified QTd or
JTd. Therefore, the investigators concluded that
increased QTd is not associated with increased
mortality in CHF patients, and it is inappropriate
to examine drug efficacy in such patients [47]. Al-
though QTd might be easily measured from the
12-lead ECG, this parameter is virtually of no di-
agnostic and prognostic value in CHF patients.
ECG-derived parameters for assessing sympa-
thetic nervous system activity–heart rate variability
include: (1) time domain parameters: root-mean-
square of successive differences, percent differ-
ence between adjacent normal to normal relative
risk (RR) intervals O50 milliseconds computed
over the entire 24-hour ECG recording (pNN50),
percent difference between adjacent normal RR
intervals O50 milliseconds computed over the
entire 24-hour ECG recording (pNN50) for short
term; standard deviation of the averages normal to
normal RR intervals in all 5-minute segments of
a 24-hour ECG recording, standard deviation of
all normal to normal RR intervals (SDNN), and
mean of standard deviation for all 5-minute
segments of RR intervals in 24 hours for long-
term heart rate variability. (2) Frequency domain
parameters of heart rate variability: ultralow-
frequency power, very low-frequency power,
low-frequency power, high-frequency power, bar-
oreflex sensitivity, and heart rate turbulence.
Activation of the sympathetic (adrenergic)
nervous system is one of the hallmarks of chronic
CHF. Adrenergic activation may be estimated by
measuring norepinephrine (NE) concentration in
the blood, determining heart rate variability, that
is, RR interval fluctuations in short- and long-
term ECG recordings, by studying baroreflex
sensitivity (the control of vagal and sympathetic
outflow to the heart and peripheral vessels), and
by quantifying heart rate turbulence.
Heart rate variability may be measured in
short 1- to 5- to 10-minute ECG strips or 24-
hour long-term ECG recordings, both in time and
HOMBACH
the frequency domain. Baroreflex sensitivity as-
sesses the reflex heart rate response to physiologic
activation and deactivation of the baroreceptors
secondary to drug-induced changes in arterial
pressure. These are measured by correlating
spontaneous fluctuations of systolic blood
pressure with the corresponding heart rate (RR
interval) both at rest and after administration
of the pure alpha-adrenoreceptor stimulant
phenylephrine.
The method of heart rate turbulence quantifies
the fluctuations of sinus rhythm cycles after single
ventricular premature complexes (VPCs). In nor-
mal or low-risk groups, sinus rhythm exhibits
a characteristic biphasic pattern of early acceler-
ation and subsequent deceleration after a VPC.
However, in high-risk (mostly postinfarction)
patients, hear rate turbulence is attenuated or
missing.
Heart rate variability
Increased sympathetic tone in cardiovascular
conditions was elegantly demonstrated by Grassi
and colleagues [47] in 243 subjects: 38 normoten-
sive healthy controls, 113 with untreated essential
hypertension, 27 normotensive and obese, and 65
with CHF. Heart rate was correlated with venous
plasma NE, and postganglionic muscle sympa-
thetic nerve activity (microneurography at a pero-
neal nerve). In the whole group, the heart rate was
correlated with both plasma NE (r ¼ 0.32,
P ! 0.0001) and muscle sympathetic nerve activ-
ity (r ¼ 0.38, P ! 0.0001). Heart rate values were
greater in the obese and CHF patients than in
controls, as was plasma NE and muscle sympa-
thetic nervous activity (all differences statistically
significant). In the subjects with essential hyper-
tension, no significant relationship was found be-
tween these three indices of sympathetic activity.
The results suggested that supine heart rate can
be regarded as a marker of intersubject differences
in sympathetic tone in the general population and
individuals with cardiovascular disease.
In patients with CHF, heart rate variability
(HRV) is frequently decreased, with a significant
increase in the low-frequency component of the
frequency domain HRV measurements, together
with high levels of catecholamines, BNP, and an
altered baroreflex sensitivity. The prognostic sig-
nificance of HRV in CHF patients remains un-
clear. In one of the first studies in 40 patients with
CHF, comparing normal individuals and patients
with a history of nonsustained ventricular
 ELECTROCARDIOGRAPHY OF THE FAILING HEART
tachycardia and normal cardiac function, there
was no significant difference in the HRV param-
eters despite a significantly lowered HRV in CHF
patients [48]. In a prospective study of 71 patients
with idiopathic dilated cardiomyopathy and
CHF, Hoffman and colleagues [49] studied HRV
by time and frequency domain methods. After
a follow-up period of 15 G 5 months, there was
no significant difference of time or frequency do-
main indices of HRV in patients with arrhythmic
events compared with those without major ar-
rhythmic events. In a further study of 159 patients
with idiopathic dilated cardiomyopathy and
CHF, 30 patients died during follow-up, and there
was a significant correlation between LVEF, in-
creased SDNN and the percentage of adjacent
RR intervals that vary by more than 50 millisec-
onds (pNN50) and an increased risk of cardiac
death. However, the risk of sudden cardiac death
correlated only with LVEF. This was in contrast
to an increased low-frequency power and impair-
ment of pNN50 that strongly correlated with an
increased risk of death from progressive pump
failure [50]. Another two studies showed that
HRV was reduced in patients with CHF with
a strong correlation between HRV and LV func-
tion and peak oxygen consumption [51,52].
HRV may have prognostic impact in hospital-
ized patients for decompensated CHF, as shown
by Aronson and colleagues [53] in 199 patients
with NYHA class III or IV CHF using 24-hour
Holter recordings, and time and frequency HRV
parameters. During a mean follow-up of 312 G
150 days after discharge from the hospital, 40 pa-
tients (21.1%) died. Kaplan-Meier analysis indi-
cated that patients with abnormal values of
standard deviation of the RR intervals over
a 24-hour period (P ¼ 0.027), of SD of all 5-minute
mean RR intervals (P ¼ 0.043), of total power
(P ¼ 0.022), and of ultralow-frequency power
(P ¼ 0.008) in the lower tertile were at higher risk
of death. In a multivariate Cox regression model,
the same indexes in the lower tertile were indepen-
dent predictors of mortality (RR from 2.2 to 2.6).
The severity of autonomic dysfunction during hos-
pitalization for CHF can thus predict survival after
hospital discharge using measurements of overall
HRV.
The predictive power of HRV for cardiac as
well as sudden cardiac death was shown in three
recent studies. In the VA trial [54] of 179 patients
with CHF, the lowest quartile of patients were
compared with the remaining, using SDNN as
the sole HRV parameter. Among 127 patients
423
meeting the inclusion criteria, SDNN !65.3 milli-
seconds (the lowest quartile) was the sole indepen-
dent factor predictive of survival in a multivariate
model (P ¼ 0.0001). A Cox proportional-hazards
model revealed that each increase in 10 millisec-
onds in SDNN conferred a 20% decrease in risk
of mortality (P ¼ 0.0001). Furthermore, patients
with SDNN !65.3 milliseconds had a significantly
increased risk of sudden death (P ¼ 0.016). Thus,
in this study HRV was the sole independent pre-
dictor of overall mortality, and was significantly
associated with sudden death. La-Rovere and col-
leagues [55] developed a multivariate survival
model for the identification of sudden (presum-
ably arrhythmic) death from data of 202 con-
secutive patients with moderate to severe CHF.
Time- and frequency-domain HRV parameters ob-
tained from 8-minute ECG recordings at baseline
and during controlled breathing were challenged
against clinical and functional parameters. This
model was then validated in 242 consecutive pa-
tients referred for CHF as the validation sample.
Sudden death was independently predicted by
a model that included low frequency power of
HRV during controlled breathing %13 millisec-
onds2 and left ventricular end-diastolic diameter
R77 mm (RR 3.7, 95% CI, 1.5 to 9.3, and RR
2.6, 95% CI, 1.1 to 6.3, respectively). Low-
frequency power %11 milliseconds2 during con-
trolled breathing and R83 VPCs/hr on Holter
monitoring were both independent predictors of
sudden death (RR 3.0, 95% CI, 1.2 to 7.6,
and RR 3.7, 95% CI, 1.5 to 9.0, respectively).
The results showed that reduced short-term low-
frequency power during controlled breathing is
a powerful and independent predictor of sudden
death in CHF patients. Similar results by the
same group were reported by Guzzetti and col-
leagues [56], who studied 330 consecutive CHF pa-
tients. Using time, frequency, and fractal analyses,
the risk of pump failure and of sudden death could
be differentiated as follows: depressed power of
night time HRV (%509 milliseconds2) very low-
frequency power (!0.04 Hz), high pulmonary
wedge pressure (R18 mmHg), and low LVEF
(%24%) were independently related to pump fail-
ure, while reduction of low frequency power
(%20 milliseconds2) and increased LV end-systolic
diameter (R61 mm) were linked to sudden (ar-
rhythmic) mortality.
From a technical point of view it is interesting
that HVR parameters may also be retrieved from
implanted devices. In this respect Adamson and
colleagues [57] studied 397 patients where
424 HOMBACH
continuous HRV was measured as the standard
deviation of 5-minute median atrial–atrial inter-
vals (SDAAM) sensed by the biventricular pacing
device. SDAAM !50 milliseconds when averaged
over 4 weeks was associated with increased mor-
tality risk (HR 3.2, P ¼ 0.02), and SDAAM
were persistently lower over the entire follow-up
period in patients who required hospitalization
or died. Automated detection of decreases in
SDAAM was 70% sensitive in detecting cardio-
vascular hospitalisations, with 2.4 false-positives
per patient-year follow-up.
Last, it should be mentioned that HRV may be
improved by exercise training [58], by drugs like
candesartan [59] or valsartan, whose effect was
comparable with lisinopril [60], and by biventricu-
lar pacing for cardiac resynchronization [61].
Whether these effects influence the prognosis of
CHFpatients remains to be investigated in larger
studies. In conclusion, based on present experi-
ence assessing HRV either by time or frequency
domain parameters may be a valuable tool for
predicting an adverse prognosis in CHF patients
irrespective of the underlying heart disease.
References
[1] Lonn E, Mathew J, Pogue J, et al, and the Heart
Outcomes Prevention Evaluation Study Investiga-
tors. Relationship of electrocardiographic left
ventricular hypertrophy to mortality and cardiovas-
cular morbidity in high-risk patients. Eur J Cardio-
vasc Prev Rehab 2003;10:420–8.
[2] Fonseca C, Oliveira AG, Mota T, et al. The value of
the electrocardiogram and chest X-ray for confirm-
ing or refuting a suspected diagnosis of heart failure
in the community. Eur J Heart Fail 2004;6:807–12.
[3] Sanders P, Morton JB, Davidson NC, et al. Electri-
cal remodeling of the atria in congestive heart
failure: electrophysiological and electroanatomical
mapping in humans. Circulation 2003;108:1461–8.
[4] Song J, Kalus JS, Caron MF, et al. Effect of diuresis
on P-wave duration and dispersion. Pharmaco-
therapy 2002;22:564–8.
[5] Camsari A, Pekdemir H, Akkus MN, et al. Long-
term effects of beta blocker therapy on P-wave dura-
tion and dispersion in congestive heart failure
patients: a new effect? J Electrocardiol 2003;36:
111–6.
[6] Faggiano P, Dáloia A, Zanelli E, et al. Contribu-
tion of left atrial pressure and dimension to signal-
averaged P-wave duration in patients with chronic
congestive heart failure. Am J Cardiol 1997;179:
219–22.
[7] Yamada T, Fukunami M, Shimonagata T, et al. Pre-
diction of paroxysmal atrial fibrillation in patients
with congestive heart failure: a prospective study.
J Am Coll Cardiol 2000;35:405–13.
[8] Dixen U, Wallevik L, Hansen MS, et al. Prolonged
signal-averaged P wave duration as a prognostic
marker for morbidity and mortality in patients
with congestive heart failure. Scan Cardiovasc
J 2003;37:193–8.
[9] Shamim W, Yousufuddin M, Xiao HB, et al. Septal
q waves as an indicator of risk of mortality in elderly
patients with chronic heart failure. Am Heart J 2002;
44:740–4.
[10] Krüger S, Filzmaier K, Graf J, et al. QRS prolonga-
tion on surface ECG and brain natriuretic peptide as
indicators of left ventricular systolic dysfunction.
J Intern Med 2004;255:206–12.
[11] Bode-Schnurbus L, Bocker D, Block M, et al. QRS
duration: a simple marker for predicting cardiac
mortality in ICD patients with heart failure. Heart
2003;89:1157–62.
[12] Iuliano S, Fisher SG, Karasik PE, et al. Department
of Veterans Affairs Survival Trial of Antiarrhythmic
Therapy in Congestive Heart Failure. QRS duration
and mortality in patients with congestive heart
failure. Am Heart J 2002;143:1085–91.
[13] Morgera T, Di Lenarda A, Sabbadini G, et al.
Idiopathic dilated cardiomyopathy: prognostic sig-
nificance of electrocardiographic and electro-
physiological findings in the nineties. Ital Heart
J 2004;5:593–603.
[14] Shenkman HJ, Pampati V, Khandelwal AK, et al.
Congestive heart failure and QRS duration: estab-
lishing prognosis study. Chest 2002;122:528–34.
[15] Shamim W, Yousufuddin M, Cicoria M, et al. Incre-
mental changes in QRS duration in serial ECGs over
time identify high risk elderly patients with heart
failure. Heart 2002;88:47–51.
[16] Bloomfield DM, Steinman RC, Namerow PB, et al.
Microvolt T-wave alternans distinguishes between
patients likely and patients not likely to benefit
from implanted cardiac defibrillator therapy: a solu-
tion to the Multicenter Automatic Defibrillator
Implantation Trial (MADIT) II conundrum. Circu-
lation 2004;110:1885–9.
[17] Nicolic G, Marriott HJL. Left bundle branch
block with right axis deviation: a marker of conges-
tive cardiomyopathy. J Electrocardiol 1985;18:
395–404.
[18] Childers R, Lupovich S, Sochanski M, et al. Left
bundle branch block and right axis deviation: a re-
port of 36 cases. J Electrocardiol 2000;33:93–102.
[19] Leclercq C, Kass DA. Retiming the failing heart: prin-
ciples and current clinical status of cardiac resynchro-
nization. J Am Coll Cardiol 2002;39:194–201.
[20] Leclercq C, Kass DA. Retiming the failing heart:
principles and current clinical status of cardiac
resynchronization. J Am Coll Cardiol 2002;39:
194–201.
[21] Kass DA. Pathology of cardiac dyssynchrony and
resynchronization. In: Ellenbogen KA, Kay GN,
 ELECTROCARDIOGRAPHY OF THE FAILING HEART
Wilkoff BL, editors. Device therapy for congestive
heart failure. Philadelphia (PA): WB Saunders;
2004. p. 27–46.
[22] Tedrow U, Sweeney MO, Stevenson WG. Physiol-
ogy of cardiac resynchronization. Curr Cardiol
Rep 2004;6:189–93.
[23] Vassallo JA, Cassidy DM, Miller JM, et al. Left ven-
tricular endocardial activation during right ventricu-
lar pacing: effect of underlying heart disease. J Am
Coll Cardiol 1986;7:1228–33.
[24] Cazeau S, Leclercq C, Lavergne T, et al, and the
Multisite Stimulation in Cardiomyopathies (MUS-
TIC) Study Investigators. Effects of multisite biven-
tricular pacing in patients with heart failure and
intraventricular conduction delay. N Engl J Med
2001;344:873–80.
[25] Abraham WT, Fisher WG, Smith AL, et al, and the
MIRACLE Study Group. Multicenter InSync Ran-
domized Clinical Evaluation. Cardiac resynchroni-
zation in chronic heart failure. N Engl J Med 2002;
346:1845–53.
[26] Bristow MR, Saxon LA, Boehmer J, et al, and the
Comparison of Medical Therapy, Pacing, and Defi-
brillation in Heart Failure (COMPANION) Investi-
gators. Cardiac-resynchronization therapy with or
without an implantable defibrillator in advanced
chronic heart failure. N Engl J Med 2004;350:
2140–50.
[27] Cleland JG, Daubert JC, Erdmann E, et al, and
the Cardiac Resynchronization-Heart Failure
(CARE-HF) Study Investigators. The effect of
cardiac resynchronization on morbidity and mor-
tality in Heart failure. N Engl J Med 2005;352:
1539–49.
[28] Strickberger SA, Conti J, Daoud EG, et al. Patient
selection for cardiac resynchronization therapy:
from the Council on Clinical Cardiology Subcom-
mittee on Electrocardiography and Arrhythmias
and the Quality of Care and Outcomes Research In-
terdisciplinary Working Group, in collaboration
with the Heart Rhythm Society. Circulation 2005;
111:2146–50.
[29] Sutton MG, Plappert T, Hilpisch KE, et al. Sus-
tained reverse left ventricular structural remodeling
with cardiac resynchronization at one year is a func-
tion of etiology: quantitative Doppler echocardio-
graphic evidence from the Multicenter InSync
Randomized Clinical Evaluation (MIRACLE). Cir-
culation 2006;113:266–72.
[30] Knight BP, Desai A, Coman J, et al. Long-term re-
tention of cardiac resynchronization therapy. J Am
Coll Cardiol 2004;44:72–7.
[31] Kashani A, Barold SS. Significance of QRS complex
duration in patients with heart failure. J Am Coll
Cardiol 2005;46:2183–92.
[32] Yu CM, Wing-Hong Fung J, Zhang Q, et al. Under-
standing nonresponders of cardiac resynchroniza-
tion therapydcurrent and future perspectives.
J Cardiovasc Electrophysiol 2005;16:1117–24.
425
[33] Barrabes JA, Figuears J, Moure C, et al. Q-wave
evolution of a first acute myocardial infarction with-
out significant ST segment elevation. Int J Cardiol
2001;77:55–62.
[34] Mager A, Sclarovky S, Herz I, et al. Value of the ini-
tial electrocardiogram in patients with inferior-wall
acute myocardial infarction for prediction of multi-
vessel coronary artery disease. Coron Artery Dis
2000;11:415–20.
[35] Barrabes JA, Figueras J, Moure C, et al. Prognostic
value of lead aVR in patients with a first non-ST-
segment elevation acute myocardial infarction.
Circulation 2003;108:814–9.
[36] Shah A, Wagner GS, Granger CB, et al. Prognostic
implications of TIMI flow grade in the infarct re-
lated artery compared with continuous 12-lead ST-
segment resolution analysis: reexamining the ‘‘gold
standard’’ for myocardial reperfusion assessment.
J Am Coll Cardiol 2000;35:666–72.
[37] Hochrein J, Sun F, Pieper KS, et al. Higher T-wave
amplitude associated with better prognosis in pa-
tients receiving thrombolytic therapy for acute myo-
cardial infarction (a GUSTO-I substudy). Global
Utilization of Streptokinase and Tissue plasminogen
Activator for Occluded Coronary Arteries. Am J
Cardiol 1998;81:1078–84.
[38] Harding JD, Piacentino V, Gaughan JP, et al. Elec-
trophysiological alterations after mechanical circu-
latory support in patients with advanced cardiac
failure. Circulation 2001;104:1241–7.
[39] Gaudron P, Kugler I, Hu K, et al. Time course of
cardiac structural, functional and ellectrical changes
in asymptomatic patients after myocardial infarc-
tion: their inter-relation and prognostic impact.
J Am Coll Cardiol 2001;38:33–40.
[40] Vrtovec B, Delgado R, Zewail A, et al. Prolonged
QTc interval and high B-type natriuretic peptide
levels together predict mortality in patients with
advanced heart failure. Circulation 2003;107:
1764–9.
[41] Berger RD, Kasper EK, Baughman KL, et al. Beat-
to-beat QT interval variability: novel evidence for
repolraization lability in ischemic and nonischemic
dilated cardiomyopathy. Circulation 1997;96:
1557–65.
[42] Grimm W, Steder U, Menz V, et al. QT dispersion
and arrhythmic events in idiopathic dilated cardio-
myopathy. Am J Cardiol 1996;78:458–61.
[43] Bodi-Peris V, Monmeneu-Menadas JV, Marin-
Ortuno F, et al. QT interval dispersion in hospital
patients admitted with cardiac insufficiency. Deter-
minants and prognostic value. Rev Esp Cardiol
1999;52:563–9.
[44] Pinsky DJ, Sciacca RR, Steinberg JS. QT dispersion
as a marker of risk in patients awaiting heart trans-
plantation. J Am Coll Cardiol 1997;29:1576–84.
[45] Brooksby P, Batin PD, Nolan J, et al. The relation-
ship between QT intervals and mortality in ambulant
patients with chronic heart failure. The United
426
Kingdom heart failure evaluation and assessment of
risk trial (UK-HEART). Eur Heart J 1999;20:
1335–41.
[46] Gang Y, Ono T, Hnatkova K, et al. ELITE II inves-
tigators. QT dispersion has no prognostic value in
patients with symptomatic heart failure: an ELITE
II substudy. Pacing Clin Electrophysiol 2003;26:
394–400.
[47] Grassi G, Vailati S, Bertinieri G, et al. Heart rate as
a marker of sympathetic activity. J Hypertens 1998;
16:1635–9.
[48] Fei L, Keeling PJ, Gill JS, et al. Heart rate variability
and it’s relation to ventricular arrhythmias in con-
gestive heart failure. Br Heart J 1994;71:322–8.
[49] Hoffman J, Grimm W, Menz V, et al. Heart rate var-
iability and major arrhythmias in patients with
idiopathic dilated cardiomyopathy. Pacing Clin
Electrophysiol 1996;19:1841–4.
[50] Szabo BM, van Veldhuisen DJ, van der Veer N, et al.
Prognostic value of HRV in chronic congestive heart
failure secondary to idiopathic or ischemic dilated
cardiomyopathy. Am J Cardiol 1997;79:978–80.
[51] Yi G, Goldman JH, Keeling PJ, et al. Heart rate var-
iability in idiopathic dilated cardiomyopathy: rela-
tion to disease severity and prognosis. Heart 1977;
77:108–14.
[52] Fauchier L, Babuty D, Cosnay P, et al. Heart rate
variabilty in idiopathic dilated cardiomyopathy:
characteristics and prognostic value. J Am Coll
Cardiol 1997;30:1009–14.
[53] Aronson D, Mittleman MA, Burger AJ. Measures of
heart period variability as predictors of mortality in
hospitalized patients with decompensated conges-
tive heart failure. Am J Cardiol 2004;93:59–63.
[54] Bilchick KC, Fetics B, Djoukeng R, et al. Prognostic
value of heart rate variability in chronic congestive
heart failure (Veterans Affaiŕs Survival Trial of
HOMBACH
Antiarrhythmic Therapy in Congestive Heart Fail-
ure). Am J Cardiol 2002;90:24–8.
[55] La Rovere MT, Pinna GD, Maestri R, et al. Short-
term heart rate variability strongly predicts sudden
cardiac death in chronic heart failure patients. Circu-
lation 2003;107:565–70.
[56] Guzzetti S, La Rovere MT, Pinna GD, et al. Differ-
ent spectral components of 24 h heart rate variability
are related to different modes of death in chronic
heart failure. Eur Heart J 2005;26:357–62.
[57] Adamson PB, Smith AL, Abraham WT, et al, and
the InSync III Model 8042 and Attain OTW Lead
Model 4193 Clinical Trial Investigators. Continu-
ous autonomic assessment in patients with symp-
tomatic heart failure: prognostic value of heart
rate variability measured by an implanted cardiac
resynchronization device. Circulation 2004;110:
2389–94.
[58] Larsen AI, Gjesdal K, Hall C, et al. Effect of exer-
cise training in patients with heart failure: a pilot
study on autonomic balance assessed by heart
rate variability. Eur J Cardiovasc Prev Rehab
2004;11:162–7.
[59] Tambara K, Fujita M, Sumita Y, et al. Beneficial ef-
fect of candesartan treatment on cardiac autonomic
nervous system activityin patients with chronic heart
failure: simultaneous recording of ambulatori elec-
trocardiogram and posture. Clin Cardiol 2004;27:
300–3.
[60] De Tommasi E, Iacoviello M, Romito R, et al. Com-
parison of the effect of valsartan and lisinopril on au-
tonomic nervous system activity in chronic heart
failure. Am Heart J 2003;146:E17.
[61] Livanis EG, Flevari P, Theodorakis GN, et al. Effect
of biventricular pacing on heart rate variability in
patients with chronic heart failure. Eur J Heart
Fail 2003;5:175–8.
 Cardiol Clin 24 (2006) 427–437
The 12-Lead Electrocardiogram
in Supraventricular Tachycardia
Uday N. Kumar, MD, Rajni K. Rao, MD,
Melvin M. Scheinman, MD*
Division of Cardiology, Department of Medicine, 500 Parnassus Avenue, Box 1354, University of California,
San Francisco, San Francisco, California 94143, USA
The term supraventricular tachycardia (SVT)
encompasses a range of common arrhythmias in
which the atrial or atrioventricular (AV) node is
essential for the perpetuation of the tachyarrhyth-
mia [1]. Because of misdiagnosis and inconsistent
classification, the exact prevalence of SVT is not
clear but may be as high as six to eight per 1000
people in the United States [2]. SVT affects pa-
tients of varied ages, often can lead to symptoms,
and may occur in patients with or without struc-
tural heart disease.
The diagnosis of SVT is made primarily by
using the 12-lead electrocardiogram (ECG). Cor-
rect ECG diagnosis of SVT is important for
several reasons. First, because symptomatic pa-
tients who may have SVT often require rapid and
accurate treatment, misidentification of the type
of SVT can lead to inappropriate acute manage-
ment. Second, making the correct ECG diagnosis
of the type of SVT is important for long-term
prognosis and treatment strategies, including the
selection of effective medications or the decision to
refer a patient for catheter ablation. Finally, for
patients who do go on to catheter ablation,
a correct ECG diagnosis of the type of SVT
facilitates the appropriate choice of ablation strat-
egy. The correct choice is essential because the risk,
duration, complexity, and success rate of catheter
ablation varies based on the type of SVT [3].
Despite its importance, making the correct
ECG diagnosis of SVT type may be difficult for
* Corresponding author.
E-mail address: scheinman@medicine.ucsf.edu
(M.M. Scheinman).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.04.004
numerous reasons. First, during the acute care of
highly symptomatic or unstable patients who have
possible SVT, the health care provider may be
pressed for time. Second, the correct diagnosis
may also be elusive when only a rhythm strip (as
opposed to a 12-lead ECG) is available. Similarly,
SVT may be suspected in hospitalized patients
who are monitored on telemetry, but such mon-
itoring equipment may not be able to generate
a true 12-lead ECG. Third, the ECG morphology
may be more unusual in patients taking antiar-
rhythmic medications or in patients who have had
prior ablation or surgical procedures. For in-
stance, patients who have congenital heart disease
and have undergone corrective surgeries frequently
develop SVT as a late complication [4,5]. Fourth,
variability in the placement of ECG electrodes
may affect the ECG tracing. Finally, the ECG
computer diagnosis of arrhythmias is unreliable.
These potential difficulties represent just a few of
the challenges to accurate ECG diagnosis of SVT.
General points
This article provides a stepwise approach to
the 12-lead ECG diagnosis of SVT. It first
presents an initial approach to categorization
and diagnosis. Then, the common ECG manifes-
tations of each type of SVT are discussed in-
dividually. Finally, it presents a systematic
algorithm for diagnosing suspected SVT based
on the ECG.
The various forms of SVT include sinus tachy-
cardia (ST), focal atrial tachycardia (AT), multifo-
cal atrial tachycardia (MAT), atrial fibrillation
(AF), atrial flutter (AFl), AV node reentrant
cardiology.theclinics.com
428 KUMAR
tachycardia (AVNRT), AV reentrant tachycardia
(AVRT), junctional tachycardia (JT), and perma-
nent junctional reciprocating tachycardia (PJRT).
The first step in diagnosing SVT is to catego-
rize the rhythm as narrow complex (QRS ! 120
milliseconds) or wide complex (QRS R 120
milliseconds) and as regular (fixed R-R cycle
lengths) or irregular (variable R-R cycle lengths)
(Table 1). Some tachycardias are represented in
more than one category. The next step is to look
for evidence of atrial activity. If a discrete P
wave is visible, the relationship of the P wave to
the QRS complex should be evaluated. This rela-
tionship is characterized as either a ‘‘short-RP
tachycardia’’ or a ‘‘long-RP tachycardia.’’ In
short-RP tachycardia, the time from the P wave
to the preceding R wave is less than the time
from the P wave to the following R wave (RP !
PR). In long-RP tachycardia, the opposite is true
(RP O PR) (see Table 1).
In addition to cycle regularity and QRS width,
the following ECG criteria may help differentiate
the various types of SVT: heart rate during
tachycardia; mechanism of initiation/termination;
P wave/QRS/ST morphology; changes in cycle
length with the appearance of bundle-branch
block; and ECG changes in response to maneu-
vers such as carotid sinus massage or adenosine
administration. One study reports that viewing
Table 1
Initial categorization of tachycardic rhythms using the 12-lead electrocardiogram
Rhythm Narrow
Regular Short-RP
Typical (slow-fast) AVNRT
Atypical (slow-slow) AVNRT (usually)
Orthodromic AVRT
JT
AT (less commonly)
Long-RP
Atypical (fast-slow) AVNRT
Orthodromic AVRT (slowly conducting
accessory pathway)
ST
AT (more commonly)
PJRT
Flutter waves
AFl (fixed AV block)
Irregular AF
AFl (variable AV block)
MAT
Abbreviations: AF, atrial fibrillation; AFl, atrial flutter; AT, atrial tachycardia; AVNRT, atrioventricular node reen-
trant tachycardia; AVRT, atrioventricular reentrant tachycardia; JT, junctional tachycardia; MAT, multifocal atrial
tachycardia; PJRT, permanent junctional reciprocating tachycardia; ST, sinus tachycardia.
et al
the ECG at both normal (25 mm/s) and faster
(50 mm/s) paper speeds may improve diagnostic
accuracy [6]. The findings and criteria helpful in
the diagnosis of each particular SVT are discussed
in greater detail later.
Sinus tachycardia
ST is a long-RP tachycardia that presents with
a heart rate over 100 beats/min (bpm). The
P wave morphology is identical to that of normal
sinus rhythm. There usually is a gradual onset and
termination that often can be appreciated on
Holter or telemetry monitoring. A variant of ST,
sinus node reentrant tachycardia, comprises a re-
entrant loop originating in the sinus node, behaves
clinically like an SVT, and has a sudden onset and
abrupt termination. In sinus node reentry tachy-
cardia, the PR interval during tachycardia may
differ from normal sinus rhythm, although the
P wave morphology is identical. True sinus node
reentrant tachycardia seems to be uncommon,
with most of the reported cases originating from
high in the crista terminalis [7]. Patients who have
another variant of ST, termed ‘‘inappropriate si-
nus tachycardia,’’ show persistent tachycardia
during the day with marked increases in heart
rate in response to exercise. This diagnosis is
made from the 24-hour Holter recording, and
beta-blockers are the drugs of choice.
Wide
ST, JT, AT, AVNRT, orthodromic AVRT,
or AFl with aberrancy
Antidromic AVRT
Monomorphic ventricular tachycardia
AF, AFl, or MAT with aberrancy
AF, AFl with pre-excitation
Polymorphic ventricular tachycardia
 THE 12-LEAD ECG IN SUPRAVENTRICULAR TACHYCARDIA
Focal atrial tachycardia
AT is initiated and sustained within the atria
and is manifested by a single P wave morphology.
The P wave morphology depends of the site of
origin of the tachycardia. AT may be caused by
automaticity, triggered activity, or microreentry
[8–10]. Rarely, digitalis toxicity can result in AT,
usually with accompanying AV nodal block [11].
AT is usually a long-RP tachycardia with an atrial
rate usually less than 250 bpm and a ventricular
rate dependent on the presence or absence of
AV nodal block [1]. A classic example of focal
AT is shown in Fig. 1.
AT caused by abnormal automaticity often
exhibits a ‘‘warm-up’’ phenomenon in which the
heart rate gradually speeds up before reaching
a fixed peak. The diagnosis of AT caused by
triggered activity is often made in the electrophys-
iology laboratory when the AT can be initiated by
pacing the atrium at a critical rate. AT caused by
microreentry is often seen in patients who have
structural heart disease, usually is initiated by
a premature atrial beat, and typically is paroxys-
mal. It often is impossible to distinguish triggered
from reentrant AT. For AT caused by abnormal
automaticity, adenosine usually results in AV
nodal block without terminating the AT. On the
ECG, this AT is manifested as discernible P waves
continuing at the AT rate with occasionally
conducted QRS complexes. In contrast, triggered
ATs usually are terminated with adenosine. Al-
though these general aspects of the different types
of AT can be useful in diagnosis, significant
overlap of these properties exists.
To help localize the origin of the AT, one can
begin by analyzing the P wave polarity. To
Fig. 1. Focal atrial tachycardia with 1:1 conduction. Note the narrow, regular tachycardia. The P waves are marked
with arrows (long-RP interval), and the P wave morphology differs from that of sinus rhythm.
429
distinguish a right atrial from a left atrial location,
aVL and V1 are the essential leads. P waves that
are positive in V1 usually arise from the left
atrium, whereas P waves that are positive in
aVL usually arise from the right atrium or
occasionally from the right pulmonary vein [12–
15]. Positive P wave polarity in leads II, III, and
aVF predicts a superior origin; negative P wave
polarity in these leads predicts an inferior origin.
The polarity of the P wave may be difficult to in-
terpret if the P wave is inscribed upon the ST seg-
ment or T wave. Thus, the P wave morphology is
interpreted best when the heart rate or ventricular
rate is slowed, such as after adenosine administra-
tion. The finer points of P wave polarity during
AT that are most relevant for the site of ablation
are described in Table 2.
Multifocal atrial tachycardia
MAT is thought to be caused by enhanced
automaticity of multiple competing atrial foci.
MAT is almost always associated with pulmonary
disease and rarely, if ever, is seen in the setting of
digitalis toxicity. To make the diagnosis of MAT,
at least three different P wave morphologies must
be observed, which typically do not have a fixed
relationship to one another. The heart rate is
greater than 100 bpm and usually is irregular
(varying R-R intervals); the PR intervals may
vary; and AV nodal block may be present.
Atrial flutter
AFl is caused by a rapid macroreentrant circuit
occurring in the atria. The atrial rate is usually
240 to 340 bpm; the ventricular rate varies,
430 KUMAR
Table 2
P wave morphology in atrial tachycardiadgeneral principles
Location P wave polarity during atrial tachycardia
Right atrium þ or
in aVL Crista terminalis (CT)
and  in V1
Tricuspid annulus
Right atrial appendage
Left atrium  in aVL and Right pulmonary veins
þ in V1 Left pulmonary veins
Mitral annulus
Left atrial appendage
Interatrial septum Specific to Coronary sinus ostium
and adjacent location
structures Anteroseptum
Mid septum
Left septum
Abbreviations: þ, positive; , negative;
, biphasic.
From Refs. [12,15,42].
depending on the degree of AV nodal block. Of
note, chronic antiarrhythmic therapy may pro-
long atrial refractoriness and slow the AFl rate
[16]. The atrial-to-ventricular ratio can be regular
(eg, 2:1, 4:1) or irregular if the degree of AV nodal
block is variable. Unless the atrial rate is relatively
slow, 1:1 conduction is unusual unless it occurs by
an accessory pathway, which can result in wide
pre-excited QRS complexes.
The nomenclature of AFl is based on electro-
anatomic relationships observed during electro-
physiology studies; if the circuit is dependent
upon the cavotricuspid isthmus, it is defined as
typical AFl [17–20]. The reentrant circuit in typ-
ical AFl traverses the musculature around the tri-
cuspid valve. If this circuit activates the lateral
right atrium from superior to inferior and the
septum from inferior to superior, it is termed
counterclockwise typical AFl. If the lateral and
septal activations occur in reverse, while still be-
ing dependent on the cavotricuspid isthmus, it is
termed clockwise typical AFl. In contrast to dis-
crete P waves, typical AFl usually results in flut-
ter waves having a saw-tooth pattern. An
analysis of their polarity demonstrates that coun-
terclockwise typical flutter waves are positive in
V1 and negative in leads II, III, and aVF
et al
May be þ,
, or  in II, II, aVF in the high,
mid, and inferior CT, respectively;  or
in V1;
 in aVR
 in V1-V2; þ in aVL and frequently in I
Insufficient data
þ in V1–V6; þ in I; narrow P wave in V1
þ in V1–V6; /isoelectric in aVL; broad or
notched P wave in V1

with initially narrow negative deflection
in V1; /isoelectric in I, aVL; slightly
þ/isoelectric in II, III, aVF
Broadly þ in V1–V6;  in I, aVL; þ in II,
III, aVF
þ in V1; isoelectric in I;  in II, III, aVF; þ in aVL,
aVR
/
in V1; þ/
in II, III, aVF
/
in V1;  in two of three inferior
leads (II, III, aVF)
Uncommon; variable findings
(Fig. 2). Clockwise typical flutter waves are neg-
ative in V1 and positive in leads II, III, and aVF.
Of note, although the macroreentrant circuit is
located in the right atrium, the saw-tooth pattern
and its polarity primarily reflect the axis and se-
quence of left atrial activation.
In contrast to typical AFl, atypical AFl is any
other AFl that does not display a typical activa-
tion sequence. On the ECG, the saw-tooth flutter
waves may be absent, and the flutter wave polarity
differs from typical AFl. The flutter waves may be
quite varied in appearance and may appear
similar to discrete P waves. For example, in
atypical AFl arising from the left atrial septum,
the flutter waves are prominent in V1 but are flat
in the other leads and may appear similar to AT
[21,22]. The atrial rate of atypical AFl may also be
quite varied, depending on the location and spa-
tial extent of the macroreentrant circuit. Atypical
AFl can originate in the left or right atrium,
around a scar or surgical site, or around other
structures, which partly explains the variability
in its morphology and rate. The administration
of adenosine can be important in the diagnosis
of any type of AFl by virtue of its effect on the
AV node: the slowing of the ventricular rate can
make flutter waves more readily discernible.
 THE 12-LEAD ECG IN SUPRAVENTRICULAR TACHYCARDIA
Fig. 2. Typical, counterclockwise atrial flutter. The flutter waves are marked with arrows. Four-to-one atrioventricular
block is present, which facilitates examination of the flutter wave morphology. Note the characteristic saw-tooth appear-
ance in the inferior leads (II, III, aVF) and the positive flutter wave polarity in lead V1.
ECG diagnosis of typical versus atypical flutter is
clinically relevant, because the approach and suc-
cess rate of catheter ablation differ [3,23,24].
Atrial fibrillation
The ECG hallmarks of AF include an irregu-
larly irregular variation in R-R interval and the
absence of organized atrial activity. Underlying
AF may still be present even if the R-R intervals
are regular when concomitant complete AV nodal
block with a junctional or subjunctional escape
rhythm is present (occasionally seen with digitalis
toxicity). Coarse AF may be confused with AFl;
very fine AF may be confused with atrial paral-
ysis. In addition to AF, irregular junctional
rhythms, MAT, and AFl with variable block are
also in the differential diagnosis for irregularly
irregular rhythms (see Table 1). The atrial rates
in AF are variable but are usually greater than
350 bpm. The ventricular rate is usually signifi-
cantly slower and varies depending on AV nodal
function, unless an accessory pathway capable of
antegrade conduction is present.
Atrioventricular node reentrant tachycardia
Initiation of AVNRT is dependent on the
presence of dual AV node physiology, with two
pathways having differing conduction and refrac-
tory times. AVNRT can be divided into typical
and atypical forms, with the typical form being
more common [2,25]. Typical AVNRT, also
known as slow-fast AVNRT, uses a slow AV
nodal pathway for antegrade conduction and
a fast pathway for retrograde conduction
431
(Fig. 3). Atypical AVNRT can be slow-slow
AVNRT, using a slow antegrade and another
slow retrograde pathway (with different proper-
ties), or fast-slow, using a fast antegrade and
a slow retrograde pathway. The heart rate in
AVNRT can vary from 118 to 264 bpm (mean,
181
35) and is similar to the rates seen in
AVRT [26]. There typically is a 1:1 AV relation-
ship in AVNRT, but because the reentrant circuit
resides within the area of the AV node, 2:1 AV
block can be seen. Retrograde AV nodal atrial
block can occur also, although less commonly.
Typical AVNRT is initiated with an atrial prema-
ture beat and terminates with a P wave (antegrade
slow pathway). AVNRT is terminated by adeno-
sine (O90% of the time) or vagal maneuvers,
which can be useful in making a diagnosis [27].
Typical AVNRT is a short-RP tachycardia in
which the earliest retrograde atrial activity is
detected on the septum, near the AV node. The
RP interval is usually less than 70 milliseconds
[28]. Because retrograde atrial activation occurs
over a fast pathway, the retrograde P wave is
superimposed on the QRS and appears as
a pseudo S wave (present during AVNRT but
not during normal sinus rhythm) that is best
seen in leads II, III, and aVF. Similarly, a pseudo
R0 may also be present in lead V1. These ECG
findings are important because they are infre-
quently seen in AVRT [29,30]. Having a pseudo
S, a pseudo R0 , or both is 90% to 100% specific
for typical AVNRT and has an 81% positive pre-
dictive value for typical AVNRT [29,30]. These
criteria are only 42% sensitive for typical
AVNRT, however [29]. Occasionally in typical
AVNRT (20% of the time), the P wave is buried
432 KUMAR et al

Fig. 3. Typical atrioventricular node reentrant tachycardia. Note the rapid, regular, narrow complexes, the pseudo
S wave in leads I, II, and aVF, and the small pseudo R0 wave in lead V1.

within the QRS and is invisible [26]. Additionally, Orthodromic atrioventricular reentrant
ST segment depression of R2 mm is less common tachycardia
in AVNRT and is seen to a lesser extent and in
The reentrant loop in AVRT is comprised of
fewer leads than in AVRT [29]. In contrast to
the atria, AV node, ventricle, and accessory
AVRT, QRS alternans is an uncommon finding
pathway. In orthodromic AVRT, conduction
in AVNRT and may be related more to the rapid-
occurs antegrade through the AV node and
ity of the heart rate than to the underlying SVT
retrograde through an accessory pathway. Ortho-
mechanism [29,31]. Taking into account the
dromic AVRT is usually a short-RP tachycardia
pseudo S/R0 waves, the RP interval, and the lack
with an RP interval greater than 100 milliseconds
of significant ST depression in multiple leads
[28,29]. If the accessory pathway has slow retro-
(the Jaeggi algorithm), a correct diagnosis of typ-
grade conduction, however, the RP is significantly
ical AVNRT can be made by ECG analysis 76%
longer, consistent with a long-RP tachycardia. Or-
of the time [28,29].
thodromic AVRT also is a narrow complex tachy-
In atypical (slow-slow) AVNRT, the retro-
cardia, unless bundle-branch block or aberrancy
grade atrial activation usually is seen first near
is present (Fig. 4). The heart rate in AVRT ranges
the coronary sinus ostium. The P wave may be
from 124 to 256 bpm (mean, 183
32 bpm) [32].
distinct from the QRS and appears negative in
AVRT usually is initiated with one or more ven-
leads II, III, and aVF and positive in leads V1, V2,
tricular premature beats and usually terminates
aVR, and aVL [32]. Atypical (slow-slow) AVNRT
with a QRS complex. AVRT can be terminated
is usually a short-RP tachycardia and often can-
by adenosine (O90% of the time) or vagal maneu-
not be distinguished from orthodromic AVRT,
vers [27].
as discussed later [31,33].
In normal sinus rhythm, antegrade conduction
Atypical (fast-slow) AVNRT is a long-RP
over an accessory pathway resulting in a short PR
tachycardia in which the earliest retrograde atrial
interval and a delta wave on the surface ECG is
activity is seen in the posteroseptal right atrium or
the hallmark of the Wolff-Parkinson-White ECG
in the coronary sinus. This form of AVNRT is
pattern (Fig. 5). The delta wave is the ECG man-
thought to use the same pathways as in typical
ifestation of ventricular pre-excitation. If no ante-
AVNRT, but in reverse [34]. The P wave precedes
grade conduction is evident, but the pathway is
the QRS and is negative in leads II, III, and aVF
capable of retrograde conduction, the pathway is
and is positive in leads V1, V2, aVR, and aVL.
defined as concealed. In Wolff-Parkinson-White,
Atypical (fast-slow) AVNRT may be indistin-
the disappearance of the delta wave during tachy-
guishable from a low AT or orthodromic AVRT
cardia is caused by orthodromic AVRT. The loss
with a posteroseptal pathway.
 THE 12-LEAD ECG IN SUPRAVENTRICULAR TACHYCARDIA 433

Fig. 4. Orthodromic atrioventricular reentrant tachycardia. Note the narrow, regular, rapid tachycardia. The P waves
buried in the ST segment (short-RP interval) are marked with an arrow. No pseudo S/R0 waves are seen.

of the delta wave during sinus rhythm results from
the loss of pre-excitation caused by a poorly con-
ducting accessory pathway incapable of conduct-
ing at higher rates.
Most patients who have orthodromic AVRT
(81%–87%) have visible retrograde P waves that
are best seen in leads I, II, III, aVF, and V1
[29,32]. QRS alternans (alternating QRS ampli-
tude) is common (45% of cases), unlike AVNRT
[29]. Also unlike AVNRT, ST segment depression
of R2 mm is common in AVRT, particularly in
patients who have no visible P wave or who
have an RP interval less than 100 milliseconds;
the ST depressions also are seen in several leads
(mean, 4.4
1.4 leads) [29]. By using the Jaeggi
algorithm, which takes into account the absence
of pseudo S/R0 waves, an RP interval greater
than 100 milliseconds, and the presence of ST de-
pression of R2 mm, a correct diagnosis of AVRT
can be made using the surface ECG 88% of the
time [28,29].
After diagnosing AVRT, it is helpful to localize
the pathway because the feasibility, approach, and
success of catheter ablation depend on the path-
way site. In Wolff-Parkinson-White syndrome, the
polarity of the delta wave during normal sinus
rhythm often predicts accessory pathway location
unless multiple accessory pathways are present (an
occasional finding) [36,37]. Accessory pathway
localization based on the delta wave during nor-
mal sinus rhythm has been described in detail pre-
viously [38,39]. The polarity and morphology of
the retrograde P wave during orthodromic
AVRT also may assist in pathway localization
[14,32,35]. In general, a positive P wave in leads
II, III, and aVF suggests an anterior accessory
pathway; a negative P wave in leads II, III, and
aVF suggests an inferior accessory pathway; a pos-
itive P wave in lead V1 or a negative P wave in lead
I suggests a left-sided accessory pathway; and
a negative P wave in lead V1 or a positive P wave
in lead I suggests a right-sided accessory pathway.

Fig. 5. Wolff-Parkinson-White pattern. Note the short PR interval, the delta waves (pre-excitation), which are positive
in leads I and aVL and negative in leads II, III, and aVF, and the transition of the delta wave axis from lead V1 to V2
(consistent with a right posteroseptal accessory pathway).
434
The development of bundle-branch block ab-
errancy during SVT also can be a helpful clue. If
the tachycardia cycle length slows when bundle-
branch block appears, then the SVT is an AVRT
and not AVNRT or AT, because the ventricle is
not an integral component of the circuit in these
latter two arrhythmias. More importantly, the
slowing of the tachycardia cycle length in with the
appearance of bundle branch block proves that
this AVRT uses an accessory pathway with the
pathway on the same side as the blocked bundle
branch. The increase in cycle length can be
explained by the additional conduction time re-
quired for the depolarization wave front to go
from the normally conducting bundle branch (on
the contralateral side) across the ventricular
septum and into the accessory pathway and
atrium.
Focal junctional tachycardia
JT is uncommon in adults and often is a di-
agnosis of exclusion. This arrhythmia is more
common in children and often is irregular. In
adults, JT manifests as a regular, short-RP
tachycardia (if ventriculo-atrial [VA] conduction
is present) with a narrow QRS complex, unless
bundle-branch aberrancy is present. It can be
caused by enhanced automaticity or triggered
activity. The VA relationship may be dissociated,
associated with 1:1 retrograde conduction, or
associated with various degrees of VA retrograde
block [36]. When dissociation is present, the P
waves seen on the ECG are most likely caused
by sinus rhythm. If the P waves appear associated
with the QRS rhythm, they typically are negative
in leads II, III, and aVF because of retrograde
atrial activation. The diagnosis of JT is based pri-
marily on tachycardia initiation without the need
for a critical AV nodal delay. Beta-blockers are
Fig. 6. Wide-complex tachycardia caused by atrial fibrillation in a different patient who had Wolff-Parkinson-White
syndrome. Note the very rapid rate, the irregularity, and the wide, bizarre, varying QRS morphology caused by varying
degrees of fusion from AV node conduction and antegrade accessory pathway conduction. The negative delta wave in
lead II suggests a posterior (inferior) accessory pathway location.
KUMAR et al
the initial drug of choice, and ablation may be rec-
ommended for drug-refractory cases.
Permanent junctional reciprocating tachycardia
PJRT is an incessant long-RP tachycardia,
typically seen in children but occasionally seen
in adults, which may lead to tachycardia cardio-
myopathy [37]. The cardiomyopathy frequently
abates after successful treatment of the arrhyth-
mia. In PJRT the heart rate is usually between
100 and 200 bpm but can vary significantly be-
cause of autonomic influences [37]. The arrhyth-
mia uses the atria, AV node, the ventricle, and
an accessory pathway and is similar to ortho-
dromic AVRT. PJRT enters into the differential
diagnosis of atypical AVNRT or AT originating
from the inferior atrium. The accessory pathway
usually is concealed and possesses decremental
conduction properties. In most cases, the acces-
sory pathway is within or near the coronary sinus
ostium. PJRT usually starts spontaneously after
a sinus beat and does not require a premature
beat or a change in the PR interval for initiation.
On the ECG, the QRS complexes usually are nar-
row and have a 1:1 AV relationship. The P waves
usually are broad and negative in leads II, III, and
aVF [38]. The RP interval is much longer in PJRT
than in most orthodromic AVRTs. PJRT usually
terminates in the retrograde limb, which is sensi-
tive to vagal maneuvers [38,39].
Wide-complex supraventricular tachycardias
caused by pre-excitation
It is important to differentiate patients who
have pre-excited wide-complex tachycardia
(WCT) from those who have SVT with aberrancy
or ventricular tachycardia. Of particular concern
are patients who have AF and AFl with antegrade
 THE 12-LEAD ECG IN SUPRAVENTRICULAR TACHYCARDIA
conduction through an accessory pathway. Be-
cause the accessory pathway does not possess the
same decremental conduction properties as the
AV node, AF or AFl with antegrade conduction
through an accessory pathway may conduct
rapidly, often with ventricular rates greater than
280 to 300 bpm, and may degenerate into ven-
tricular fibrillation. An irregular, rapid WCT
should raise suspicion for AF with pre-excitation
(Fig. 6). Similarly, AFl or AT with very rapid or
1:1 AV conduction should raise concern for pre-
excitation. In all cases, delta waves should be pres-
ent during tachycardia but may be overshadowed
by the wide-complex morphology. In AF, AFl, or
AT, varying degrees of fusion through the path-
way and through the AV node may produce
beat-to-beat variation in the QRS morphology.
If the rhythm is maximally pre-excited, the
QRS morphology may appear similar to that of
patients who have ventricular tachycardia, but
conversion to sinus rhythm will produce the pre-
excitation pattern in patients who have Wolff-
Parkinson-White syndrome.
Antidromic AVRT is a WCT that exhibits
maximal pre-excitation (Fig. 7). In antidromic
AVRT, conduction occurs antegrade through an
accessory pathway and retrograde through the
AV node. Delta waves should be evident both
during normal sinus rhythm and during tachycar-
dia. The retrograde P waves in antidromic AVRT
are frequently seen in a 1:1 VA relationship pre-
ceding the QRS [40]. Because dual AV node phys-
iology (having both fast and slow pathways) is
common in patients who have accessory path-
ways, retrograde conduction may occur during
antidromic AVRT by a slow AV nodal pathway,
Fig. 7. Antidromic atrioventricular reentrant tachycardia in the same patient as in Fig. 5, using a right posteroseptal
accessory pathway. Note the wide-complex, regular rhythm. The delta waves are more prominent because of maximal
pre-excitation.
435
a fast AV nodal pathway, or alternate between
fast and slow AV nodal pathways [41]. Thus, the
characteristics of the P wave and PR interval
may vary during tachycardia.
Summary of steps required for ECG diagnosis
of SVT
1. Assess QRS width and regularity.
2. Look for evidence of atrial activity.
3. If distinct P waves are seen
 Determine the RP relationship
 Evaluate the P wave morphology during
SVT and, if possible, during normal sinus
rhythm
 Evaluate the AV relationship (eg, 1:1, 2:1,
dissociated)
4. If flutter waves are seen or suspected
 Evaluate their morphology, preferably if
the ventricular rate can be slowed
5. If no P waves are seen
 Look for irregularity or the absence of an
isoelectric baseline, both suggesting AF
 Look for a pseudo S/R0 to suggest AVNRT
6. Assess the underlying atrial and ventricular
rates.
7. Examine the initiation (atrial premature beat,
ventricular premature beat, warm-up, no ini-
tiating factors) and termination. (Does the
SVT terminate with a P wave or a QRS;
does it cool down slowly?)
8. Examine the response to adenosine. (Does it
terminate the tachycardia, and, if so, how?
Does it cause AV block, and, if so, what hap-
pens to the underlying atrial rhythm?)
436 KUMAR
References
[1] Ganz LI, Friedman PL. Supraventricular tachycar-
dia. N Engl J Med 1995;332:162–73.
[2] Blomstrom-Lundqvist C, Scheinman MM, Aliot
EM, et al. ACC/AHA/ESC guidelines for the man-
agement of patients with supraventricular arrhyth-
miasdexecutive summary. A report of the American
College of Cardiology/American Heart Association
Task Force on Practice Guidelines and the Euro-
pean Society of Cardiology Committee for Practice
Guidelines (Writing Committee to Develop Guide-
lines for the Management of Patients with Supraven-
tricular Arrhythmias) developed in collaboration
with NASPE-Heart Rhythm Society. J Am Coll
Cardiol 2003;42:1493–531.
[3] Scheinman MM, Huang S. The 1998 NASPE pro-
spective catheter ablation registry. Pacing Clin Elec-
trophysiol 2000;23:1020–8.
[4] van den Bosch AE, Roos-Hesselink JW, Van Dom-
burg R, et al. Long-term outcome and quality of life
in adult patients after the Fontan operation. Am J
Cardiol 2004;93:1141–5.
[5] Weipert J, Noebauer C, Schreiber C, et al. Occur-
rence and management of atrial arrhythmia after
long-term Fontan circulation. J Thorac Cardiovasc
Surg 2004;127:457–64.
[6] Accardi AJ, Miller R, Holmes JF. Enhanced diagnosis
of narrow complex tachycardias with increased elec-
trocardiograph speed. J Emerg Med 2002;22:123–6.
[7] Marrouche NF, Beheiry S, Tomassoni G, et al.
Three-dimensional nonfluoroscopic mapping and
ablation of inappropriate sinus tachycardia. Proce-
dural strategies and long-term outcome. J Am Coll
Cardiol 2002;39:1046–54.
[8] Chen SA, Chiang CE, Yang CJ, et al. Sustained
atrial tachycardia in adult patients. Electrophysio-
logical characteristics, pharmacological response,
possible mechanisms, and effects of radiofrequency
ablation. Circulation 1994;90:1262–78.
[9] Haines DE, DiMarco JP. Sustained intraatrial reen-
trant tachycardia: clinical, electrocardiographic and
electrophysiologic characteristics and long-term fol-
low-up. J Am Coll Cardiol 1990;15:1345–54.
[10] Wu D, Amat-y-leon F, Denes P, et al. Demonstra-
tion of sustained sinus and atrial re-entry as a mech-
anism of paroxysmal supraventricular tachycardia.
Circulation 1975;51:234–43.
[11] Smith TW, Antman EM, Friedman PL, et al. Digi-
talis glycosides: mechanisms and manifestations of
toxicity. Part II. Prog Cardiovasc Dis 1984;26:
495–540.
[12] Kistler PM, Kalman JM. Locating focal atrial tachy-
cardias from P-wave morphology. Heart Rhythm
2005;2:561–4.
[13] Tang CW, Scheinman MM, Van Hare GF, et al. Use
of P wave configuration during atrial tachycardia to
predict site of origin. J Am Coll Cardiol 1995;26:
1315–24.
et al
[14] Waldo AL, Maclean AH, Karp RB, et al. Sequence
of retrograde atrial activation of the human heart.
Correlation with P wave polarity. Br Heart J 1977;
39:634–40.
[15] Yamane T, Shah DC, Peng JT, et al. Morphological
characteristics of P waves during selective pulmo-
nary vein pacing. J Am Coll Cardiol 2001;38:
1505–10.
[16] Tai CT, Chen SA. Mechanisms of antiarrhythmic
drug action on termination of atrial flutter. Pacing
Clin Electrophysiol 2001;24:824–34.
[17] Lee KW, Yang Y, Scheinman MM. Atrial flutter:
a review of its history, mechanisms, clinical features,
and current therapy. Curr Probl Cardiol 2005;30:
121–67.
[18] Yang Y, Varma N, Keung EC, et al. Reentry within
the cavotricuspid isthmus: an isthmus dependent cir-
cuit. Pacing Clin Electrophysiol 2005;28:808–18.
[19] Scheinman MM, Yang Y, Cheng J. Atrial flutter:
part II nomenclature. Pacing Clin Electrophysiol
2004;27:504–6.
[20] Olgin JE, Kalman JM, Fitzpatrick AP, et al. Role of
right atrial endocardial structures as barriers to con-
duction during human type I atrial flutter. Activa-
tion and entrainment mapping guided by
intracardiac echocardiography. Circulation 1995;
92:1839–48.
[21] Bochoeyer A, Yang Y, Cheng J, et al. Surface elec-
trocardiographic characteristics of right and left
atrial flutter. Circulation 2003;108:60–6.
[22] Goya M, Takahashi A, Nuruki N, et al. A peculiar
form of focal atrial tachycardia mimicking atypical
atrial flutter. Jpn Circ J 2000;64:886–9.
[23] Della BP, Fraticelli A, Tondo C, et al. Atypical atrial
flutter: clinical features, electrophysiological charac-
teristics and response to radiofrequency catheter ab-
lation. Europace 2002;4:241–53.
[24] Marrouche NF, Natale A, Wazni OM, et al. Left
septal atrial flutter: electrophysiology, anatomy,
and results of ablation. Circulation 2004;109:
2440–7.
[25] Sung RJ, Styperek JL, Myerburg RJ, et al. Initiation
of two distinct forms of atrioventricular nodal reen-
trant tachycardia during programmed ventricular
stimulation in man. Am J Cardiol 1978;42:404–15.
[26] Tai CT, Chen SA, Chiang CE, et al. A new electro-
cardiographic algorithm using retrograde P waves
for differentiating atrioventricular node reentrant
tachycardia from atrioventricular reciprocating
tachycardia mediated by concealed accessory path-
way. J Am Coll Cardiol 1997;29:394–402.
[27] DiMarco JP, Miles W, Akhtar M, et al. Adenosine
for paroxysmal supraventricular tachycardia: dose
ranging and comparison with verapamil. Assess-
ment in placebo-controlled, multicenter trials. The
Adenosine for PSVT Study Group. Ann Intern
Med 1990;113:104–10.
[28] Jaeggi ET, Gilljam T, Bauersfeld U, et al. Electrocar-
diographic differentiation of typical atrioventricular
 THE 12-LEAD ECG IN SUPRAVENTRICULAR TACHYCARDIA
node reentrant tachycardia from atrioventricular
reciprocating tachycardia mediated by concealed
accessory pathway in children. Am J Cardiol 2003;
91:1084–9.
[29] Arya A, Kottkamp H, Piorkowski C, et al. Differen-
tiating atrioventricular nodal reentrant tachycardia
from tachycardia via concealed accessory pathway.
Am J Cardiol 2005;95:875–8.
[30] Kalbfleisch SJ, el-Atassi R, Calkins H, et al. Differ-
entiation of paroxysmal narrow QRS complex
tachycardias using the 12-lead electrocardiogram.
J Am Coll Cardiol 1993;21:85–9.
[31] Kay GN, Pressley JC, Packer DL, et al. Value of the
12-lead electrocardiogram in discriminating atrio-
ventricular nodal reciprocating tachycardia from
circus movement atrioventricular tachycardia utiliz-
ing a retrograde accessory pathway. Am J Cardiol
1987;59:296–300.
[32] Chen SA, Tai CT, Chiang CE, et al. Electrophysio-
logic characteristics, electropharmacologic re-
sponses and radiofrequency ablation in patients
with decremental accessory pathway. J Am Coll
Cardiol 1996;28:732–7.
[33] Oh S, Choi YS, Sohn DW, et al. Differential diagno-
sis of slow/slow atrioventricular nodal reentrant
tachycardia from atrioventricular reentrant tachy-
cardia using concealed posteroseptal accessory path-
way by 12-lead electrocardiography. Pacing Clin
Electrophysiol 2003;26:2296–300.
[34] Yamabe H, Shimasaki Y, Honda O, et al. Demon-
stration of the exact anatomic tachycardia circuit
in the fast-slow form of atrioventricular nodal reen-
trant tachycardia. Circulation 2001;104:1268–73.
437
[35] Chen SA, Tai CT, Chiang CE, et al. Role of the sur-
face electrocardiogram in the diagnosis of patients
with supraventricular tachycardia. Cardiol Clin
1997;15:539–65.
[36] Scheinman MM, Gonzalez RP, Cooper MW, et al.
Clinical and electrophysiologic features and role of
catheter ablation techniques in adult patients with
automatic atrioventricular junctional tachycardia.
Am J Cardiol 1994;74:565–72.
[37] Lindinger A, Heisel A, von Bernuth G, et al. Perma-
nent junctional re-entry tachycardia. A multicentre
long-term follow-up study in infants, children and
young adults. Eur Heart J 1998;19:936–42.
[38] Gaita F, Haissaguerre M, Giustetto C, et al. Cathe-
ter ablation of permanent junctional reciprocating
tachycardia with radiofrequency current. J Am
Coll Cardiol 1995;25:648–54.
[39] Prystowsky EN, Yee R, Klein GJ. Wolff-Parkinson-
White syndrome. In: Zipes DP, Jalife J, editors.
Cardiac electrophysiology: from cell to bedside.
4th edition. Philadelphia: Saunders; 2004. p. 874.
[40] Packer DL, Gallagher JJ, Prystowsky EN. Physio-
logical substrate for antidromic reciprocating tachy-
cardia. Prerequisite characteristics of the accessory
pathway and atrioventricular conduction system.
Circulation 1992;85:574–88.
[41] Chen YJ, Chen SA, Chiang CE, et al. Dual AV node
pathway physiology in patients with Wolff-Parkin-
son-White syndrome. Int J Cardiol 1996;56:275–81.
[42] Marrouche NF, SippensGroenewegen A, Yang Y,
et al. Clinical and electrophysiologic characteristics
of left septal atrial tachycardia. J Am Coll Cardiol
2002;40:1133–9.
 Cardiol Clin 24 (2006) 439–451

Value of the 12-Lead ECG in Wide QRS Tachycardia

John M. Miller, MD*, Mithilesh K. Das, MD, Anil V. Yadav, MD,
Deepak Bhakta, MD, Girish Nair, MD, Cesar Alberte, MD
Indiana University School of Medicine, Krannert Institute of Cardiology, 1801 N. Capitol Avenue,
Indianapolis, IN 46202, USA

Mr. V., A 64-year-old man, came to the emer-
gency room because of sudden onset of palpitations
2 hours earlier. He was mildly short of breath but
had no chest pain; there was neither history of prior
episodes nor even any cardiovascular illness aside
from hypertension. On examination, he appeared
somewhat anxious but comfortable, had blood
pressure 110/70 and a regular heart rate at 155
beats per minute (bpm); he had no evidence of heart
failure except for ‘‘occasional elevation in jugular
venous pressure.’’ ECG showed a regular rhythm at
145 bpm with a QRS duration of 160 milliseconds
(Fig. 1); he had no prior ECGs. The staff doctors
argued among themselves whether this was supra-
ventricular tachycardia (SVT) or ventricular tachy-
cardia (VT), eventually reasoning that with his
normal blood pressure and mentation, it must be
SVT. Just as he was about to receive an injection
of adenosine, the tachycardia reverted to sinus
rhythm at 75 bpm. After another hour’s observa-
tion to ensure he was stable, Mr. V. was sent
home with a diagnosis of SVT and told to make
an appointment with a cardiologist. Two weeks
later, he had another episode of palpitations but
suddenly became very short of breath and passed
out. His wife called 911, but when rescue workers
arrived, refractory ventricular fibrillation was pres-
ent and sinus rhythm could not be restored.
This story illustrates several of the difficulties
encountered when a patient develops a wide QRS
tachycardia (WQRST). A fundamental errord
assuming that with a normal blood pressure and
mental status, the rhythm must be SVTdled to an
* Corresponding author.
E-mail address: jmiller6@iupui.edu (J.M. Miller).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.03.003
incorrect diagnosis with a relatively benign progno-
sis, whereas Mr. V. actually had VT (his peculiar
neck vein pattern was cannon A waves) that put him
at risk of sudden cardiac deathdwhich, sadly,
occurred 2 weeks later. In fact, his ECG in the
emergency room (Fig. 1) clearly demonstrated most
of the criteria that strongly favor a diagnosis of VT.
Few clinical situations cause more anxiety among
physicians than caring for a patient with an ongoing
episode of WQRST; one must act quickly and
decisively (but often without the confidence that
the action is correct), and the consequences of
making an incorrect diagnosis are potentially di-
sastrous. Because optimal long-term management
depends on a correct initial diagnosis, the impor-
tance of correctly identifying the nature of the
WQRST is obvious. In this article, we will review
the ECG criteria that can help the clinician distin-
guish among the possible diagnoses. Most of these
criteria have been in the literature for more than 15
years, but are still poorly recognized or applied by
clinicians [1–8]. Clinical history [9] as well as nonelec-
trocardiographic methods, such as evaluation of
neck veins for cannon A waves, variable intensity
of S1, and use of carotid sinus massage, are also valu-
able in the differential diagnosis of WQRST and
should not be ignored; however, a detailed discus-
sion of these is beyond the scope of this article.
Definitions
In the following discussion, we will use these
definitions:
Wide QRS complex tachycardia: rhythm with
QRS duration R120 milliseconds and rate
R100 bpm
cardiology.theclinics.com
440 MILLER et al

Fig. 1. ECG of patient in emergency room diagnosed as supraventricular tachycardia because of stable blood pressure.
Correct diagnosis is ventricular tachycardia, by virtue of QRS duration, rightward superior axis, monophasic R in V1,
R/S ratio !1 in V6, and atrioventricular dissociation.

Ventricular tachycardia: tachycardia requiring Diagnostic possibilities

participation of structures below the bundle
Possible causes of WQRST include the follow-
of His
ing (examples of leads V1 and V6 shown in
Supraventricular tachycardia: tachycardia re-
Fig. 2):
quiring participation of structures above
bundle of His 1) Ventricular tachycardiadThis is the most
Left bundle branch block (LBBB) configura- common cause of WQRST in all published
tion: QRS duration R120 milliseconds with series, accounting for 70% to 80% of cases
predominantly negative terminal portion of depending on the population studied. Most
lead V1 cases occur in the setting of significant struc-
Right bundle branch block (RBBB) configura- tural heart disease, and many patients are
tion: QRS duration R120 milliseconds with quite ill at the time of the rhythm. These fea-
predominantly positive terminal portion of tures and the implications of the diagnosis of
lead V1 VT are in part responsible for the anxiety

Fig. 2. Examples of leads V1 and V6 in both LBBB and RBBB configurations in each cause of WQRST. See text for
further clarification.
 VALUE OF THE 12-LEAD ECG IN THE WIDE QRS TACHYCARDIA
provoked by seeing a patient with WQRST.
Because ventricular activation is not medi-
ated to any significant extent by the normal
His-Purkinje system (HPS; His bundle, bun-
dle branches, and fascicles), but is instead de-
pendent on muscle–muscle conduction, the
QRS complex during VT generally does not
resemble true LBBB or RBBB.
2) Supraventricular tachycardia with:
a. Aberrant interventricular conduction
(SVT-A)dBBB may be present at all
times, including during the resting rhythm
(‘‘permanent’’), or only during some epi-
sodes of WQRST due to refractoriness of
one of the bundle branches. This may be
purely due to the rapid rate or the sudden-
ness of rate increase (‘‘rate-related’’) from
the baseline rhythm (‘‘functional’’). Be-
cause ventricular activation is mediated
by the nonblocked portion of the HPS,
the QRS complex in SVT resembles known
patterns of BBB and fascicular block. In
most WQRST series, SVT-A accounts for
the second largest group after VT.
b. Abnormal ventricular activation using an
atrioventricular (AV) accessory pathway
(AP), such as with Wolff-Parkinson-White
syndrome. In these so-called preexcited
tachycardias, ventricular activation pro-
ceeds, at least in part, over an AP that by-
passes the AV node. These pathways
almost always insert into the epicardial as-
pect of ventricular muscle near the AV
groove, remote from the HPS; muscle–
muscle conduction is responsible for ven-
tricular activation, as is the case with
VT. Preexcited SVT is a relatively uncom-
mon cause of WQRST among adults.
c. Abnormal baseline QRS configurationd
Patients with a variety of disorders that alter
the QRS complex during resting rhythm may
have episodes of SVT conducted to the ven-
tricles with the same abnormal QRS pattern.
Disorders that produce bizarre QRS pat-
terns include dilated and hypertrophic car-
diomyopathy and repaired congenital heart
disease. This group currently accounts for
a small portion of all WQRSTs, but will
probably increase in prevalence as more pa-
tients with repaired congenital heart disease
reach adulthood, and patients with severe
cardiomyopathies have improved longevity.
d. Nonspecific QRS widening due to electro-
lyte or drug effectsdTransient metabolic
441
abnormalities such as hyperkalemia and
acidosis, as well as drug effects (procaina-
mide and other Type IA or IC antiar-
rhythmic drugs, amiodarone, and so on)
can diffusely widen the QRS complex
and transform an otherwise narrow com-
plex SVT into one with a wide QRS.
This is an uncommon but important
group of WQRSTs; correct diagnosis re-
lies on a good history and strong clinical
suspicion of specific disorders. One clue
to the presence of hyperkalemia is a nor-
mal or even short QT interval (usually
with characteristic peaked T waves) dur-
ing WQRST, in contrast to the slight pro-
longation of QT in the presence of a wide
QRS complex.
3) Ventricular pacingdAlthough patients with
ventricular pacemakers can certainly have
ventricular rates O100 bpm with a wide
QRS, pacemaker stimulus artifacts (‘‘spikes’’)
are usually easily visible, and the nature of the
rhythm is evident. However, in some cases
pacemaker artifacts may be very small, such
that the rhythm is not clearly recognized as
paced. This relatively uncommon cause of
WQRST is likely to become somewhat more
prevalent with continuing advances in pace-
maker technology (voltage regulation to
pace just above threshold, and so on).
In all large series concerned with the differential
diagnosis of WQRST, the overwhelming majority
of cases are VT (up to 80%); the next largest group
is SVT-A, with other diagnoses contributing a small
fraction of cases. From a practical standpoint, it is
thus clear that the major differential diagnosis of
WQRST is between VT and SVT-A. In most other
diagnoses, the QRS complex does not resemble any
typical form of aberration. Thus, one approach to
establishing the correct diagnosis for WQRST is to
pose the question, ‘‘is the QRS complex compatible
with some form of aberration?’’ If it is, the rhythm is
likely SVT; if it is not, the rhythm is more than likely
VT, with other possible diagnoses depending on
clinical circumstances (pacemaker present, suspi-
cion of hyperkalemia or drug effect, and so on).
ECG criteria
Many ECG criteria have been proposed to
distinguish VT from SVT-A. These may be
difficult to remember in the urgency of a clinical
442 MILLER
setting; however, if one recalls their basisdis the
QRS compatible with aberration, or not?dit is
easier to ‘‘reconstruct’’ the criteria and apply them
correctly. These criteria have had variable sensi-
tivity and specificity in prior studies, many of
which contained relatively few cases. We therefore
examined a series of 650 WQRSTs and evaluated
the accuracy of the established criteria. In each
case, the correct diagnosis was confirmed by
electrophysiology study; ECGs were obtained
from 385 individuals (279 [73%] men), aged 9–
89 years (mean 53 G 19 years) among whom
structural heart disease was present in 54% (prior
myocardial infarction in 38%, cardiomyopathy in
13%, and repaired congenital heart disease in
2%). Within this series, 473 (73%) WQRSTs were
VT, 132 (20%) were SVT-A, 37 (6%) were
preexcited SVT, 6 (!1%) were SVT with abnor-
mal baseline QRS complex, 2 (!1%) were
ventricular pacing with poorly evident stimulus
artifacts; none in this series of WQRST evaluated
at electrophysiology study were SVT with drug/
electrolyte-based QRS widening. The proportion
of preexcited SVT is probably higher than one
would expect in a consecutive series of adult
patients presenting to the health care system
with WQRST, due to the selected nature of the
population (referred for electrophysiology study).
The following paragraphs discuss the estab-
lished criteria for the diagnosis of WQRST and
particularly their ability to distinguish VT from
SVT-A (Table 1).
1) QRS duration. It has been noted that in most
cases of SVT-A, the QRS duration is %140
milliseconds in RBBB aberration, and up to
160 milliseconds in LBBB aberration [1].
Therefore, QRS complexes wider than this
are less likely to be aberration, and thus VT
is the more probable diagnosis. As noted
above, influences that nonspecifically widen
the QRS such as drug effects can decrease
the value of this rule. In patients with VT oc-
curring in the absence of structural heart dis-
ease, or in VTs with earliest activation in the
septum in patients with structural heart dis-
ease, the QRS complex can be relatively nar-
row (sometimes even !120 milliseconds).
Among SVT-A cases in our series, the
mean (GSD) QRS duration was 136 G 18
milliseconds (for VT, 166 G 38 milliseconds,
P ! 0.001 compared with SVT-A); 77/132
(58%) of SVT-As had QRS duration %140
milliseconds versus 125/473 (26%) VTs
et al
(sensitivity 0.58, specificity 0.73, negative
predictive accuracy 0.86); thus, 348/473
(74%) of VTs had QRS duration O140 milli-
seconds. Further, only 28 (21%) of SVT-As
in our series had a QRS duration in excess of
160 milliseconds, compared with 302 (64%)
VTs (P ! 0.001; sensitivity 0.64, specificity
0.79).
2) QRS axis. In most cases of SVT-A, the QRS
axis is either normal (0
to þ 90
) or shows
left anterior fascicular block (0
to 90
) or
left posterior fascicular block (þ90
to
180
). In our series, 124/132 (94%) of SVT-
A ECGs had a frontal plane axis within these
bounds versus 378/473 (80%) of VTs. How-
ever, an axis between 90
and 180
(right-
ward superior) is not readily achieved with
any combination of BBB or fascicular block,
and is thus unlikely to be SVT-A (therefore,
VT). This is easily recognized on the stan-
dard 12-lead ECG, because leads 1, 2, and
3 are all predominantly negative (Fig. 1) [3].
In our series, 93/473 (20%) of VTs had an
axis between 90
and 180
versus 5/132
(4%) of SVT-As (P ! 0.001; sensitivity
0.20, specificity 0.96). Thus, although this is
a highly specific criterion, preexisting QRS
abnormalities can produce a rightward supe-
rior axis; a previous resting ECG, if avail-
able, can aid the analysis.
3) QRS concordance. A concordant pattern is
one in which the predominant QRS deflec-
tion is either all positive or all negative across
the precordial leads. This is a relatively un-
common pattern in SVT-A because with
LBBB aberration, the negative QRS complex
in V1 and V2 becomes positive in V4-6, and
in RBBB aberration, tall terminal R waves in
V1 and V2 diminish by V3 and V4 before in-
creasing again in V5 and V6. In our series,
72/81 (88%) of ECGs with a concordant pat-
tern were VT (Fig. 3). Although relatively
specific for VT, only 15% of VTs demon-
strated this finding. We found this to be
more useful in RBBB-type QRS complexes
(‘‘positive’’ concordance): concordance was
present in 46/254 (18%) of VTs with
RBBB-type pattern, only 5/93 (5%)
RBBB-type SVT-As showed concordance
(P ! 0.005, sensitivity 0.18, specificity 0.95).
On the other hand, a negative concordant
pattern discriminated poorly: only 26/219
(12%) of LBBB-type VTs showed negative
concordance while 4/39 (10%) of SVT-As
 VALUE OF THE 12-LEAD ECG IN THE WIDE QRS TACHYCARDIA 443

Table 1
Summary of electrocardiographic criteria to distinguish VT from SVT-A
Criterion VT SVT-A P value Sensitivity Specificity PPA NPA
QRS duration
Mean G SD 166 G 38 ms 136 G 18 ms !0.001 d d d d
%140 ms 125/473 (26%) 77/132 (58%) !0.001 0.58 0.73 0.38 0.86
%160 ms 171/473 (36%) 104/132 (79%) !0.001 0.64 0.79 0.38 0.92
QRS Axis
Right superior 93/473 (20%) 5/132 (4%) !0.001 0.20 0.96 0.95 0.25
(90
to 180
)
Right inferior 44/219 (20%) 1/39 (3%) !0.01 0.20 0.97 0.98 0.18
(90
to 180
) w/LBBB
Normal (0
to þ 90
) 7/254 (3%) 21/93 (23%) !0.001 0.23 0.97 0.75 0.77
w/RBBB
QRS concordance
Positive (RBBB-type) 46/254 (18%) 5/93 (5%) !0.005 0.18 0.95 0.90 0.30
Negative (LBBB-type) 26/219 (12%) 4/39 (10%) NS 0.12 0.90 0.87 0.15
Atrioventricular relationship
AV dissociation 145/473 (31%) 0/132 (0%) !0.001 0.31 1.00 1.00 0.29
1:1 AV relationship 38/473 (8%) 69/132 (52%) !0.001 0.08 0.48 0.36 0.13
2:1 VA conduction 19/473 (4%) 0/132 (0%) !0.02 0.04 1.00 1.00 0.22
VA Wenckebach 4/473 (1%) 0/132 (0%) NS 0.01 1.00 1.00 0.22
Atrial fibrillation/flutter 20/473 (4%) 6/132 (5%) NS 0.04 0.95 0.77 0.22
Indeterminate 247/473 (52%) 57/132 (43%) NS 0.52 0.57 0.81 0.25
Dissociated, 2:1 or 168/473 (36%) 0/132 (0%) !0.001 0.36 1.00 1.00 0.30
wenckebach
QRS configuration in V1
VT criteriaa 460/473 (97%) 16/132 (12%) !0.001 0.97 0.88 0.97 0.90
Onset-S Nadir O60 ms 170/219 (77%) 6/39 (15%) !0.001 0.78 0.85 0.97 0.40
w/LBBB
‘‘Rs’’ complex 26/473 (5%) 0/132 (0%) !0.01 0.05 1.00 1.00 0.23
‘‘W’’ complex 26/473 (5%) 0/132 (0%) !0.01 0.05 1.00 1.00 0.23
SVT-A criteriaa 13/473 (3%) 116/132 (88%) !0.001 0.97 0.88 0.97 0.90
QRS configuration in
V6 in RBBB-type
R/S ratio !1 188/254 (74%) 20/94 (24%) !0.001 0.73 0.79 0.90 0.52
RS complex in
precordial leads
RS absent 139/473 (29%) 16/132 (12%) !0.001 0.29 0.88 0.90 0.26
RS present, R-S 241/334 (75%) 9/116 (8%) !0.001 0.72 0.92 0.96 0.54
interval O100 ms
QRS alternans
Present 7/473 (2%) 6/132 (5%) !0.05 0.1 0.95 .054 0.21
Abbreviations: LBBB, left bundle branch block-type tachycardia; NPA, negative predictive accuracy; PPA, positive
predictive accuracy; RBBB, right bundle branch block-type tachycardia; SVT-A, supraventricular tachycardia with
aberration; VT, ventricular tachycardia.
a
See text for details.

with LBBB pattern did so (P ¼ NS; sensi-
tivity 0.12, specificity 0.90). Of note, posi-
tive concordance can be observed in
preexcited SVT because APs activate the
ventricles from base to apex (positive com-
plexes in precordial leads); right ventricular
apical pacing can sometimes produce a neg-
ative concordant pattern.
4) Atrioventricular relationship. In SVT-A, with
extremely rare exceptions, one must have at
least as many atrial complexes as ventricular;
in VT, atrial activation is not necessary
for continuation of the rhythm, and thus
one may observe a variety of non-1:1 AV
relationships in VT. These include complete
dissociation (generally sinus rhythm in the
444
Fig. 3. Concordant precordial QRS patterns in two cases of VT. (A) Positive concordance (all precordial leads positive).
(B) Negative concordance (all precordial leads negative). Note dissociated atrial activity in V1 of each ECG.
atria; Figs. 1 and 3), 2:1 retrograde (VA) con-
duction, and retrograde Wenckebach block.
All of these are best discerned on a long
rhythm strip rather than analyzing individual
ECG leads (2–3 seconds’ worth). ECG leads of
greatest value in finding P waves are 2, 3, aVF,
V1, and aVR. It is important to note that VT
can have 1:1 retrograde conduction, mimick-
ing SVT-A; this is especially problematic in
young patients with VT in the absence of
structural heart disease. Of equal importance
is realizing that atrial activation may be unre-
lated to that of the ventricles, but one cannot
discern this on the ECG either because of
a rapid ventricular rate resulting in too much
QRS complex, ST segment, and T wave to be
able to see distinct P waves, the presence of
atrial activation ‘‘buried’’ in a wide QRS com-
plex, or the presence of atrial fibrillation or
flutter. In our series, 145/473 (31%) of VTs
MILLER et al
showed AV dissociation, 19 (4%) showed 2:1
retrograde conduction, and 4 (1%) had retro-
grade Wenckebach; thus, in 36%, the AV rela-
tionship was diagnostically helpful. On the
other hand, 38 (8%) of VTs showed 1:1 retro-
grade conduction, atrial fibrillation was evi-
dent in 20 (4%), and in 247 (52%), the AV
relationship could not be determined. Thus,
in over half of VT cases, this criterion was
not helpful. However, among available crite-
ria for distinguishing SVT-A from VT, all
but the AV relationship depend on having
a relatively normal baseline ECG (which is
not available for comparison in the vast ma-
jority of cases). Of note, not all cases of
SVT-A show 1:1 conduction: surprisingly,
this was clearly present in only 69/132 (52%)
of SVT-A cases. In 6 (5%), atrial flutter was
evident, while in 57 (47%), atrial activity
could not clearly be discerned (occurring
 VALUE OF THE 12-LEAD ECG IN THE WIDE QRS TACHYCARDIA 445

simultaneously with the wide QRS or other-
wise obscured).
5) Fusion and capture complexes. Fused QRS
complexes are those in which the QRS is
a blend of two sources of activation
(Fig. 4). This is typically seen in VT when
an atrial complex encounters an AV node
and HPS that have recovered from refracto-
riness related to the prior QRS and can acti-
vate some of the ventricular muscle over
a portion of the HPS, while the next VT com-
plex is also occurring. Rarely, a complete
supraventricular capture complex occurs
(all ventricular muscle activated from HPS).
Understandably, these phenomena depend
on (1) a non-1:1 AV relationship during VT
and (2) a relatively slow VT rate; in the ab-
sence of these, the AV node would be main-
tained in refractoriness, and no atrial
activations could penetrate the AV conduc-
tion system. Fusion complexes can occur
during SVT-A (a premature ventricular com-
plex occurring during SVT), but this is ex-
ceedingly rare. Although fusion/capture
complexes are practically diagnostic of VT,
they are very uncommon; only two VTs in
our series showed this phenomenon (0.5%).
6) Specific patterns in leads V1 and V6. Config-
urations of the QRS in V1 and V6 can aid in
distinguishing VT from SVT-A (also based
on the principle that only certain patterns
are compatible with aberration); examples
of leads V1 and V6 from each diagnostic cat-
egory are shown in Fig. 2 and more exten-
sively in Fig. 5. These morphologic criteria
are as follows:
a. V1 in RBBB-type QRS. Normally, initial
ventricular activation occurs independent
of the right bundle branch; thus, RBBB
aberration should only affect the latter
portion of the QRS, not its onset. In
accordance with this principle, several pat-
terns have been recognized as consistent
with aberration (rR0 , rsR0 , rSr0 , rSR0 ) [1],
and others inconsistent therewith, indicat-
ing VT instead (qR, Rsr0 , monophasic R
wave). Within our series, 116/132 (88%)
SVT-As had a V1 configuration consis-
tent with one of the listed aberration pat-
terns compared with 13/473 (3%) VTs
(P ! 0.001; sensitivity 0.97, specificity
0.88). However, this criterion can occa-
sionally be difficult to apply because the
end of the T wave may distort the onset
of the QRS in V1, simulating or obscuring
a Q wave. Similarly, atrial activity (espe-
cially atrial flutter or fibrillation) may dis-
tort the initial portion of V1.
b. V1 in LBBB-type QRS. In true LBBB ab-
erration, the initial portion of the QRS in
V1 shows rapid activation, with an R
wave duration (if present) %30 millisec-
onds and interval from QRS onset to
S-wave nadir %70 milliseconds [4]. This
pattern was observed in 33/39 (85%) of
LBBB-type SVT-A ECGs, but only 49/
219 (22%) 170/176 of LBBB-type VT
ECGs. In contrast, a broad initial R
wave (O30 milliseconds) or longer inter-
val from QRS onset to S-nadir (O70 mil-
liseconds) is not compatible with typical
aberration, and thus WQRST with this
pattern is more likely VT; 170/176 (97%)
of ECGs with such a pattern were VT
(P ! 0.001; sensitivity 0.78, specificity
0.85, positive predictive accuracy 0.97).
In addition, notching or slurring off the
QRS complex suggests myocardial dis-
ease, the presence of which suggests
a higher likelihood of VT.
c. Other specific patterns in V1. In cases of
a tall R/small S (‘‘Rs’’) as well as so-called

Fig. 4. Fusion and capture complexes. A rhythm strip of lead V1 is shown in a patient with RBBB-type VT; black dots
indicated dissociated atrial activity (white circle in center of dots indicates a probable P wave hidden within a QRS com-
plex). A fortuitously timed P wave (black dot with vertical line) can conduct and depolarize some of the ventricles (ar-
rowheads), or rarely (black dot with white cross), completely capture the ventricles (arrow).
446 MILLER et al

Fig. 5. Typical QRS configurations in SVT with aberration and VT in both LBBB and RBBB-type complexes. Note that
with LBBB aberration, initial deflections are relatively rapid, whereas they occur much more slowly in VT. See text for
further clarification.

‘‘W’’ configuration in V1 (examples in With RBBB aberration, this small contri-

Fig. 6), VT is the overwhelming diagnosis;
in our series, all cases (26 [5%] ‘‘Rs,’’ 26
[5%] ‘‘W’’) with either of these patterns
were VT (specificity 1.0). It is difficult to
achieve either of these patterns with any
type of aberration.
d. V6 in RBBB-type QRS. Normally, the
right ventricle’s relatively small muscle
mass contributes a small deflection in its
direction, toward V1 and away from V6.
bution to the QRS is shifted later to ap-
pear as a small S wave. QRS patterns
consistent with aberration are thus qRs
or Rs. In contrast, because practically all
RBBB-type VTs arise in the left ventricle,
all of the right ventricular voltage plus
some left ventricular voltage travels away
from V6, yielding a different set of QRS
patterns (qRS, qrS, rS, QS). This is the or-
igin of the familiar R/S !1 rule (suggests

Fig. 6. Uncommon but highly specific configurations of lead V1 found only in VT. (Top) The ‘‘Rs’’ configuration;
(bottom) the ‘‘W’’ configuration. Neither is compatible with standard aberration.
 VALUE OF THE 12-LEAD ECG IN THE WIDE QRS TACHYCARDIA 447

Fig. 7. Explanation for R/S ratio in V6 rule. In each panel, a heart diagram (atria, AV node/His bundle/bundle
branches) and right (RV) and left ventricles (LV) is shown above representative ECGs of V1 and V6 in normal conduc-
tion, RBBB aberration, and RBBB-type VT. At the left of each ECG panel, the relative contribution of each ventricle to
the ECG complex is depicted. The normal RV has a smaller mass and contributes less voltage than the LV. In normal
conduction, the RV and LV components occur simultaneously, resulting in a narrow QRS. In RBBB aberration, the RV
component is shifted later, resulting in the standard small terminal R in V1 and small terminal S in V6. In RBBB-type
VT, all of the RV plus some of the LV voltage proceeds toward V1 and away from V6, resulting in a tall R wave in V1
and large S wave in V6 (R/S ratio !1).

VT; see Fig. 7). Among cases in our series
showing RBBB-type QRS in which an RS
complex was present in V6, 188/254
(74%) of VT ECGs had an R/S ratio !1,
whereas 20/94 (24%) of SVT-A ECGs
showed this. In 188/208 (90%) of ECGs
with an R/S ratio !1, the correct diagnosis
was VT (P ! 0.001; sensitivity 0.73, speci-
ficity 0.79, positive predictive accuracy 0.9).
e. V6 in LBBB-type QRS. In typical LBBB
aberration, V6 usually shows a monophasic
R wave with a slow upstroke. This is, un-
fortunately, quite common in VT as well;
however, a qR or QS should not be seen
in V6 with true LBBB aberration and these
strongly suggest VT when present.
7) Absent RS in precordial leads [10]. In most
cases of either RBBB or LBBB aberration,
at least one precordial lead has an ‘‘RS’’ con-
figuration, and the interval from R wave on-
set to S wave nadir is %100 milliseconds; in
a WQRST in which none of the precordial
leads has an RS complex, or in which the
R–S interval exceeds 100 milliseconds, aber-
ration is unlikely to be the cause, and
therefore, VT is the likely diagnosis
(Fig. 3A and Fig. 8). In the original series in-
corporating these criteria, the ‘‘absent RS’’
feature was not frequently present (only
15% of VTs, but no SVTs); adding the R–S
interval O100 milliseconds criterion allowed
a diagnosis in almost half of their cases (AV
dissociation and traditional patterns in V1
and V6 were also used in the algorithm). In
our series, an RS complex was absent in
140/473 (30%) VT ECGs as opposed to 9/
132 (8%) SVT-A ECGs. Thus, 140/156
(90%) ECGs in our series showed no RS
complex in the precordial leads (positive pre-
dictive accuracy 0.9). The utility of the R-S
interval was somewhat less than the origina-
tors or the criterion found (sensitivity 0.99,
specificity 0.97). As with the criteria for
lead V1 noted above, other features of the
ECG (particularly T waves) can distort the
QRS complex to either simulated or obscure
a small Q wave, the presence of which would
make the criterion ‘‘negative.’’
8) Miscellaneous features. Rightward inferior
axis deviation in LBBB SVT-A is extremely
rare (1/39 LBBB SVTs [3%] in our series);
448
Fig. 8. ‘‘Absent RS’’ complexes in precordial leads. In each case, no precordial lead has an ‘‘RS’’ complex (hence, ‘‘ab-
sent RS’’). Although considered to be highly suggestive of VT, the bottom panel shows a LBBB SVT tachycardia with no
RS complexes in precordial leads.
this combination should suggest VT as the
WQRST diagnosis (present in 44/219 [20%]
of LBBB-type VTs in our series [P ! 0.01,
specificity 0.97]). In addition, RBBB-type
VT rarely has a normal axis (0
to þ90
);
this combination should likewise suggest
SVT-A (combination was present in 21/93
[23%] of RBBB-type SVT-As in our series
versus 7/254 [3%] RBBB-type VTs;
[P ! 0.01, specificity 0.97]).
9) QRS alternans (Fig. 9). Beat-to-beat alterna-
tion of the QRS amplitude is often associated
with narrow-complex orthodromic SVT in-
corporating an accessory AV pathway; alter-
nans of R0.1 mV can also be seen in
WQRST, whether SVT-A or VT. In one
study of WQRST, alternans was present in
about equal frequency in SVT-A and VT
MILLER et al
(35%) but more leads showed alternans in
SVT-A (mean seven leads) than in VT
(mean three leads) [11]. In our series, alter-
nans was far less common; only 15 cases
showed this feature, with a disproportionate
number in the SVT-A group (6/132 [5%])
as opposed to the VT group (7/473 [2%];
P ! 0.05, sensitivity 0.05, specificity 0.99).
It is not clear why the findings in the current
series differ so greatly from the prior one.
The presence of alternans did not appear to
have any relationship to tachycardia cycle
length, and of interest, patients with multiple
WQRSTs typically showed alternans in only
one. Further, two episodes of the same
tachycardia did not necessarily both show al-
ternans. Reasons for these peculiar features
are not known.
 VALUE OF THE 12-LEAD ECG IN THE WIDE QRS TACHYCARDIA
Fig. 9. QRS alternans during VT. Voltage changes in alternate complexes are noted in all limb leads as well as V1 and to
a lesser extent, all other leads except V3 and V4 (alternans is also noted in Fig. 8A).
The preceding criteria can be applied when
only the WQRST ECG is available; as health
information systems become more integrated, it
should be possible to readily access a WQRST
patient’s prior ECGs for comparison. In such
cases, the following additional criteria can be
applied [2,5]:
1) Identical QRS configuration between baseline
and WQRST [2,12,13]. In cases in which the
WQRST complexes are identical to those of
the baseline ECG, it is very likely that the
WQRST is SVT. Only very rare cases of
VT show this similarity; the one important
exception is bundle branch reentrant VT, in
which both the resting ECG and VT have
a LBBB pattern [14,15].
2) Contralateral BBB patterns in baseline versus
WQRST ECGs. If a patient has resting
RBBB in sinus rhythm and has an episode
of LBBB-type tachycardia, the tachycardia
is very unlikely to be SVT with LBBB aberra-
tion, because (at least in principle) both con-
duction pathways would be blocked and one
should see only P waves. However, it is clear
that some cases of apparent BBB are instead
bundle branch delay (especially in a diseased
left bundle branch), and thus, it is not uni-
formly true that contralateral BBB in resting
rhythm versus WQRST is diagnostic of VT.
3) WQRST complexes narrower than baseline
ECG [14]. In this situation, the baseline
ECG must have a very wide BBB pattern,
with a narrower QRS complex during tachy-
cardia caused by a septal tachycardia origin
(activating both ventricles nearly simulta-
neously) or early engagement of the HPS.
Because of these considerations, this unusual
449
phenomenon suggests VT as the cause of the
WQRST.
It is worth noting in passing that rate
and regularity of the rhythm are not helpful
discriminators between SVT-A and VT. A ven-
tricular rate of 150 bpm should prompt one to
think of atrial flutter, but otherwise, the degree of
overlap in ventricular rate among WQRSTs is
such that it has no differential diagnostic value.
An irregular WQRST is almost always atrial
fibrillation; other SVTs and VT are only rarely
irregular (enough R-R variation to be evident on
the surface ECGdusually O40 milliseconds).
Some authors have attempted to use combina-
tions of ECG criteria to yield a higher proportion
of correct diagnoses. In one such effort, Griffith
and colleagues [16] found that a history of myo-
cardial infarction, QRS morphology in aVF and
V1, and frontal plane axis deviation O40
differ-
ent from the baseline ECG correctly identified
90% of WQRSTs. This algorithm relies on clinical
history and prior ECG, one or both of which may
not be available.
Application of criteria
Correct application of established diagnostic
criteria remains a significant problem in manage-
ment of patients with WQRST. As an example of
this, two cardiologists and two emergency physi-
cians were asked to apply the algorithm by
Brugada and colleagues [17] to 157 WQRSTs;
the sensitivity and specificity of diagnoses were
0.85–0.91 and 0.55–0.60, respectively, for the car-
diologists, while the emergency physicians’ num-
bers were 0.79–0.83 and 0.43–0.70, respectively.
450 MILLER
These contrast to the sensitivity (0.987) and spec-
ificity (0.965) in the original paper by the algo-
rithm’s developers. This disparity does not call
into question either the validity of the algorithm
or its designers’ honesty; expert ECG interpreters
who are focused on the task of carefully applying
an algorithm would naturally be expected to yield
better results than those who are less practiced in
this exercise. These results do illustrate that even
relatively simple and straightforward criteria for
differentiating among causes of WQRST often
fall short in their ‘‘real world’’ application.
Time and again, instead of intelligently apply-
ing valid ECG criteria to a WQRST patient
(perhaps because of the difficulty in either recall-
ing or correctly applying them), clinicians fall
back on the appearance of the patient to make
their diagnosis: if he has a good blood pressure
and mental status, the WQRST must be SVT.
Inappropriate treatment based on incorrect di-
agnoses has led to disastrous results [18,19]. How-
ever, if one recalls nothing else about diagnosing
the cause of WQRST, in most clinical settings
a simple guess of VT will be correct O70% of
the time [20]. A corollary to this is that if there
is doubt as to the diagnosis of WQRST, it is safest
to treat as VT.
Summary
A wealth of useful diagnostic criteria is avail-
able to assist the health care worker in arriving at
the correct diagnosis in cases of WQRST. Despite
the abundance of good criteria for determining the
diagnosis in cases of WQRST, they are of no use if
they cannot be readily applied in an urgent clinical
situation because they cannot be easily recalled or
are too complex and cumbersome to use. It may be
that refresher courses in the differential diagnosis
of WQRST, especially for emergency physicians
who are often the ‘‘first responders’’ to patients
with WQRST, can improve physicians’ diagnostic
accuracy in this important disorder.
References
[1] Wellens HJ, Bär FW, Lie KI. The value of the elec-
trocardiogram in the differential diagnosis of
a tachycardia with a widened QRS complex. Am
J Med 1978;64:27–33.
[2] Dongas J, Lehmann MH, Mahmud R, et al. Value of
preexisting bundle branch block in the electrocardio-
graphic differentiation of supraventricular from
et al
ventricular origin of wide QRS tachycardia. Am
J Cardiol 1985;55(6):717–21.
[3] Reddy GV, Leghari RU. Standard limb lead QRS
concordance during wide QRS tachycardia. A new
surface ECG sign of ventricular tachycardia. Chest
1987;92(4):763–5.
[4] Kindwall KE, Brown J, Josephson ME. Electro-
cardiographic criteria for ventricular tachycardia
in wide complex left bundle branch block mor-
phology tachycardias. Am J Cardiol 1988;61(15):
1279–83.
[5] Kremers MS, Black WH, Wells PJ, et al. Effect of
preexisting bundle branch block on the electrocar-
diographic diagnosis of ventricular tachycardia.
Am J Cardiol 1988;62(17):1208–12.
[6] Marriott HJ. Differential diagnosis of supraventric-
ular and ventricular tachycardia. Cardiology 1990;
77(3):209–20.
[7] Antunes E, Brugada J, Steurer G, et al. The differen-
tial diagnosis of a regular tachycardia with a wide
QRS complex on the 12-lead ECG: ventricular
tachycardia, supraventricular tachycardia with aber-
rant intraventricular conduction, and supraventricu-
lar tachycardia with anterograde conduction over an
accessory pathway. Pacing Clin Electrophysiol 1994;
17(9):1515–24.
[8] Alberca T, Almendral J, Sanz P, et al. Evaluation
of the specificity of morphological electrocardio-
graphic criteria for the differential diagnosis of
wide QRS complex tachycardia in patients with in-
traventricular conduction defects. Circulation
1997;96(10):3527–33.
[9] Baerman JM, Morady F, DiCarlo LA Jr, et al. Dif-
ferentiation of ventricular tachycardia from supra-
ventricular tachycardia with aberration: value of
the clinical history. Ann Emerg Med 1987;16(1):
40–3.
[10] Brugada P, Brugada J, Mont L, et al. A new ap-
proach to the differential diagnosis of a regular
tachycardia with a wide QRS complex. Circulation
1991;83(5):1649–59.
[11] Kremers MS, Miller JM, Josephson ME. Elec-
trical alternans in wide complex tachycardias. Am
J Cardiol 1985;56(4):305–8.
[12] Halperin BD, Kron J, Cutler JE, et al. Misdiagnos-
ing ventricular tachycardia in patients with under-
lying conduction disease and similar sinus and
tachycardia morphologies. West J Med 1990;
152(6):677–82.
[13] Olshansky B. Ventricular tachycardia masquerading
as supraventricular tachycardia: a wolf in sheep’s
clothing. J Electrocardiol 1988;21(4):377–84.
[14] Miller JM. The many manifestations of ventricular
tachycardia. J Cardiovasc Electrophys 1992;3:88–107.
[15] Oreto G, Smeets JL, Rodriguez LM, et al. Wide
complex tachycardia with atrioventricular dissocia-
tion and QRS morphology identical to that of sinus
rhythm: a manifestation of bundle branch reentry.
Heart 1996;76(6):541–7.
 VALUE OF THE 12-LEAD ECG IN THE WIDE QRS TACHYCARDIA
[16] Griffith MJ, De Belder MA, Linker NJ, et al. Multi-
variate analysis to simplify the differential diagnosis
of broad complex tachycardia. Br Heart J 1991;
66(2):166–74.
[17] Isenhour JL, Craig S, Gibbs M, et al. Wide-complex
tachycardia: continued evaluation of diagnostic cri-
teria. Acad Emerg Med 2000;7(7):769–73.
[18] Stewart RB, Bardy GH, Greene HL. Wide com-
plex tachycardia: misdiagnosis and outcome after
451
emergent therapy. Ann Intern Med 1986;104(6):
766–71.
[19] Buxton AE, Marchlinski FE, Doherty JU, et al. Haz-
ards of intravenous verapamil for sustained ventricu-
lar tachycardia. Am J Cardiol 1987;59(12):1107–10.
[20] Steinman RT, Herrera C, Schuger CD, et al. Wide
QRS tachycardia in the conscious adult. Ventricular
tachycardia is the most frequent cause. JAMA 1989;
261(7):1013–6.
 Cardiol Clin 24 (2006) 453–469
Electrocardiographic Markers of Sudden Death
Peter Ott, MD*, Frank I. Marcus, MD
Sarver Heart Center, University of Arizona Health Sciences Center, 1501 N. Campbell Avenue,
Tucson, AZ 85724, USA
Sudden cardiac death (SCD) is defined as abrupt
loss of consciousness due to cardiac arrest in
a person who was previously in a stable condition.
That most of these events are due to ventricular
arrhythmias is supported by data from (a) fortu-
itous Holter ECG recordings of patients at the time
of sudden death, (b) electrocardiographic record-
ings at the scene of a patient in cardiac arrest, and
(c) stored electrogram data from patients with an
implantable cardioverter defibrillator (ICD). How-
ever, it can be difficult to know the actual cause of
death even in patients who die suddenly. For
example, autopsy studies in patients with known
heart disease and at high risk for ventricular
arrhythmias showed that sudden death events
were due to nonarrhythmic or noncardiac death
mechanisms in 7 of 17 sudden death victims [1].
Sudden death accounts for O60% of all
cardiac deaths, and remains a significant public
health problem. The SCD rates range from 270/
100,000 in female to 410/100,000 in male adults
O35 years of age. For each gender group the rate
is higher in African Americans than in the White
population. Each year, approximately 400,000 to
450,000 patients die form this condition in the
United States alone [2]. Citing retrospective data,
Myerberg and colleagues [3] have drawn attention
to the fact that, although the risk for SCD is
greatest in those patients with significant struc-
tural heart disease, these high-risk patients con-
tribute a relatively low number of total SCD
victims. The greatest number of SCD occurs in
a population with minimal or no prior history of
heart disease. In those patients, SCD is the first
presentation of heart disease.
* Corresponding author.
E-mail address: ottp@email.arizona.edu (P. Ott).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.03.004
A prospective study in The Netherlands [4] col-
lected clinical data in patients suffering sudden
death over a 3-year period. A total of 2030 deaths
occurred in inhabitants age 25 to 75; of those, 375
(18%) were sudden deaths. A history of prior
heart disease was absent in half of all sudden
death victims, and less the half of those with
a prior myocardial infarction (MI) had significant
left ventricular (LV) dysfunction (ejection fraction
[EF] !30%). These prospective data confirm that
the majority of sudden death victims cannot be
identified before the event.
During the last 10 years, the ICD has been
shown to be effective in terminating sustained
ventricular arrhythmias, thus aborting sudden
arrhythmic death. Several clinical trials [5,6]
have shown that ICD therapy, when added to op-
timal medical therapy, significantly improves sur-
vival in selected patients with decreased LV
function. To significantly decrease the total num-
ber of SCD events, risk stratification is needed to
identify patients at risk from a large pool of adults
with no or minimal overt symptoms of heart dis-
ease, because it is this group that contributes the
largest number of SCD victims.
Sudden death in the general population
Autopsy studies of victims of SCD show
a healed MI in more than one half of these cases,
even if there is no prior history of coronary artery
disease. Significant coronary artery disease is
present in 70% to 80% of SCD victims, while
10% to 15% have a dilated cardiomyopathy
(CMP) [7,8]. Other abnormalities seen are ventric-
ular hypertrophy, inflammation, and infiltration.
In clinical reports structural abnormalities were
absent in up to 10% of SCD survivors [9,10].
cardiology.theclinics.com
454 OTT & MARCUS
A routine 12-lead ECG may identify a prior MI
by showing typical Q-wave infarct pattern, or it
may show increased R-wave voltage, suggesting
primary or secondary ventricular hypertrophy.
Nonspecific ST/T wave changes with intraventric-
ular conduction delay may suggest the presence of
a possible CMP. These ECG abnormalities raise
the clinical suspicion of structural heart disease, but
they are not specific predictors for sudden death,
because the incidence of SCD is relatively low even
in the presence of such structural heart disease.
Sudden cardiac death in patients with coronary
artery disease
Acute ischemia, during an MI, leads to major
metabolic and cellular abnormalities, which may
result in ventricular fibrillation (VF) and SCD
(Fig. 1). This can be the first manifestation of cor-
onary artery disease (CAD); VF due to acute is-
chemia is a major cause of SCD in the general
population [11].
The typical Q-wave infarct pattern on the 12-
lead ECG persists indefinitely in O80% of pa-
tients but disappears after months to years in
approximately 15% of patients. In 15% to 40% of
infarcts the classical Q-wave pattern does not
develop (non-Q-wave infarction).
The presence of a remote or recent MI is not
specific for predicting future SCD events. In the
Fig. 1. Three-lead Holter ECG recording (25 mm/s, 10 mm/mV) from a 65-year-old male with CAD and normal LV
function. He died suddenly while wearing the Holter recorder. Note the ST segment elevation, indicating myocardial
ischemia, preceding the onset of VF.
era of aggressive reperfusion therapy for MI, the
incidence of sudden death early after MI is low. In
the GISSI-2 trial [12], 8600 patients were followed
after thrombolytic therapy for acute MI. During
a 6-month follow-up, the total mortality rate
was only 3%, with one third of these deaths being
classified as SCD. Although the presence of ven-
tricular arrhythmias was a predictor of both sud-
den death (relative risk [RR] 2.2) and total
mortality (RR 1.4), the absolute risk of SCD
was low. In another trial [13], in which 1400 MI
survivors with frequent nonsustained ventricular
arrhythmias were randomized to placebo or amio-
darone therapy, the annual sudden death rate was
only 3.5% over a 2-year follow-up in either group,
and was unaffected by amiodarone therapy. Thus,
given the low absolute risk for SCD in most pa-
tients post-MI, it remains difficult to attempt to
reliably identify those at high risk for SCD.
In a recent ICD trial [5], the benefit for ICD
therapy seemed to be greatest in those patients
with a prolonged QRS duration (O120 millisec-
onds) on the baseline 12-lead ECG. This raised
the question as to the predictive value of QRS
duration with regard to arrhythmic death.
Zimetbaum and colleagues [14] reviewed electro-
cardiographic data of patients from the MUSTT
trial. In this trial, 1638 patients with chronic ische-
mic heart disease, who had a reduced LVEF
(!40%) and nonsustained V, and who were not
inducible during electrophysiologic study, were
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 455

followed in a registry for 5 years. Indeed, electro- mutations in genes encoding for proteins of the
cardiographic findings of left bundle branch block beta-myosin heavy chain, myosin binding protein
(LBBB) (but not right bundle branch block C and cardiac troponin-T.
[RBBB]) or LV hypertrophy (LVH) were associ- Patients may present with dyspnea due to
ated with increased risk (hazard ratio of 1.49 diastolic LV dysfunction, palpitations due to
and 1.35) for arrhythmic death. atrial dysrhythmias or syncope, and sudden death
There is no reliable ECG marker that is both due to ventricular arrhythmias. A hallmark of this
sensitive and specific to predict SCD in patients disease is ventricular hypertrophy, which may be
with CAD. The combination of prior MI and regionally pronounced such as in the septum or at
reduced LVEF identified patients that benefit the LV apex. The main role of the 12-lead ECG is
form an ICD as primary prophylaxis for SCD. to provide a clue for the presence of this condition
by showing clinically unexpected signs of LVH
with or without precordial voltage and ST/T wave
Sudden cardiac death and congestive heart failure changes (Fig. 2). These abnormalities can be very
striking [21]. Up to one third of patients may have
Despite optimal medical therapy, the mortality
abnormal Q waves, mostly in the anterolateral
rates are high (15–35% at 2 years) for patients
leads (I, aVL, V4–V6), often mimicking MI. These
with symptomatic heart failure, and one third to
Q waves are thought to be related to septal hyper-
two thirds of these deaths are sudden [15,16]. Al-
trophy and abnormal septal forces at the onset of
though clinical data show that most of these
ventricular depolarization. Left atrial enlargement
events are due to ventricular arrhythmias, a subset
and atrial arrhythmias are common [22]. The as-
of these patients with severe heart failure have
sociation of hypertrophic CMP (HCMP) and ac-
sudden death due to bradyarrhythmias and elec-
cessory pathways remains poorly defined, and
tromechanical dissociation [17]. In patients with
atrioventricular (AV) conduction disturbances
chronic systolic LV dysfunction, the 12-lead
are uncommon.
ECG may show evidence of a prior MI or nonspe-
Based on data from tertiary centers, the annual
cific changes: 25% to 50% have an intraventricu-
mortality rate is 3% to 6% [23]. However, recent
lar conduction delay or BBB pattern, and many
data from nontertiary centers, representing com-
show nonspecific ST/T wave changes [18]. Atrial
munity cohorts of patients, not influenced by re-
and ventricular arrhythmias are frequently seen.
ferral bias, report a 1% annual mortality rate
However, there are no specific 12-lead ECG
[24]. Except for accidental death, HCMP is the
markers that would predict the risk for sudden
most common cause of SCD in young people in-
death. In particular, patients with chronic heart
cluding athletes [25]. VF is the mechanism of sud-
failure frequently have premature ventricular
den death that has been supported by stored
beats (PVCs) and/or nonsustained ventricular
electrogram data from ICDs. Several clinical
tachycardia (VT), which may be seen on routine
markers have been found to identify high-risk pa-
ECG. However, while asymptomatic nonsus-
tients: history of prior cardiac arrest or sustained
tained ventricular arrhythmias predict total mor-
VT, family history of sudden death, unexplained
tality, they have not been shown to be an
syncope, decreased blood pressure response to ex-
independent specific predictor of SCD [19].
ercise, severe LVH (O30 mm). Of note, the QRS
There is no reliable ECG marker that is both
voltage on a 12-lead ECG correlates poorly with
sensitive and specific to predict SCD in patients
echocardiographic LVH [26]. Electrocardio-
with congestive heart failure (CHF). The combi-
graphic evidence of progressive LVH, however,
nation of prior CHF and impaired LV systolic
portends a poor overall prognosis [27]. The apical
function identified patients that benefit form an
form of HCMP [28], which may show impressive
ICD as primary prophylaxis for SCD [6].
T-wave inversions in the precordial leads, has
a low risk for sudden death.
Sudden cardiac death in hypertrophic
cardiomyopathy
Sudden cardiac death and arrhythmogenic right
This is an autosomal dominant disease with
ventricular dysplasia/cardiomyopathy
variable penetrance, affecting specific proteins of
the cardiac myocyte contractile apparatus [20]. The pathologic hallmark of arrhythmogenic
Over one half of genotyped patients show right ventricular dysplasia/cardiopmyopathy
456 OTT & MARCUS

Fig. 2. A 12-lead ECG (25 mm/s; 10 mm/mV) from a 19-year-old male with HCMP and a family history of sudden
death. Note the increased S-wave voltage in leads V1 to V5 consistent with LVH. Diffuse ST/T wave changes are
also present. This patient’s LV wall thickness measured 28 mm on echocardiography.

(ARVD/C) is patchy, fibrofatty replacement of Repolarization abnormalities, such as precor-

myocytes altering myocardial depolarization and
repolarization. Patients may present with PVCs or
nonsustained or sustained VT, which may result
in syncope. These arrhythmias typically have
a LBBB morphology [29]. Some individuals may
have cardiac arrest as a result of ventricular ar-
rhythmias. In endemic areas for ARVD/C, au-
topsy studies suggest that up to 20% of
unexpected deaths in individuals younger than
35 years may be due to ARVD/C [30]. This condi-
tion is genetically determined, and a familial his-
tory of syncope premature sudden death can be
found in a sizeable minority of patients. Genetic
studies have revealed both autosomal dominant
and autosomal recessive transmission patterns.
Eight genetic loci have been identified that pri-
marily affect genes encoding for cell junction pro-
teins such as desmosomes as well as the gene
encoding the cardiac ryanodine receptor, suggest-
ing altered intracellular myocyte calcium handling
[31,32].
Important clinical clues for the diagnosis are
abnormal findings in the 12-lead ECG in in-
dividuals with apparently normal hearts: (1)
T-wave inversion in V1 to V3, (2) QRS duration
R110 milliseconds in V1 to V3, (3) presence of
epsilon waves. These ECG findings are included
as diagnostic criteria [33].
dial T-wave inversion in leads V1 to V3 can be
seen frequently in normal children. Although
T-wave inversion in V1 is also seen frequently in
young adults of both genders, T-wave inversion in
leads V2 and V3 is not seen in healthy male adults
and only rarely seen in female adults [34]. Thus, in
a young adult, presenting with PVCs or VT of
LBBB pattern, precordial T-wave inversion in
V1 to V3 should raise the suspicion for ARVD/C
(Fig. 3). At least one half of ARVD/C patients
presenting with VT will have precordial T-wave
inversion, and the extent of T-wave inversion
correlates with the degree of RV dysplasia.
Abnormalities in depolarization are common
in patients with ARVD/C. Fontaine [35] reported
that a prolonged QRS duration (O110 millisec-
onds) in lead V1 had a sensitivity of 55% and
specificity of 100% for ARVD/C. About one third
of patients show an epsilon wave: a discrete late
potential after the end of the QRS (Fig. 4). Its rec-
ognition is enhanced by recording the ECG at
double speed (50 mm/s), double scale (20 mm/mV),
and altered filter settings (40 Hz) [36]. A pro-
longed S wave upstroke in V1 to V3 (S wave nadir
to isoelectric baseline: R55 milliseconds) was
identified in 37 of 39 (95%) patients with
ARVD/C. Other ECG features relate to the delay
in RV depolarizationdthe so-called parietal
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 457

Fig. 3. A 12-lead ECG (25 mm/s; 10 mm/mV) from a 39-year old female, who had sustained VT with LBBB pattern. She
was diagnosed with ARVD/C. Note the T-wave inversion in leads V1 to V4.

Fig. 4. A 12-lead ECG (25 mm/s; 10 mm/mV) in a 47-year-old male with ARVD. Note the T-wave inversion in leads V1
and V2 and low-amplitude, low-frequency ‘‘epsilon waves’’ following the QRS in leads V1 and V2.
458 OTT & MARCUS
block: these include a QRS duration in lead V1 to
V3 exceeding QRS duration in V6 and the ratio of
QRS duration V1 to V3/V4 to V6 R1.2.
When evaluating young adults with LBBB/
inferior axis PVCs or VT, the challenge is to
determine the etiology of the arrhythmia. The
differential diagnosis is: RV outflow tract (RVOT)
PVC/VT or ARVD/C. The former tends to be
benign, and is not genetically determined; the
latter may be associated with SCD, and may be
genetically determined. In a recent study [36] eval-
uating ECG features in both clinical conditions,
precordial T-wave inversion in V1 to V3 was pres-
ent in 33 of 39 (85%) patients with ARVD/C but
in none of 28 patients with RVOT PVC/VT. Sim-
ilarly, delayed S-wave upstroke in V1 to V3 was
present in 37 or 39 (95%) of ARVD/C patients
and only in 1 of 28 (7%) RVOT PVC patients.
This suggests that these two ECG markers could
be useful in determine the etiology of the arrhyth-
mia and clinical decision making.
These ECG findings, together with morpho-
logic evaluation of RV, structure and function aid
in the clinical diagnosis of ARVD/C. Risk strat-
ification for sudden death has been evaluated by
two groups [37,38]. In general, low-risk patients
have the following characteristics: (1) upright
T waves in V1 to V6, (2) less QRS and QT disper-
sion, (3) no history of syncope, (4) no LV involve-
ment, or (5) normal or minimal RV involvement.
However, prospective evaluation of these risk fac-
tors alone or in combination has not been done.
Sudden cardiac death and long QT syndrome
In the mid-1950s, Jervell and Lange-Nielson
[39] described the association of an abnormally
long QT interval on the 12-lead ECG and sudden
death in individuals with congenital hearing loss.
This was found to be an autosomal recessive con-
dition. Later, a familial long QT syndrome
(LQTS) was also recognized in those with normal
hearing and found to be an autosomal dominant
disorder. The disease prevalence is estimated at
1:7000. Patients, often with familial clustering,
show QT interval prolongation on the 12-lead
ECG, and are prone to polymorphic VT, which
may result in syncope or SCD. Structural heart
disease is absent. The normal corrected QT inter-
val (QTc; QT interval corrected for heart rate, Ba-
zett formula) in 578 adults is !430 milliseconds in
men and !450 milliseconds in women. A pro-
longed QTc interval, seen in only 1% of
individuals is R460 milliseconds for men and
R480 milliseconds for woman [40]. QT interval
prolongation, in the absence of structural heart
disease, electrolyte abnormalities, and drugs that
prolong the QT interval is suggestive of LQTS
(for a list see www.QTdrug.org). However, the
measurement of the QTc interval to diagnose the
LQTS reveals overlap with normal values. A re-
cent study [41] examined the QTc intervals of 83
genotyped carriers and 119 noncarriers. The
QTc ranged from 410 to 590 milliseconds (mean
490 milliseconds) for gene carriers and 380 to
470 milliseconds (mean 420 millseconds) for non-
carriers. A diagnostic QTc cutoff of O440 milli-
seconds would have misdiagnosed 22 of 199
(11%) noncarrier individuals, whereas a diagnostic
cutoff of O470 milliseconds would have resulted
in a false negative diagnosis in 40% of male and
20% of female gene carriers. Thus, a diagnostic
scoring system [42] using clinical factors and
ECG features, including QTc interval and
T-wave abnormalities, has been proposed to
estimate the likelihood of LQTS.
Although originally attributed to an ‘‘imbal-
ance’’ in sympathetic cardiac innervations, it is
now established that mutations of genes coding
for various cardiac ion channels are responsible
for this disease. These genetic defects result in
prolongation of the action potential repolariza-
tion and thus an increase in the QT interval on the
12-lead ECG. Genetic analyses have identified
defects in six genes of genotyped patients; 90%
have defects in the genes coding for ion channels
carrying repolarizing potassium currents: the I-Ks
ion channel (LQT1, chromosome 11) (Fig. 5) and
the I-Kr ion channel (LQT2, chromosome 7). A
smaller proportion of patients carry genetic muta-
tions in the gene coding for the sodium channel
(LQT3, chromosome 5). The T-wave morphology
on the 12-lead ECG can provide clues to the un-
derlying genetic defect: broad-based prolonged
T-wave in LQT1, low-amplitude T wave in
LQT2 and late onset T wave in LQT 3 [43]
(Fig. 6). Clinical data has shown gene specific
risk [44] and triggers [45] for arrhythmias.
Cardiac potassium channels are sensitive to
catecholamines, and cardiac events frequently oc-
cur during stress or exertion particular in LQT1.
Thus, several groups have evaluated infusion of
epinephrine on the QT interval and T waves in
patients with LQTS. In one study [46], the infusion
of low-dose epinephrine (0.5 mg/kg/min) resulted in
an increase in the QT interval (þ82 G 34 millisec-
onds) in 19 patients with LQT1 syndrome,
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 459

Fig. 5. A 12-lead ECG (25 mm/s; 10 mm/mV) from a 35-year-old female with a history of recurrent syncope. Note a pro-
longed QT interval (QTc ¼ 540 milliseconds) with a broad-based T wave. Genotype testing in this patient documented
a defect in the I-Ks ion channel gene (LQT1 syndrome). The patient later had syncope while swimming due to docu-
mented VF and survived due to successful ICD therapy.

compared to no change (
7 G 13 milliseconds) in
27 control patients. The same group [47] also de-
scribed epinephrine-induced qualitative T-wave
changes during infusion of low-dose epinephrine;
notched T waves appeared in 75% of LQT2 pa-
tients, but were also seen in 26% and 34% of
LQT1 and controls, respectively. T-wave notching
beyond the peak of the T wave, however, was seen
exclusively in patients with LQT2 syndrome. These
data suggest a role of epinephrine infusion in diag-
nosing LQTS, especially in patients with borderline
electrocardiographic findings.
Risk stratification in patients with suspected
LQTS relies on clinical and family history: syn-
cope, sudden death, and documented torsade de
pointes. Electrocardiographic parameters such as

Fig. 6. ECG tracings in patients with different long QT syndrome genotype. LQT3: late-onset T waves of normal
duration; LQT2: flat low-amplitude T waves; LQT1: early-onset of broad-based T waves. (From Moss AJ, Zareba W,
Benhorin J, et al. ECG T-wave patterns in genetically distinct forms of the hereditary long QT syndrome. Circulation
1995;92:2929–34, with permission.)
460
QT duration and T-wave morphology also appear
to be useful. A recent report [48] on data form 647
genotyped patients, of whom 90% were LQT1
and LQT2, correlate the risk for arrhythmic
events with sex, QTc duration, and genotype. In
a multivariate analysis, a QTc O500 milliseconds
was highly predictive of arrhythmic events in pa-
tients with LQT1 and LQT2. The presence of
T-wave alternans on an ECG has long been recog-
nized as a predictor for arrhythmic events. A re-
cent study [49] from the LQT registry confirmed
a high risk of clinical events in 30 patients with
T-wave alternans (Fig. 7). However, the presence
of T-wave alternans was strongly related to QTc
prolongation, and after adjustment for this, it
was no longer an independent marker for arrhyth-
mic events. T-wave notching [50], independent of
QT duration, may be a predictor for arrhythmic
events. In a small study, T-wave notching, present
in 33 of 53 (62%) LQTS patients was more prev-
alent in 30 of 37 (81%) patients with a history of
syncope or cardiac arrest versus 3 of 16 (195) pa-
tients those without symptoms (P ! 0.001).
Sudden cardiac death and short QT syndrome
Since its original description in 2000 [51],
a small number of cases with familial sudden
death and strikingly short QT interval (220–290
Fig. 7. Six-lead ECG (25 mm/s; 10 mm/mV) from a 40-year-old female with syncope and prolonged QT interval. T-wave
alternans is present, best seen in lead I and lead II.
OTT & MARCUS
milliseconds) have been reported. The QRS com-
plex is normal, and the ST segment is virtually
nonexistent. The T waves are narrow based, tall,
and symmetric (Fig. 8). The clinical spectrum
ranges from aborted sudden death due to VF in
an infant, to recurrent syncope in otherwise
healthy adults [52]. Structural heart disease was
absent, and secondary causes for QT interval
shortening (hyperkalemia, hypercalcemia, and
digitalis therapy) were carefully excluded. Some
patients also had atrial fibrillation. During elec-
trophysiologic evaluation, short atrial and ven-
tricular refractory periods were noted and VF
was easily induced with programmed ventricular
stimulation. Recent genetic studies revealed muta-
tions in genes coding for subunits of cardiac po-
tassium channels, resulting in ‘‘gain of function’’
and thus an enhanced action potential repolariza-
tion [53]. A clear diagnostic cutoff for the QT in-
terval has not yet been established. In a clinical
series of 27 patients [54] the QT interval ranged
from 210 to 320 milliseconds (mean 267 G 33 mil-
liseconds) and the QTc interval ranged from 250
to 340 milliseconds (mean 298 G 21 milliseconds).
Recent data [55] suggests that therapy with quin-
idine (but not sotalol or ibutilide) can prolong
the abnormally short QT and render VF nonindu-
cible. It is not known if this will be a successful
long-term therapy.
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 461

Fig. 8. Twelve-12 lead ECG of a 37-year-old victim of sudden cardiac death. Note: very short QT interval (266 milli-
seconds), normal QRS duration, and absence of ST segment. (From Gussak I. Brugada P, Brugada J, et al. Idiopathic
short QT interval: a new clinical syndrome? Cardiology 2000;94:99–102; with permission.)

Sudden cardiac death and Brugada syndrome been a recent attempt to standardize this diagnos-
tic ECG feature as follows: type I; coved-shaped
In 1992, Brugada and colleagues [56] described
right precordial ST elevation, in more than one
a clinical syndrome consisting of apparently
right precordial lead (V1–V3), of R2 mm fol-
healthy adults presenting with syncope or sudden
lowed by a negative T wave: type II: saddle-
death. Their ECGs showed coved ST segment ele-
shaped ST elevation: and type III: ST elevation
vation and negative T waves in leads V1 and V2
of !2 mm (Fig. 9). Only type I is said to be diag-
(V3) in the absence of RBBB pattern (Fig. 9).
nostic of Brugada syndrome [59]. These ST eleva-
This ECG pattern has become to be known as
tions are sensitive to heart rate and the effects of
the ‘‘Brugada-ECG.’’ These patients were thought
autonomic stimulation [60]. Furthermore, sodium
to have no evidence of structural heart disease,
channel-blocking antiarrhythmic drugs such as
and were at high risk for recurrent syncope or
procainamide or flecainide have been shown to
sudden death. In 15% to 20% of patients a genetic
‘‘provoke’’ these ECG changes in patients with
defect in the cardiac sodium channel was found
otherwise normal baseline ECG, and have been
[57]. It has been proposed that there is a premature
proposed as a screening tool in patients suspected
termination of the epicardial action potential with
to have the Brugada syndrome [61,62]. However,
loss of the phase-2 dome pattern. The ECG
the sensitive, specificity and reproducibility of
changes are believed to be due to regional epicar-
this drug provocation are not well established.
dial–endocardial voltage gradient during the pla-
Of interest, certain gene mutations are sensitive
teau phase of the action potential. This voltage
to temperature with regard to their phenotypic ex-
gradient may give rise to VT due to phase-2 reen-
pression [63]. Indeed, Brugada-type ECG changes
try caused by inhomogeneity of the voltage in ad-
have been seen in patients only during febrile ill-
jacent areas of the myocardium [58]. In patients
ness (Fig. 10). Some speculate that seizures during
with the Brugada syndrome, syncope and sudden
febrile illness may be related to self-terminating
death are due to polymorphic VT. Since the orig-
ventricular arrhythmias due to a temperature sen-
inal description, it has been observed that the
sitive mutation in the sodium channel, giving rise
‘‘Brugada-ECG’’-type ST segment elevation in
to the ‘‘Brugada syndrome.’’
the right precordial leads can be (1) cove shaped,
Patients with resuscitated sudden death or
(2) saddle shaped, (3) transient, and (4) be present
syncope or with a family history of premature
in a multitude of clinical conditions. There has
sudden death and typical type I Brudaga ECG
462 OTT & MARCUS

Fig. 9. Precordial ECG leads (25 mm/s; 10 mm/mV) of different ECG patterns in Brugada syndrome. Type I: coved ST
segment elevation in leads V1 and V2 (and V3) followed by a negative T wave. Type II: saddleback-type ST segment
elevation R2 mm. Type III: saddleback-type ST segment elevation !2 mm. (From Wilde AA, Antzelevitch C,
Borggrefe M, et al. Proposed diagnostic criteria for the Brugada syndrome: consensus report. Circulation 2002;106:
2514–8; with permission.)

changes are at increased risk of recurrent events,
and should be considered for ICD therapy.
Asymptomatic patients who have an unprovoked
typical Brudaga ECG (type I) are also at increased
risk for arrhythmic events, but the magnitude of
the risk is unclear. In one study the event rate over
a mean of 24 months follow-up was 8% in these
patients. Those presenting with typical Brugada
ECG (type I) only after drug exposure or those
presenting with nondiagnostic right precordial
ECG abnormalities (type II, type III) are believed
to be at low risk [59]. The role of electrophysio-
logic study and prognostic value of inducible VF
remain controversial [64–66].
Sudden cardiac death and idiopathic ventricular
fibrillation
A small group of patients with may have VF
but have no demonstrable heart disease and have
a normal 12-lead ECG. In particular, the QT
interval will be normal, there are no changes
suggestive of Brugada syndrome, and there is no
family history of either condition. Minor struc-
tural and electrocardiographic abnormalities are
permitted [67]. This category is likely a heteroge-
neous group of patients, and further clinical and
genetic studies may be able to focus on specific
disorders within this group.
Based on careful analysis of clinical data, Coumel
and his group described two distinct entities:
(1) Short coupled variant of torsade de pointes
[68]: in this description, 14 patients (mean
age 34 G 10 years) of either sex, presented
with syncope (one with resuscitated sudden
death) and one third had a family history
of sudden death. Structural heart disease
was absent and the 12-lead ECG was nor-
mal. In particular, the QT interval was nor-
mal. Close coupled PVCs (coupling interval
to preceding QRS: 245 G 25 milliseconds)
were noted and observed to degenerate
into torsade de pointes (Fig. 11). Over a
7-year follow-up, five patients died (four
suddenly) and nine were alive, three of
whom had an ICD.
(2) Cathecholaminergic polymorphic VT [69]: in
this series, 21 children (mean age 10 G 4
years) of either sex presented with exertional
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 463

Fig. 10. (A) A12-lead ECG (25 mm/s; 10 mm/mV) of a 34-year-old female presenting with high fever (103.4 F). Note the
coved ST elevation (3 mm) in leads V1 and V2, followed by a negative T wave – consistent with type I Brugada ECG
pattern. (B) The ECG completely normalized once the fever resolved. Neither the patient nor her family had any history
of syncope or sudden death.

syncope. Again, there was no structural heart stimulation reproducibly resulted in a pro-
disease and the QT interval was normal. One gressive pattern of polymorphic extrasystole,
third of patients had a family history of bidirectional tachycardia, and polymorphic
syncope or sudden death. Adrenergic VT degenerating into VF (Fig. 12). A recent
464 OTT & MARCUS

Fig. 11. Single-lead ECG strip from a 15-year-old male with syncope. Note: frequent close coupled PVCs, initiating non-
sustained polymorphic VT degenerating into VF. (From Leenhardt A, Glaser E, Burguera M, et al. Short coupled var-
iant of Torsade de Pointes. A new electrocradiographic entity in the spectrum of idiopathic ventricular fribrillation.
Circulation 1994;89:206–15; with permission.)

study [70] in three affected families revealed
mutations of the ryanodine receptor, indicat-
ing abnormalities in intracellular calcium
handling as the mechanism of ventricular
arrhythmias.
In a recent study [71], 18 patients with unex-
plained cardiac arrest and no evident cardiac dis-
ease (normal LV function, coronary arteries, and
resting corrected QTc) underwent pharmacologic
challenges with adrenaline and procainamide

Fig. 12. Single-lead ECG strip during stress testing in a 7-year-old male with a history of exertional syncope. Note the
progressive appearance of ventricular ectopy and bidirectional ventricular tachycardia. These arrhythmias disappeared
with termination of exercise. (From Leenhardt A, Lucet V, Denjoy I, et al. Catecholaminergic polymorphic ventricular
tachycardia in children. A 7-year follow-up of 21 patients. Circulation 1995;91:1512–9; with permission.)
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 465

Fig. 13. Twelve-lead ECG (25 mm/s; 10 mm/mV) from a 35-year-old man with recurrent palpitations and documented
SVT. The ECG shows a classic preexcitation pattern: short PR interval, slurred onset of QRS (delta wave), and increased
QRS duration. Note: pseudo-Q waves in the inferior leads due to negative delta waves.

infusion to unmask subclinical primary electrical
disease. The final diagnose was catecholaminergic
VT in 10 patients (56%), Brugada syndrome in
two patients (11%), and unexplained VF in six
patients (33%).
Sudden cardiac death and Wolff-Parkinson-White
syndrome
The presence of one or more accessory path-
ways spanning the AV groove gives rise to the
typical preexcited pattern of the QRS complex on

Fig. 14. A 12-lead ECG (25 mm/s; 10 mm/mV) from a 17-year-old male with near syncope and palpitations. Note the
fast, irregularly irregular wide and narrow QRS complexes, typical for atrial fibrillation with fast AV conduction via the
accessory pathway. The shortest preexcited RR interval is 200 milliseconds.
466 OTT & MARCUS
the 12-lead ECG: (a) short PR interval, (b) slurred
onset of QRS complex (delta wave), and (c) pro-
longed QRS duration (Fig. 13). The prevalence
of preexcitation on routine ECG is variably
reported between approximately 3/1000 [72] to
5/10,000 [73]. Several algorithms, primarily using
the delta wave polarity on the 12-lead ECG, allow
accurate prediction the location of the accessory
pathway [74]. This is quite useful when planning
radiofrequency catheter ablation to treat patients
with Wolff-Parkionson-White (WPW) syndrome.
The need for possible transseptal puncture (left-
sided accessory pathways) or the risk for AV
block (anteroseptal pathway) can be anticipated.
Although orthodromic AV reentry tachycardia
is the most common tachycardia in patients with
WPW syndrome, these patients may also develop
atrial fibrillation. In atrial fibrillation, the acces-
sory pathway may allow rapid conduction to the
ventricles, which can result in VF and sudden
death. The risk for sudden death in patients with
WPW sysdrome is estimated at 1 per 1000 patient-
year follow-up [75]. The conduction properties of
the accessory pathway are believed to be the prime
determinant of VF induction during atrial fibrilla-
tion (Fig. 14). When compared to WPW patients
with atrial fibrillation but with out sudden death,
WPW with atrial fibrillation and survived sudden
death have been found to have shorter minimum
preexcited RR intervals (180 G 30 milliseconds
versus 240 G 60 milliseconds) and shorter mean
RR intervals (270 G 60 milliseconds versus 340
G 80 milliseconds) during induced atrial fibrilla-
tion [76]. However, a large overlap between these
groups exists and a short (!250 millisecond) pre-
excited RR interval during induced atrial fibrilla-
tion was observed in approximately 17% of
clinically asymptomatic WPW, resulting in a
high false positive rate. Patients with VF are
more likely to have multiple accessory pathways
[77]. Thus, close attention to the preexcitation pat-
tern at rest or during atrial fibrillation might pro-
vide a clue to the presence of more than one
pathway. A clinically useful observation, although
seen in only 5% to 15% of patients, is the loss of
antegrade pathway conduction (loss of preexcita-
tion) on a resting 12-lead ECG or during exercise
stress testing [78]. This finding correlates well with
an accessory pathway refractory period of O300
milliseconds and a mean RR interval (during
atrial fibrillation) of O300 milliseconds. The loss
of preexcitation has to be abrupt, and be associ-
ated with an increase in the PR interval to rule
out enhanced AV nodal conduction. Such findings
characterize a pathway with poor antegrade con-
duction properties, which does not allow rapid
AV conduction during atrial fibrillation; thus,
there is a very low risk for VF and sudden death.
References
[1] Pratt CM, Grennway PS, Schoenfield MH, et al. Ex-
ploration of the precision of classifying sudden car-
diac death. Implications for the interpretation of
clinical trials. Circulation 1996;93:519–24.
[2] Zheng ZJ, Croft JB, Giles WH, et al. Sudden death
in the United States 1989–1998. Circulation 2001;
104:2158–63.
[3] Myerburg RJ, Kessler KM, Castellanos A. Sudden
death: epidemiology, transient risk and intervention
assessment. Ann Intern Med 1993;119:1187–97.
[4] de Vreede-Swagemakers JJ, Gorgels APM, Dubois-
Arbouw WI, et al. Out-of hospital cardiac arrest in
the 1990s: a population based study in the Maas-
tricht area on incidence, characteristics and survival.
J Am Coll Cardiol 1997;30:1500–5.
[5] Moss AJ, Zareba W, Hall J, et al. Prophylactic im-
plantation of a defibrillator in patients with myocar-
dial infarction and reduced ejection fraction. N Engl
J Med 2002;346:877–83.
[6] Bardy GH, Lee KL, Mark DB, et al. Amiodarone or
an implantable cardioverter-defibrillator for conges-
tive heart failure. N Engl J Med 2005;352:225–37.
[7] Kuller L, Cooper M, Perper J. Epidemiology of sud-
den death. Arch Intern Med 1972;129:714–9.
[8] Newman WP, Strong JP, Johnson WD. Community
pathology of atherosclerosis and coronary artery
disease in New Orleans. Morphologic findings in
young black and white men. Lab Invest 1981;44:
496–501.
[9] Kudenchuck PJ, Cobb LA, Greene HL, et al. Late
outcome of survivors of out of hospital cardiac ar-
rest with left ventricular functions O50% and with-
out significant coronary arterial narrowing. Am
J Cardiol 1991;67:704–8.
[10] Freedman RA, Swerdlow CD, Soderholm-Difatte V,
et al. Prognostic significance of arrhythmia induc-
ibility or noninducibility at initial electrophysiologic
study in survivors of cardiac arrest. Am J Cardiol
1988;61:578–82.
[11] Gorgeles APM, Gijsbers C, de Vreede-Swagemakers
J, et al. Out of hospital cardiac arrestdthe relevance
of heart failure: the Maastricht Circulatory Arrest
Registry. Eur Heart J 2003;24:1204–9.
[12] Maggioni AP, Zuanetti G, Franzosi MG, et al. Prev-
alence and prognostic significance of ventricular ar-
rhythmias after acute myocardial infarction in the
fibrinolytic era. GISSI-2 results. Circulation 1993;
87:312–22.
[13] Cairns JA, Connolly SJ, Roberts R, et al. Random-
ized trial of outcome after myocardial infarction in
patients with frequent or repetitive ventricular
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH
premature depolarizations: CAMIAT. Lancet 1997;
349:675–82.
[14] Zimetbaum PJ, Buxton AE, Batsford W, et al. Elec-
trocardiographic predictors of arrhythmic death and
total mortality in the Multicenter Unsustained
Tachycardia Trial. Circulation 2004;110:766–9.
[15] Pitt B, Zannad F, Remme WJ, et al. The effect of spi-
ronolocatone on morbidity and mortality in patients
with severe heart failure. N Engl J Med 1999;341:
709–17.
[16] MERIT-HF study group. Effect of metoprolol CR/
XL in chronic heart failure: metoprolol CR/XL
randomized intervention trial in congestive heart
failure. Lancet 1999;353:2001–7.
[17] Luu M, Stevenson WG, Stevenson LW, et al.
Diverse mechanisms of unexpected cardiac arrest in
advanced heart failure. Circulation 1989;80:1675–80.
[18] Dec GW, Fuster V. Medical progress: idiopathic di-
lated cardiomyopathy. N Engl J Med 1993;331:1564.
[19] Teerlink JR, Jalaluddin M, Anderson S, et al. Am-
bulatory ventricular arrhythmias in patients with
heart failure do not specifically predict in increased
risk of sudden death. Circulation 2000;101:40–6.
[20] Schwartz K, Carrier L, Guicheney P, et al. Molecu-
lar basis for familial cardiomyopathies. Circulation
1995;91:532–40.
[21] Yamaguchi H, Ishimura T, Nishiyama S, et al.
Hypertrophic non-obstructive cardiomyopathy with
giant negative T waves (apical hypertrophy):
ventriculographic and echocardiographic features
in 30 patients. Am J Cardiol 1979;44:401–12.
[22] Frank S, Braunwald E. Idiopathic subaortic ste-
nosis: clinical analysis of 126 patients with em-
phasis on the natural history. Circulation 1968;
37:759.
[23] Maron BJ. Hypertrophic cardiomyopathy: a system-
atic review. JAMA 2002;287:1308–20.
[24] Maron BJ, Casey SA, Poliac LC, et al. Clinical
course of hypertrophic cardiomyopathy in a regional
United States cohort. JAMA 1999;281:650–5.
[25] Maron BJ, Shirani J, Poliac LC, et al. Sudden
death in young competitive athletes. Clinical, de-
mographic and pathological profiles. JAMA
1996;276:199–204.
[26] Montgomery JV, Harris KM, Casey SA, et al. Rela-
tion of electrocardiographic patterns to phenotypic
expression and clinical outcome in hypertrophic car-
diomyopathy. Am J Cardiol 2005;96:270–5.
[27] McKenna WJ, Borggrefe M, England D, et al. The
natural history of left ventricular hypertrophy in
hypertrophic cardiomyopathy: an electrocardio-
graphic study. Circulation 1982;66:1233.
[28] Erickson MJ, Sonnenberg B, Woo A, et al. Long
term outcome in patients with apical hypertrophic
cardiomyopathy. J Am Coll Cardiol 2002;39:
638–45.
[29] Marcus FI, Fontaine G. Arrhythmogenic right ven-
tricular dysplasia/cardiomyopathy: a review. PACE
1995;18:1298–314.
467
[30] Thienne G, Nava A, Corrado D, et al. Right ventric-
ular cardiomyopathy and sudden death in young
people. N Engl J Med 1988;318:129–33.
[31] Rampazzo N, Nava A, Malacrida S, et al. Mutation
in human desmoplakin domain binding to plakoglo-
bin causes a dominant form of arrhythmogenic right
ventricular cardiomyopathy. Am J Hum Genet
2002;71:1200–6.
[32] Tiso N, Stephan DA, Nava A, et al. Identification of
mutation in the cardiac ryanodine receptor gene in
families affected with arrhythmogenic right ventricu-
lar cardiomyopathy type 2 (ARVD2). Hum Mol
Genet 2001;10:189–94.
[33] McKenna WJ, Thienne G, Nava A, et al. Diagno-
sis of arrhythmogenic right ventricular dysplasia/
cardiomyopathy: task force of the working group
myocardial and pericardial disease of the Euro-
pean Society of Cardiology and the Scientific
Council on Cardiomyopathies of the international
society and federation of cardiology. Br Heart
J 1994;71:215–8.
[34] Marcus FI. Prevalence of T wave inversion beyond
V1 in young normal individuals and usefulness for
the diagnosis of arrhythmogenic right ventricular
dysplasia/cardiomyopathy. Am J Cardiol 2005;95:
1070–1.
[35] Fontaine G, Umemura J, DiDonna P, et al. La duree
des complexes QRS dans la dysplasie ventriculaire
droite arythmogene. Un noveau marqueur diagnos-
tique non-invasif. Ann Cardiol Angeiol (Paris) 1993;
42:399–405.
[36] Nasir K, Bomma C, Tandri H, et al. Electrocardio-
graphic features of arrhythmogenic right ventricular
dysplasia/cardiomyopathy according to disease
severity. A need to broaden diagnostic criteria. Cir-
culation 2004;110:1527–34.
[37] Turrini P, Corrado D, Basso C, et al. Noninvasive
risk stratification in arrhythmogenic right ventri-
cular cardiomyopathy. Ann Noninvasive Electro-
cardiol 2003;8:161–9.
[38] Peters S, Peters H, Thierfelder L. Risk stratification
of sudden cardiac death and malignant ventricular
arrhythmias in right ventricular dysplasia-cardiomy-
opathy. Int J Cardiol 1999;71:243–50.
[39] Jervell A, Lange-Nielsen F. Congenital deaf mutism,
functional heart disease with prolonged QT interval
and sudden death. Am Heart J 1957;54:59–68.
[40] Moss AJ, Robinson J. Clincial features of idiopathic
long QT syndrome. Circulation 1992;85(suppl I):
II40–4.
[41] Vincent GM, Timothy KW, Leppert M, et al. The
spectrum of symptoms and QT intervals in carriers
of the gene for the long-QT syndrome. N Engl
J Med 1992;327:846–52.
[42] Schwartz PJ, Moss AJ, Vincent GM, et al. Diagnos-
tic criteria for the long QT syndrome: an update.
Circulation 1993;88:782–4.
[43] Moss AJ, Zareba W, Benhorin J, et al. ECG T-wave
patterns in genetically distinct forms of the
468 OTT & MARCUS
hereditary long QT syndrome. Circulation 1995;92:
2929–34.
[44] Zareba W, Moss AJ, Schwartz PJ, et al. Influence of
the genotype on the clinical course of the long QT
syndrome. N Engl J Med 1998;339:960–5.
[45] Schwartz PJ, Priori SG, Spazzolini C, et al. Geno-
type-phenotype correlation in the long QT syn-
drome. Gene specific triggers for life-threatening
arrhythmias. Circulation 2001;103:89–95.
[46] Ackerman MJ, Khositseth A, Tester D, et al. Epi-
nephrine induced QT interval prolongation:
a gene specific paradoxical response in congenital
long QT syndrome. Majo Clin Proc 2002;77:
413–21.
[47] Khosiseth A, Hejlik J, Shen WK, et al. Epinephrine
induced T wave notching in congenital long QT syn-
drome. Heart Rhythm 2005;2:141–6.
[48] Priori SG, Schwartz PJ, Napolitano C, et al. Risk
stratification in the long-QT syndrome. N Engl
J Med 2003;348:1866–74.
[49] Zareba W, Moss AJ, Vessie C, et al. T wave alter-
nans in idiopathic long QT syndrome. J Am Coll
Cardiol 1994;23:1541–6.
[50] Malfatto G, Beria G, Sala S, Bonazzi O, et al. Quan-
titative analysis of T wave abnormalities and their
prognostic implications in the long QT syndrome.
J Am Coll Cardiol 1994;23:296–301.
[51] Gussak I, Brugada P, Brugada J, et al. Idiopa-
thic short QT interval: a new clinical syndrome?
Cardiology 2000;94:99–102.
[52] Gaita F, Giustetto C, Bianchi F, et al. Short QT syn-
drome. A familial cause of sudden death. Circulation
2003;108:965–70.
[53] Bellocq C, Ginneken GC, Bezzina C, et al. Mutation
in the KCNQ1 Gene leading to the short QT interval
syndrome. Circulation 2004;109:2394–7.
[54] Giusetto C, Wolpert C, Anttonnen OM, et al. Clin-
ical presentation of the patients with short QT syn-
drome. Heart Rhythm 2005;2(II):S61.
[55] Gaita F, Giustetto C, Bianchi F, et al. Short QT syn-
drome: pharmacological treatment. J Am Coll
Cardiol 2004;43:1494–9.
[56] Brugada P, Brugada J. Right bundle branch block
persistent ST segment elevation and sudden death:
a distinct clinical and electrocardiographic syn-
drome. A multicenter report. J Am Coll Cardiol
1992;20:1391–6.
[57] Chen Q, Kirsch GE, Zhang D, et al. Genetic basis
and molecular mechanism for idiopathic ventricular
fibrillation. Nature 1998;392:293–5.
[58] Antzelevitch C. The Brugada syndrome: ionic basis
and arrhythmia mechanisms. J Cardiovasc Electro-
physiol 2001;12:268–72.
[59] Antzelevitch C, Brugada P, Borggrefe M, et al. Bru-
gada syndrome. Report of the second consensus
conference. Heart Rhyhtm 2005;2:429–40.
[60] Miyazaki T, Mitamura H, Miyoshi S, et al. Auto-
nomic and antiarrhythmic drug modulation of ST
segment elevation in patients with Brugada syn-
drome. J Am Coll Cardiol 1996;27:1061–70.
[61] Fujiki A, Usui M, Nagasawa H, et al. ST segment el-
evation in the right precordial leads induced with
class Ic antiarrhythmic drugs: insight into the mech-
anism of Brugada syndrome. J Cardiovasc Electro-
physiol 1999;10:214–8.
[62] Gasparini M, Priori S, Mantica M, et al. Flecainide
test in Brugada syndrome: a reproducible but risky
tool. PACE 2003;26(pt II):338–41.
[63] Dumaine R, Towbin JA, Brugada P, et al. Ionic
mechanism responsible for the electrocardiographic
phenotype of the Brugada syndrome are tempera-
ture dependent. Circ Res 1999;85:803–9.
[64] Priori SG, Napolitano C, Gasparini M, et al. Natu-
ral history of Brugada syndrome. Insights for risk
stratification and management. Circulation 2002;
105:1342–7.
[65] Brugada P, Brugada R, Mont L, et al. Natural his-
tory of Brugada syndrome: the prognostic value of
programmed electrical stimulation of the heart.
J Cardiovasc Electrophysiol 2003;14:455–7.
[66] Sarkozy A, Brugada P. Sudden cardiac death and
inherited arrhythmia syndromes. J Cardiovasc Elec-
trophysiol 2005;16(suppl):S8–20.
[67] Consensus Statement of the Joint Steering Com-
mittees of UCARE and of IVF-US. Survivors
of out-of-hospital cardiac arrest with apparently
normal heart: Need for definition and standard-
ized clinical evaluation. Circulation 1997;95:
265–72.
[68] Leenhardt A, Glaser E, Burguera M, et al. Short
coupled variant of torsade de pointes. A new elec-
trocardiographic entity in the spectrum of idio-
pathic ventricular fibrillation. Circulation 1994;
89:206–15.
[69] Leenhardt A, Lucet V, Denjoy I, et al. Catechol-
aminergic polymorphic ventricular tachycardia in
children. A 7 year follow-up of 21 patients. Circula-
tion 1995;91:1512–9.
[70] Priori SG, Napolitano C, Natascia T, et al. Muta-
tions of the cardiac ryanodine receptor gene
(hRyR2) underlie catecholaminergic polymorphic
ventricular tachycardia. Circulation 2001;103:
196–200.
[71] Krahn AD, Gollob M, Yee R, et al. Diagnosis of
unexplained cardiac arrest: role of adrenaline and
procainamide infusion. Circulation 2005;112:
2228–34.
[72] Krahn AD, Manfreda J, Tate RB. The natural his-
tory of electrocardiographic preexcitation in men.
The Manitoba follow-up study. Ann Intern Med
1992;116:456–60.
[73] Orejerana LA, Vidaillet HJ, DeStefano F. Popula-
tion prevalence of Wolff-Parkinson-White syn-
drome. J Am Coll Cardiol 1995; abstr 327A.
[74] Fitzpatrick AP, Gonzales RP, Lesh MD, et al. New
algorithm for the localization of accessory
 ELECTROCARDIOGRAPHIC MARKERS OF SUDDEN DEATH 469

atrioventricular connections using a baseline electro-
cardiogram. J Am Coll Cardiol 1994;23:107–16.
[75] Klein GJ, Prystowsky EN, Yee R, et al. Asymptom-
atic Wolff-Parkinson-White. Should we intervene?
Circulation 1989;80:1902–5.
[76] Klein GJ, Bashore TM, Sellers TD, et al. Ventricular
fibrillation in the Wolff-Parkinson-White syndrome.
N Engl J Med 1979;301:1080–5.
[77] Pappone C, Santinelli V, Rosanio S, et al. Usefulness
of invasive electrophysiologic testing to stratify the
risk of arrhythmic events in asymptomatic patients
with Wolff-Parkinson-White syndrome. J Am Coll
Cardiol 2003;41:239–44.
[78] Levy S, Broustedt JP. Exercise testing in the Wolff-
Parkinson-White syndrome. Am J Cardiol 1981;48:
976–9.
 Cardiol Clin 24 (2006) 471–490

Diagnostic Value of the 12-Lead Electrocardiogram

during Conventional and Biventricular Pacing
for Cardiac Resynchronization
S. Serge Barold, MDa,*, Michael C. Giudici, MDb,
Bengt Herweg, MDa, Anne B. Curtis, MDa
a
Division of Cardiology, University of South Florida College of Medicine and Tampa General Hospital,
Tampa, FL 33615, USA
b
Division of Cardiology, Genesis Heart Institute, Davenport, IA, USA

The ‘‘low-tech’’ paced 12-lead surface ECG has
fallen into disuse for routine pacemaker evaluation
because it adds expense that is not usually re-
imbursed, and requires an additional piece of
hardware. It is widely believed that the majority
of pacemaker evaluation can be appropriately
performed with a single-channel rhythm strip in
conjunction with ‘‘high-tech’’ ECG/marker sys-
tems of pacemaker programmers. Intellectually,
a 12-lead ECG would be ideal for all pacemaker
evaluations, but most physicians do not have
adequate support staff, time, and resources to do
this when it will not be reimbursed by third-party
payers. It is, however, important for the physician
to recognize when a single-lead evaluation is in-
adequate and when a 12-lead ECG is essential to
perform a proper evaluation [1]. This is especially
important to evaluate patients with biventricular
devices for cardiac resynchronization [2–7].
The pacemaker stimulus
Digital recorders distort pacemaker stimuli and
produce striking changes in amplitude and polar-
ity. Digital recorders can also miss some or all the of
the pacemaker stimulus because of sampling char-
acteristics involving a relatively narrow window.
Some digital recorders process stimuli into an
* Corresponding author.
E-mail address: ssbarold@aol.com (S.S. Barold).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.05.001
artifactual standard-size ECG deflection. Diagnos-
tic evaluation of the pacemaker stimulus was
possible with old analog-writing systems but no
longer with modern digital technology.
Normal QRS patterns during right ventricular
pacing
Pacing from the right ventricle (RV), regard-
less of site, virtually always produces a left bundle
branch block (LBBB) pattern in the precordial
leads (defined as the absence of a positive complex
in lead V1 recorded in the fourth or fifth intercos-
tal space) [2,8,9]. Pacing from the RV apex pro-
duces negative paced QRS complexes in the
inferior leads (II, III, and aVF) because depolar-
ization begins in the inferior part of the heart
and travels superiorly away from the inferior leads
(Fig. 1). The mean paced QRS frontal plane axis
is always superior, usually in the left, or less com-
monly in the right, superior quadrant.
Pacing from the right ventricular outflow tract
Primary lead placement in the right ventricular
outflow tract (RVOT), septum, or lead displace-
ment from the RV apex toward the RVOT shifts
the frontal plane paced QRS axis to the left
inferior quadrant, a site considered normal for
spontaneous QRS complexes [8]. The inferior
leads become positive. The axis then shifts to the
right inferior quadrant as the stimulation site
moves more superiorly toward the pulmonary
cardiology.theclinics.com
472 BAROLD et al

Fig. 1. Diagrammatic representation of the 12-lead ECG during apical RV pacing. (Reproduced from Barold SS,
Stroobandt RX, Sinnaeve AF. Cardiac pacemakers step by step. An illustrated guide. Malden [MA]: Blackwell-Futura;
2004, with permission.)

valve (Fig. 2). With the backdrop of dominant R
waves in the inferior leads, RVOT pacing may
generate qR, QR, or Qr complexes in leads I
and aVL (Fig. 3) [8]. Occasionally with slight dis-
placement of the pacing lead from RV apex to the
RV outflow tract, leads I and aVL may register
a qR complex in conjunction with the typical neg-
ative complexes of RV apical stimulation in the
inferior leads (Figs. 4 and 5). This qR pattern in
leads I and aVL must not be interpreted as
a sign of myocardial infarction [10].
qR and Qr complexes in inferior
and precordial leads
RV pacing from any site never produces qR
complexes in V5 and V6 in the absence of
myocardial infarction or ventricular fusion with

Fig. 2. Twelve-lead ECG during pacing the RV outflow tract. The frontal plane axis points to the left inferior quadrant,
and there is a LBBB pattern in the precordial leads because there is no dominant R wave in lead V1.
 DIAGNOSTIC VALUE OF THE 12-LEAD ECG IN CARDIAC PACING 473

Dominant R wave of the paced QRS complex

QS
qR
Fig. 3. Diagram showing the difference between a QS
complex and a qR complex.
a spontaneous conducted QRS complex. A qR or
Qr (but not QS, as shown in Fig. 3) complex in the
precordial or inferior leads is always abnormal
during RV pacing from any site in the absence
of ventricular fusion (Fig. 6). In contrast, a q
wave is common in the lateral leads (I, aVL, V5,
and V6) during uncomplicated biventricular pac-
ing (using the RV apex), and should not be inter-
preted as representing myocardial infarction or
RVOT displacement of an RV apical lead. On
the other hand, uncomplicated RV apical pacing
rarely displays a qR complex in lead I (but not
aVL) [2,8].
in lead V1 during conventional right
ventricular apical pacing
A dominant R wave in V1 during RV pacing
has been called a right bundle branch block
(RBBB) pattern of depolarization, but this termi-
nology is potentially misleading because this pat-
tern may not be related to RV activation delay
(Box 1). In our experience, a dominant R wave
of a paced ventricular beat in the right precordial
leads (V1 and V2 recorded in the fourth intercostal
space) occurs in approximately 8% to 10% of pa-
tients with uncomplicated RV apical pacing
[8,9,11,12]. The position of precordial leads V1
and V2 should be checked, because a dominant
R wave can sometimes be recorded at the level
of the third or second intercostal space during un-
complicated RV apical pacing (Figs. 7 and 8). The
pacing lead is almost certainly in the RV (apex or
distal septal site) if leads V1 and V2 show a nega-
tive QRS complex when recorded one space lower
(fifth intercostal space). However, a dominant R
wave may not be always eliminated at the level
of the fifth intercostal space if RV pacing origi-
nates from the midseptal region [13].

Fig. 4. Lead displacement from RV apex to mid-septal pacing simulating anterior myocardial infarction. ECG during
VVI pacing in a patient with idiopathic cardiomyopathy, atrial fibrillation, complete heart block, and normal coronary
arteries. After implantation of a pacemaker lead at the RV apex, the ECG showed a dominant QRS complex in leads I
and aVL without a q wave. Three days after implantation this ECG shows qR complexes only in leads I and aVL, with
major negativity in the inferior leads. The pacing lead was slightly displaced upwards toward the outflow tract as shown
in Fig. 5. (Reproduced from Barold SS. Complications and follow-up of cardiac pacemakers. In: Singer I, Kupersmith,
editors. Clinical manual of electrophysiology. Baltimore [MD]: Williams & Wilkins; 1993, with permission.)
474 BAROLD et al

conduction. However, left ventricular (LV) pacing

Fig. 5. Same patient as Fig. 4. Lateral chest X-ray
showing displacement of the pacing lead toward the
RV outflow tract.
Furthermore, in the normal situation with the
ventricular lead in the RV, the ‘‘RBBB’’ pattern
from pacing RV sites results in a vector change
from positive to negative by lead V3 in the precor-
dial sequence. Therefore, a tall R wave in V3 and
V4 signifies that a pacemaker lead is most proba-
bly not in the RV after excluding ventricular
fusion from spontaneous atrioventricular
generating a positive complex in lead V1 may not
necessarily be accompanied by a positive complex
in leads V2 and V3. The ECG pattern with a truly
posterior RV lead has not been systematically in-
vestigated as a potential cause of a tall R wave in
V1 during RV pacing.
We have never seen a so-called RBBB pattern
in lead V1 during uncomplicated RV outflow tract
pacing, and it has never been reported so far.
Right axis deviation of the ventricular paced beats
in the frontal plane with a deep S wave in leads I
and aVL does not constitute a RBBB pattern
without looking at lead V1.
Significance of a small r wave in lead V1
during uncomplicated right
ventricular pacing
A small early (r) wave (sometimes wide) may
occasionally occur in lead V1 during uncompli-
cated RV apical or outflow tract pacing. There
is no evidence that this r wave represents a conduc-
tion abnormality at the RV exit site. Furthermore,
an initial r wave during biventricular pacing does
not predict initial LV activation [1].
Left ventricular endocardial pacing
Passage of a pacing lead into the LV rather
than the RV occurs usually via an atrial septal
defect (patent foramen ovale) or less commonly

Fig. 6. Twelve-lead ECG showing ventricular fusion of pacemaker-induced ventricular depolarization with the native
QRS complex generated by spontaneous atrioventricular conduction. The fusion QRS complex is narrow. Note the
QR complexes in leads II, III, aVF, V5, and V6. The pattern simulates myocardial infarction during DDD pacing
(AS-VP) in a patient with sick sinus syndrome, relatively normal AV conduction, and no evidence of coronary artery
disease. The spontaneous ECG showed a normal QRS pattern.
 DIAGNOSTIC VALUE OF THE
Box 1. Causes of a dominant R wave
in lead V1 during conventional
ventricular pacing

Ventricular fusion.

Pacing in the myocardial relative
refractory period.

Left ventricular pacing from the
coronary venous system.

Left ventricular endocardial or
epicardial pacing.

Lead perforation of the right ventricle
or ventricular septum with left
ventricular stimulation.

Uncomplicated right ventricular pacing
(Lead V1 recorded too high or in the
correct place)
via the subclavian artery [13–28]. The diagnosis of
a malpositioned endocardial LV lead will be
missed in a single-lead ECG. The problem may
be compounded if the radiographic malposition
of the lead is not obvious or insufficient
Fig. 7. Diagram showing evaluation of a dominant R
wave in lead V1 during uncomplicated RV pacing.
When a dominant R wave occurs when V1 is recorded
one or two ICS too high, a negative QRS complex will
often be recorded in the fourth ICS which is the correct
site for V1. If the dominant R wave persists or is initially
recorded in the fourth ICS, a negative QRS complex will
be recorded one ICS lower in the fifth ICS. (Reproduced
from Barold SS. Complications and follow-up of cardiac
pacemakers. In: Singer I, Kupersmith J, editors. Clinical
manual of electrophysiology. Baltimore [MD]: Williams
& Wilkins; 1993, with permission.)
12-LEAD ECG IN CARDIAC PACING 475
projections are taken. A 12-lead paced ECG will
show an RBBB pattern of paced ventricular depo-
larization, commonly with preserved QRS positiv-
ity in the right precordial leads or at least V1
(Fig. 9). The positive QRS complexes are unal-
tered when leads V1 and V2 are recorded one in-
tercostal space lower. During LV pacing, the
frontal plane axis of paced beats can indicate the
site of LV pacing, but as a rule, with a RBBB con-
figuration the frontal plane axis cannot differenti-
ate precisely an endocardial LV site from one in
the coronary venous system. The diagnosis of an
endocardial LV lead is easy with transesophageal
echocardiography. In the usual situation, it will
show the lead crossing the atrial septum then pass-
ing through the mitral valve into the LV. An en-
docardial LV lead is a potential source of
cerebral emboli. Most patients with neurologic
manifestations do not exhibit echocardiographic
evidence of thrombus on the pacing lead. In symp-
tomatic patients, removal of the lead after a period
of anticoagulation should be considered. A
chronic LV lead in asymptomatic or frail elderly
patients is sometimes best treated with long-term
anticoagulant therapy.
A Medline search from the years 2000 to 2005
revealed a substantial number of reports docu-
menting inadvertent endocardial LV lead place-
ment (pacing and implantable cardioverter-
defibrillator leads). The true incidence of this
problem is unknown, but the Medline data
suggest that there are probably many unreported
cases. It is disturbing that this serious but avoid-
able complication (simply by looking at a 12-lead
ECG at the time of implantation) is still being
recognized at late follow-up when lead extraction
can be problematic.
ECG patterns recorded during LV pacing
from the coronary venous system
An RBBB pattern in a correctly positioned lead
V1 occurs with few exceptions when the stimulation
site is in the posterior or posterolateral vein
(Fig. 10) [2–5,7,9,29,30]. Leads V2 and V3 may or
may not be positive. With apical sites, leads V4 to
V6 are typically negative. With basal locations,
leads V4 to V6 are usually positive, as with the con-
cordant positive R waves during overt preexcita-
tion in left-sided accessory pathway conduction in
the Wolff-Parkinson-White syndrome [3]. Pacing
from the middle cardiac vein or the great (anterior)
vein may occasionally produce a LBBB pattern, de-
pending on the site of stimulation [31].
476 BAROLD et al

Fig. 8. Dominant R wave in lead V1 during uncomplicated RV pacing. (A) Lead V1 shows a dominant R wave in the
third and fourth intercostal spaces (ICS) but a negative QRS complex is recorded one space lower in the fifth ICS. (B)
Dominant R wave in lead V1 in ECG when leads V1 and V2 were recorded in the second ICS. The dominant R wave in
V1 disappeared when V1 and V2 were recorded in the fourth ICS. (C) Tall R wave recorded in lead V1 placed too high
and its disappearance when V1 was recorded in the correct fourth ICS. (Reproduced from Barold SS, Falkoff MD, Ong
LS, et al. Electrocardiographic diagnosis of pacemaker malfunction. In: Wellens HJJ, Kulbertus HE, editors. What’s
new in electrocardiography? The Hague [The Netherlands]: Martinus Nijhoff; 1981, with permission. With permission
from Springer Science and Business Media.)

LV pacing from the traditional site for re-
synchronization produces a RBBB pattern in lead
V1 virtually without exception. When lead V1
shows a negative QRS complex during LV pacing,
one should consider incorrect ECG lead place-
ment (lead V1 too high as in Fig. 11), location in
the middle or great (anterior) cardiac vein, or an
undefined mechanism requiring elucidation. The
frontal plane axis often points to the right inferior
quadrant (right axis deviation) and less commonly
to the right superior quadrant. In an occasional
patient with uncomplicated LV pacing with a typ-
ical RBBB pattern in lead V1, the axis may point
to the left inferior or left superior quadrant. The

Fig. 9. ECG of a patient who received a dual-chamber pacemaker for sick sinus syndrome. Three years after pacemaker
implantation, he presented with several transient ischemic attacks (TIAs). An ECG taken when the pacemaker was pro-
grammed to the VVI mode showed ventricular paced beats with a dominant R wave in leads V1 to V3 and right axis
deviation in the frontal plane. There was a ventricular fusion beat in leads V4 to V6. A transesophgeal echocardiogram
confirmed the position of the ventricular lead in the LR passing from the right atrium to the left atrium and crossing the
mitral valve. The lead was successfully extracted percutaneously without complications using a modified technique to
prevent embolization. A new lead in the RV produced ventricular-paced beats with the typical left bundle branch pattern
and left superior axis deviation in the frontal plane.
 DIAGNOSTIC VALUE OF THE 12-LEAD ECG IN CARDIAC PACING 477

Fig. 10. Twelve-lead ECG showing monochamber LV pacing from the coronary venous system. There is typical right
bundle branch pattern and right axis deviation. Note the dominant R wave from V1 to V6 consistent with basal LV
pacing. LV pacing shown in all the subsequent figures was performed from the coronary venous system. (Reproduced
from Barold SS, Herweg B, Giudici M. Electrocardiographic follow-up of biventricular pacemakers. Ann Noninvasive
Electrocardiol 2005;10:231–55; with permission.)

reasons for these unusual axis locations are
unclear.
ECG patterns and follow-up of biventricular
pacemakers
So far, evaluation of the overall ECG patterns
of biventricular pacing has focused mostly on
simultaneous RV and LV stimulation [2,5,7,32–
35]. A baseline 12-lead ECG should be recorded
at the time of implantation during assessment of
the independent capture thresholds of the RV
and LV to identify the specific morphology of
the paced QRS complexes in a multiplicity of
leads [1]. This requires having the patient con-
nected to a multichannel 12-lead ECG during
the implantation procedure. A total of four 12-
lead ECGs are required. (1) Intrinsic rhythm
and QRS complex before any pacing. (2) Paced
QRS associated with RV pacing. (3) Paced QRS

Fig. 11. (A) ECG recorded during LV pacing with leads V1 and V2 recorded at the level of the second intercostal space in
a patient with a thin patient with an elongated chest during LV pacing. There is no dominant R wave in lead V1. The
ECG during biventricular pacing also failed to show a dominant R wave in V1 at the level of the second intercostals
space. (B) The dominant R wave in V1 becomes evident only when lead V1 is recorded in the fourth ICS. The R
wave in V1 recorded in the fourth ICS during biventricular pacing also became dominant. (Reproduced from Barold
SS, Herweg B, Giudici M. Electrocardiographic follow-up of biventricular pacemakers. Ann Noninvasive Electrocardiol
2005;10:231–255; with permission.)
478
Table 1
Change in frontal plane axis of paced QRS when programming from biventricular to monochamber LV and RV pacing
Pacing site QRS in lead I
BiV / RV Greater positivity
BiV / LV Greater negativity
a
QRS in lead III is more negative than in lead II.
Abbreviations: BiV, biventricular, LV, left ventricle; RV, right ventricle.
associated with LV pacing. (4) Paced QRS associ-
ated with biventricular pacing. The four tracings
should be examined to identify the lead configura-
tion that best demonstrates a discernible and obvi-
ous difference between the four pacing states
(inhibited, RV only, LV only, and biventricular).
This ECG lead should then be used as the surface
monitoring lead for subsequent evaluations. Loss
of capture in one ventricle will cause a change in
the morphology of ventricular paced beats in the
12-lead ECG similar to that of either single cham-
ber RV pacing or single-chamber LV pacing. A
shift in the frontal plane axis may be useful to cor-
roborate loss of capture in one of the ventricles
[2,3,6,7]. If both the native QRS and the biventric-
ular paced complex are relatively narrow, then
a widening of the paced QRS complex will iden-
tify loss of capture in one chamber with effectual
capture in the other.
Paced QRS duration and status of mechanical
ventricular resynchronization
The paced QRS during biventricular pacing is
often narrower than that of monochamber RV or
LV pacing. Thus, measurement QRS duration
during follow-up is helpful in the analysis of
appropriate biventricular capture and fusion
with the spontaneous QRS [3,4,7]. If the biventric-
ular ECG is virtually similar to that recorded with
RV or LV pacing alone and no cause is found, one
should not automatically conclude that one of the
leads does not contribute to biventricular depolar-
ization without a detailed evaluation of the pacing
system.
Chronic studies have shown that the degree of
narrowing of the paced QRS duration is a poor
predictor of the mechanical cardiac resynchroni-
zation response [7,36,37]. In other words, the de-
gree of QRS narrowing or its absence does not
correlate with the long-term hemodynamic benefit
of biventricular pacing [7,36,37], because the paced
QRS does not reflect the underlying level of me-
chanical dyssynchrony. In this respect some pa-
tients with monochamber LV pacing exhibit an
BAROLD et al
QRS in lead III Axis shift
a
Greater negativity Clockwise
Greater positivity Counterclockwise
equal or superior degree of mechanical resynchro-
nization compared with biventricular pacing de-
spite a very wide-paced QRS complex [7,36,38].
Usefulness of the frontal plane axis
of the paced QRS complex
Table 1 and Fig. 12 show the importance of the
frontal plane axis of the paced QRS complex in de-
termining the arrangement of pacing during testing
of biventricular pacemakers [2,5,7]. The shift in the
frontal plane QRS axis during programming the
ventricular output is helpful in determining the
site of ventricular stimulation in patients with
first-generation devices without separately pro-
grammable RV and LV outputs (see Table 1).
Biventricular pacing with the right ventricular
lead located at the apex
The frontal plane QRS axis usually moves
superiorly from the left (RV apical pacing) to
the right superior quadrant (biventricular pacing)
in an anticlockwise fashion if the ventricular mass
is predominantly depolarized by the LV pacing
lead (Figs. 12 and 13) [2,3,6,7]. The frontal plane
axis may occasionally reside in the left superior
rather than the right superior quadrant during un-
complicated biventricular pacing.
The QRS is often positive in lead V1 during bi-
ventricular pacing when the RV is paced from the
apex (Fig. 13). A negative QRS complex in lead
V1 may occur under the following circumstances:
incorrect placement of lead V1 (too high on the
chest), lack of LV capture, LV lead displacement,
or marked latency (exit block or delay from the
stimulation site, an important but poorly studied
phenomenon with LV pacing) associated with
LV stimulation, ventricular fusion with the con-
ducted QRS complex, coronary venous pacing
via the middle cardiac vein (also the anterior car-
diac vein), or even unintended placement of two
leads in the RV [39]. A negative QRS complex
in lead V1 during uncomplicated biventricular
pacing probably reflects different activation of
an heterogeneous biventricular substrate (ische-
mia, scar, His-Purkinje participation in view of
 DIAGNOSTIC VALUE OF THE 12-LEAD ECG IN CARDIAC PACING 479

Fig. 12. Diagram showing the usual direction of the mean frontal plane axis during apical RV pacing, RV outflow tract
pacing, LV pacing from the coronary venous system, and biventricular pacing with LV from the coronary venous system
þ RV from the apex. The axis during biventricular pacing from the LV from the coronary sinus þ RV outflow tract
usually points to the right inferior quadrant (right axis) as with monochamber LV pacing. (Reproduced from Barold
SS, Stroobandt RX, Sinnaeve AF. Cardiac pacemakers step by step. An illustrated guide. Malden [MA]: Blackwell-
Futura; 2004; with permission.)

the varying patterns of LV activation in spontane-
ous LBBB, and so on), and does not necessarily
indicate a poor (electrical or mechanical) contri-
bution from LV stimulation.
Biventricular pacing with the right ventricular
lead in the outflow tract
In our limited experience we have found that
during biventricular pacing with the RV lead in
the outflow tract, the paced QRS in lead V1 is of-
ten negative and the frontal plane paced QRS axis
is often directed to the right inferior quadrant
(right axis deviation) (Fig. 14). Further studies
are required to confirm these preliminary findings
and to determine the significance of these ECG
patterns of biventricular pacing according to the
RV pacing site.
Q, or q and QS configuration in lead 1
Georger and colleagues [34] observed a q wave
in lead I in 17 of 18 patients during biventricular
pacing (Fig. 15). As indicated previously, a q wave

Fig. 13. ECG during biventricular pacing with the RV lead at the apex. There is a dominant R wave is V1 and a right
superior axis in the frontal plane. The QRS complex is relatively more narrow (170 milliseconds) than during single-
chamber RV or LV pacing. (Reproduced from Barold SS, Herweg B, Giudici M. Electrocardiographic follow-up of
biventricular pacemakers. Ann Noninvasive Electrocardiol 2005;10:231–55; with permission.)
480 BAROLD et al

Fig. 14. Biventricular pacing with the RV lead in the outflow tract. There was a very prominent R wave in lead V1 dur-
ing monochamber LV pacing. Note the typical absence of a dominant R wave in lead V1, and the presence of right axis
deviation, an uncommon finding during biventricular pacing with the RV lead at the apex. The presence of ventricular
fusion with the spontaneous conducted QRS complex was ruled out. (Reproduced from Barold SS, Herweg B, Giudici M.
Electrocardiographic follow-up of biventricular pacemakers. Ann Noninvasive Electrocardiol 2005;10:231–55; with
permission.)

in lead I during uncomplicated RV apical pacing

is rare, and these workers observed it in only one
patient. Loss of the q wave in lead I was 100%
predictive of loss of LV capture [34]. It therefore
appears that analysis of the Q/q wave or a QS
complex in lead I may be a reliable way to assess
LV capture during biventricular pacing.

Ventricular fusion beats with native conduction

Fig. 15. Uncomplicated biventricular pacing (RV lead
at the apex) in a patient with nonischemic cardiomyop-
athy. The interventricular (V-V) interval is 40 millisec-
onds with LV activation first. The six-lead ECG shows
a Qr complex in lead I and a QR complex in lead
aVL. This pattern does not indicate an old myocardial
infarction. The frontal plane axis lies in the right
superior quadrant as expected with this pacing
arrangement. Bottom: magnified lead aVR showing
separate RV and LV stimuli.
In patients with sinus rhythm and a relatively
short PR interval, ventricular fusion with com-
peting native conduction during biventricular
pacing may cause misinterpretation of the ECG,
and a common pitfall in device follow-up (Fig. 16)
[2,7]. QRS shortening mandates exclusion of ven-
tricular fusion with the spontaneous QRS com-
plex, especially in the setting of a relatively short
PR interval. The presence of ventricular fusion
should be ruled out by observing the paced QRS
morphology during progressive shortening of the
AS-VP (atrial sensing-ventricular pacing) interval
in the VDD mode or the AP-VP (atrial pacing-
ventricular pacing) interval in the DDD mode.
The AS-VP interval should be programmed to en-
sure pure biventricular pacing under circum-
stances that might shorten the PR interval such
as increased circulating catecholamines. It is im-
portant to remember that a very narrow paced
QRS complex may represent ventricular fusion
(possibly associated with a suboptimal hemody-
namic response) with the conducted QRS complex
rather than near-perfect electrical ventricular re-
synchronization. In this respect, remarkable
 DIAGNOSTIC VALUE OF THE
narrowing of the paced QRS complex occurs with
triventricular pacing (two RV sites þ LV), advo-
cated by the French group for heart failure
patients who have become refractory to conven-
tional biventricular pacing [40].
Long-term ECG changes
Many studies have shown that the paced QRS
duration does not vary over time as long as the
LV pacing lead does not move from its initial site
[7,36,41]. Yet, surface ECGs should be per-
formed periodically because the LV lead may be-
come displaced into a collateral branch of the
coronary sinus. Dislodgement of the LV lead
may result in loss of LV capture with the ECG
showing an RV pacing QRS pattern with an in-
creased QRS duration and superior axis devia-
tion. Ricci and colleagues [41] suggested that
variation of the QRS duration over time may
play a determinant role if correlated with remod-
eling of the ventricles by echocardiography. Fi-
nally, the underlying spontaneous ECG should
be exposed periodically to confirm the presence
of a LBBB type of intraventricular conduction
abnormality. In this respect, turning off the pace-
maker could potentially improve LV function
and heart failure in patients who have lost their
intraventricular conduction delay or block
through ventricular remodeling. In other words,
a spontaneous narrow QRS is better than biven-
tricular pacing.
Anodal stimulation in biventricular pacemakers
Although anodal capture may occur with high
output traditional bipolar RV pacing, this phe-
nomenon is almost always not discernible electro-
cardiographically. Biventricular pacing systems
may use a unipolar lead for LV pacing via
a coronary vein. The tip electrode of the LV
lead is the cathode, and the proximal electrode of
the bipolar RV lead often provides the anode for
LV pacing. This arrangement creates a common
anode for RV and LV pacing. A high current
density (from two sources) at the common anode
during biventricular pacing may cause anodal
capture manifested as a paced QRS complex
with a somewhat different configuration from
that derived from pure biventricular pacing
(Fig. 17) [42,43]. Anodal capture during biventric-
ular pacing disappears by reducing the LV output
of the pacemaker or when the device (even at high
output) is programmed to a true unipolar system
with the common anode on the pacemaker can,
12-LEAD ECG IN CARDIAC PACING 481
a function that may not be available in devices
with an ICD. Anodal capture was recognized in
first-generation transvenous biventricular pace-
makers (without separately programmable RV
and LV outputs) when three distinct pacing mor-
phologies were observed exclusive of fusion with
the spontaneous QRS complex: Biventricular
with anodal capture (at a high output), biventricu-
lar (at a lower output), and RV (with loss of LV
capture) or rarely LV (with loss of RV capture).
This form of anodal stimulation may also occur
during biventricular pacing with contemporary de-
vices [44,45] only if there is a common anode on the
RV lead.
A different form of anodal capture involving
the ring electrode of the bipolar RV lead can also
occur with contemporary biventricular pace-
makers with separately programmable ventricular
outputs (Fig. 18). During monochamber LV pac-
ing at a relatively high output, RV anodal capture
produces a paced QRS complex identical to that
registered with biventricular pacing. Occasionally,
this type of anodal capture prevents electrocardio-
graphic documentation of pure LV pacing if the
LV pacing threshold is higher than that of RV an-
odal stimulation. Such anodal stimulation may
complicate threshold testing, and should not be
misinterpreted as pacemaker malfunction. Fur-
thermore, if the LV threshold is not too high, ap-
propriate programming of the pacemaker output
should eliminate anodal stimulation in most cases.
The use of true bipolar LV leads eliminates all
forms of RV anodal stimulation.
Effect of interventricular V-V timing
on the electrocardiogram of biventricular
pacemakers
The electrocardiographic consequences of tem-
porally different RV and LV activation with
programmable V-V timing in the latest biventric-
ular devices have not yet been studied in detail
(Fig. 15). Contemporary biventricular devices per-
mit programming of the interventricular interval
usually in steps from þ80 milliseconds (LV first)
to 80 milliseconds (RV first) to optimize LV he-
modynamics. In the absence of anodal stimula-
tion, increasing the V-V interval gradually to 80
milliseconds (LV first) will progressively increase
the duration of the paced QRS complex, alter its
morphology with a larger R wave in lead V1, indi-
cating more dominant LV depolarization [46].
The varying QRS configuration in lead V1 with
482 BAROLD et al

Fig. 16. (A) Twelve-lead ECGs showing ventricular fusion. Narrowing of the paced QRS complex (well seen in V1) due
to ventricular fusion with the spontaneous conducted QRS complex. This ECG was the initial recording taken upon
arrival to the pacemaker follow-up center. AV delay ¼ 100 milliseconds. The marked narrowing of the QRS complex
in lead V1 suggests ventricular fusion rather than QRS narrowing from satisfactory biventricular pacing. (B) The
ECG taken 15 minutes later (same parameters and AV delay) when the patient was more relaxed shows no evidence
of ventricular fusion. (C) Immediately after the tracing in (B), ventricular fusion was demonstrated only when the
AV delay was lengthened to 130 milliseconds. The serial tracings illustrate the dynamic nature of AV conduction
(emotion, catecholamines, and so on) and the importance of appropriate programming of the AV delay to prevent
ventricular fusion with the spontaneous conducted QRS complex. (Reproduced from Barold SS, Herweg B, Giudici
M. Electrocardiographic follow-up of biventricular pacemakers. Ann Noninvasive Electrocardiol 2005;10:231–55,
with permission.)
 DIAGNOSTIC VALUE OF THE 12-LEAD ECG IN CARDIAC PACING 483

Fig. 17. Anodal capture during first-generation biventricular pacing. There is anodal capture on the left (three pacing
sites). It disappears on the right (two pacing sites) with reduction of the common ventricular output revealing pure
biventricular pacing. (Reproduced from Garrigue S, Barold SS Clémenty J. Electrocardiography of multisite ventricular
pacing. In: Barold SS, Mugica J, editors. The fifth decade of cardiac pacing. Elmsford [NY]: Blackwell-Futura; 2004;
with permission.)

different V-V intervals has not yet been correlated
with the hemodynamic response.
RV anodal stimulation during biventricular
pacing interferes with a programmed interventric-
ular (V-V) delay (often programmed with the LV
preceding the RV) aimed at optimizing cardiac
resynchronization because RV anodal capture
causes simultaneous RV and LV activation (The
V-V interval becomes zero). In the presence of
anodal stimulation, the ECG morphology and its
duration will not change if the device is pro-
grammed with V-V intervals of 80, 60, and 40
milliseconds (LV before RV). The delayed RV
cathodal output (80, 60, and 40 milliseconds) then
falls in the myocardial refractory period initiated
by the preceding anodal stimulation. At V-V
intervals %20 milliseconds, the paced QRS may
change because the short LV-RV interval prevents
propagation of activation from the site of RV
anodal capture in time to render the cathodal site
refractory. Thus, the cathode also captures the
RV and contributes to RV depolarization, which
then takes place from two sites: RV anode, and
RV cathode [46].

Fig. 18. Anodal pacing (on the left) in the DDD mode of a second-generation biventricular pacemaker seen during
monochamber LV pacing with an LV output of 3.5 V and 0.5-millisecond LV. The ECG pattern was identical to
that recorded during biventricular pacing. On the right, anodal stimulation alternates with pure LV pacing when the
LV output is slightly less than 3.5 V at 0.5 milliseconds. (Reproduced from Herweg B, Barold SS. Anodal capture
with second-generation biventricular cardioverter-defibrillator. Acta Cardiol 2003;58:435–6; with permission.)
484
Underlying arrhythmias
The presence of chronic atrial fibrillation is
often poorly recognized in a continuously paced
rhythm with resultant denial of anticoagulant
therapy [47–49]. In a study by Patel and colleagues
[47], atrial fibrillation was correctly identified in
only 20% of patients with continuous pacing and
in 40% with intermittent pacing. A single-lead is
clearly unreliable for the diagnosis. A routine 12-
lead ECG can provide the diagnosis of underlying
atrial fibrillation in most patients, and reprogram-
ming the pacemaker is usually unnecessary in the
presence of a continuously paced rhythm.
Hypertrophic obstructive cardiomyopathy
Selection of the longest AV delay that achieves
complete ventricular capture by QRS criteria may
not be hemodynamically optimal during pacing
for obstructive hypertrophic cardiomyopathy. In
searching for a shorter more favorable AV delay,
a 12-lead ECG is essential to verify the absence of
ventricular fusion [50,51]. Lack of improvement
may require repositioning of the RV lead with
echocardiography to ensure a distal apical posi-
tion because fluoroscopy and QRS morphology
in the 12-lead ECG during pacing are imprecise
markers of lead position.
Pacemaker alternans
Pacemaker QRS alternans is characterized by
alternate changes in paced QRS morphology [52].
Fig. 19. Three-lead ECG (II, V1, and V5) showing pacemaker alternans in a patient with sinus rhythm, 2:1 AV block,
and a DDD pacemaker. Every alternate beat is a ventricular fusion beat (pacing þ spontaneous QRS complex).
BAROLD et al
The causes include respiratory fluctuation, pericar-
dial effusion, mechanical pulsus alternans leading
to electrical alternans from varying depolarization,
and true alternating intraventricular alternans
from the pacing site. Alternans can only be diag-
nosed after ventricular bigeminy or alternate fusion
beats (Fig. 19) are ruled out by changing the pacing
rate. True alternans represents a form of exit block
(persisting over a range of rates) seen only in severe
myocardial disease under circumstances that also
cause latency and second-degree Wenckebach exit
block from the pacing site.
Atrial capture and resynchronization
Evaluation of atrial capture can be difficult in
a single-lead ECG. In conventional dual-chamber
pacing systems, the configuration, duration, and
delay (and degree of atrial latency) of the paced P
waves should be carefully evaluated in a 12-lead
ECG for evidence of capture, and interatrial
conduction delay (where the paced P wave is buried
in the QRS complex) [53,54]. The 12-lead ECG
taken at double standardization often brings out
atrial deflections unclear in a standard tracing. In-
teratrial conduction delay requires reprogramming
of the AV delay or the addition of a second atrial
lead for biatrial pacing to provide optimal mechan-
ical AV synchrony on the left side of the heart [55].
Latency
The delay from the pacing stimulus to the
onset of ventricular depolarization is called
 DIAGNOSTIC VALUE OF THE
latency (Fig. 20). An isoelectric onset of QRS
complex in one or a few leads can mimic latency.
Consequently, latency requires a 12-lead ECG for
diagnosis. The normal values measures less than
40 milliseconds. Latency may progress to type I
second-degree Wenckebach exit block with grad-
ual prolongation of the spike-to-QRS interval,
eventually resulting in an ineffectual stimulus
[56,57]. Latency can only be evaluated by looking
at a 12-lead ECG to rule out an isoelectric initial
part of the QRS complex.
Causes of latency:
Right ventricular infarction
Anterior infarction
Severe myocardial disease
Hyperkalemia
Antiarrhythmic drug toxicity
Variant (Prinzmetal’s) angina
These abnormalities usually occur in terminal
situations, often with a combination of ischemia,
acidosis, hypoxia, antiarrhythmic drugs, and hy-
perkalemia. Occasionally, latency and second-de-
gree exit block occurs secondary to potentially
reversible causes such as hyperkalemia and anti-
arrhythmic drug toxicity. In biventricular pacing,
latency related to LV pacing may produce
Fig. 20. Latency or first-degree pacemaker ventricular
block. (Above) Two pacing rhythm strips, lead II taken
at different times, showing a delay of up to 200 millisec-
onds between the unipolar stimulus artifact and the very
broad QRS. (Below) Magnified view of four complexes
from the above rhythm strip. The arrow highlights the
latency period. (Reproduced from Kistler PM, Mond
HG, Vohra JK. Pacemaker ventricular block. Pacing
Clin Electrophysiol 2003;26:1997–9; with permission.)
12-LEAD ECG IN CARDIAC PACING 485
suboptimal hemodynamics associated with an
ECG showing the pattern of RV pacing because
LV depolarization is delayed and overshadowed
by RV stimulation. The electrical and hemody-
namic problem can often be corrected by advanc-
ing LV stimulation by programming the
interventricular (V-V) delay, a feature available
only in contemporary devices.
Hyperkalemia
The paced QRS complex widens with hyper-
kalemia. Other common causes of a wide-paced
QRS complex include amiodarone therapy and
severe myocardial disease. Hyperkalemia can
cause pacemaker–myocardial block, which may
be reversible with or without a contribution from
antiarrhythmic drug toxicity, especially type IA
agents [58–65]. This produces first-degree exit
block (latency) and second-degree type I (Wenck-
ebach) exit block from the region of the pace-
maker stimulus, eventually progressing to
complete exit block with total lack of capture. In
a dual-chamber device, hyperkalemia may cause
failure of atrial capture associated with preserva-
tion of ventricular pacing (Fig. 21) [61–63]. This
differential effect on atrial and ventricular excit-
ability (pacing) correlates with the well-known
clinical and experimental observations that the
atrial myocardium is more sensitive to hyperkale-
mia than the ventricular myocardium. In this situ-
ation, the loss of atrial capture should be
demonstrated at the maximum programmable
AV delay to rule out latency.
Paced QRS duration during
conventional right ventricular pacing
Sweeney and colleagues [66] recently found
that a long-paced QRS duration was a significant,
independent predictor of heart failure hospitaliza-
tion in patients with sinus node dysfunction (haz-
ard ratio 1.15; 95% confidence interval 1.07,1.23)
for each 10-millisecond increase in paced QRS du-
ration (P ¼ 0.001). Paced QRS duration was not
significant for mortality (P ¼ 0.41) or atrial fibril-
lation (P ¼ 0.20) when baseline QRS duration and
other predictors were included [66]. The
association of a wide paced QRS complex with
the increased development of heart failure was
confirmed by Miyoshi and colleagues [67], who
also found that a prolonged paced QRS duration
(R190 milliseconds) was associated with a signi-
ficant increase in the overall morbidity of heart
486 BAROLD et al

Fig. 21. Hyperkalemia. (A) ECG (leads V1–V6) showing hyperkalemia-induced failure of atrial capture during DDD
pacing in a patient with severe congestive heart failure (K ¼ 6.3 mEq/L). The paced QRS complex is widened to 0.36
seconds. Pacemaker variables: Lower rate ¼ 70 ppm, AV delay ¼ 200 milliseconds, atrial output ¼ 8.1 V at a 1.0-mil-
lisecond pulse duration, and ventricular output ¼ 5.4 V at a 0.6-millisecond pulse duration. (B) ECG showing restoration
of atrial capture a few minutes after initial treatment of hyperkalemia. The pacemaker variables are the same as in (A).
The duration of the QRS complex has shortened to 0.30 seconds. The interval from the atrial stimulus to the isoelectric
segment of the PR interval measures approximately 0.22 seconds, and represents delay in interatrial conduction. (C) ECG
recorded 24 hours after (A). Pacemaker variables: Lower rate ¼ 90 ppm (increased from 70 because of congestive heart
failure and ventricular ectopy), AV interval ¼ 300 milliseconds, atrial output ¼ 5.4 V at a 0.6-millisecond pulse duration,
ventricular output ¼ 5.4 V at a 0.6-millisecond pulse duration. The QRS complex has further shortened to 0.24 seconds.
The interval from the atrial stimulus to the isoelectric segment of the PR interval has shortened to 0.16 seconds. (Repro-
duced from Barold SS, Falkoff MD, Ong LS, et al. Hyperkalemia-induced failure of atrial capture during dual-chamber
cardiac pacing. J Am Coll Cardiol 1987;10:467–9; with permission from American College of Cardiology Foundation.)

failure (P ! 0.05). On this basis, serial determina-
tions of the paced QRS duration may be clinically
useful to evaluate LV function and the risk of de-
veloping heart failure.
Unusual pacemaker stimuli
The ventricular triggered mode in some
biventricular devices automatically attempts to
provide ventricular resynchronization in the
presence of ventricular sensing (Fig. 22). A ven-
tricular-sensed event initiates an immediate emis-
sion of a ventricular or usually a biventricular
output (according to the programmed settings)
in conformity with the programmed upper rate in-
terval. For example, Medtronic (Minneapolis,
Minnesota) devices offer this function in the
VVIR mode, but in dual-chamber devices trigger-
ing occurs upon sensing only in the programmed
AV delay [2]. The ventricular output will be
 DIAGNOSTIC VALUE OF THE 12-LEAD ECG IN CARDIAC PACING 487

Fig. 22. Ventricular-triggered mode of third-generation Medtronic biventricular ICD (VVIR mode). The device triggers
a biventricular output upon sensing the RV electrogram in an attempt to provide resynchronization upon sensing. The
pacemaker stimuli (RV and LV) deform the sensed ventricular premature beats. The degree of electrical resynchroniza-
tion cannot be determined from this tracing. (Reproduced from Barold SS, Herweg B, Giudici M. Electrocardiographic
follow-up of biventricular pacemakers. Ann Noninvasive Electrocardiol 2005;10:231–55; with permission.)

ineffectual in the chamber where sensing was initi-
ated because the myocardium is physiologically
refractory. The stimulus to the other ventricle
thus attempts to provide a measure of resynchro-
nization. Ventricular triggering may be helpful in
some patients, but its true benefit is difficult to as-
sess as the ventricles may be activated in an order
that may not be hemodynamically favorable.
Closely coupled (double) ventricular stimuli
occur in two circumstances: (1) Biventricular
pacing with V-V programming (Fig. 15) [46]. (2)
Biventricular pacing with a conventional DDD
pacemaker in patients with persistent atrial fibril-
lation where the ‘‘atrial’’ channel usually paces the
LV and the ventricular channel paces the RV [68].
Two ventricular stimuli are often seen because
such devices do not usually permit programming
an AV interval of zero.
Artifacts resembling pacemaker stimuli
Artifacts occurring in the ECG of a paced
patient create confusion. The term ‘‘triboelectric
phenomena’’ is often used to describe high-voltage
deflections generated by static electricity. Tribo-
electric signals are usually wider and more irregular
than pacemaker stimuli, and often recognizable as
artifacts. Occasionally, the diagnosis depends on
finding subtle differences from the pacing stimulus
[69–71]. Magnification of the questionable artifact
may reveal a relatively prolonged overshoot not
present with pacemaker stimuli. Such an overshoot
is typical of electrostatic discharges, and should be
sought in the evaluation of artifacts mimicking
pacemaker stimuli to avoid the diagnosis of pace-
maker malfunction.
Summary
The paced 12-lead ECG is a valuable tool in
the assessment of patients with pacemakers, and
ideally should be recorded routinely at the time of
implantation and during follow-up. It has become
particularly important in patients undergoing
cardiac resynchronization. The multiplicity of
clinical situations described in this review high-
light the pitfalls of using a single ECG lead in the
overall evaluation of pacemaker patients. The
design of programmers capable of registering
a 12-lead ECG would obviate the need of an
additional electrocardiograph and encourage the
routine recording of the paced 12-lead ECG with
each patient encounter. Such an arrangement
would improve the care of pacemaker patients.
References
[1] Barold SS, Levine PA, Ovsyshcher IE. The paced 12-
lead electrocardiogram should no longer be
neglected in pacemaker follow-up. PACE 2001;24:
1455–8.
[2] Barold SS, Herweg B, Giudici M. Electrocardio-
graphic follow-up of biventricular pacemakers.
Ann Noninvasive Electrocardiol 2005;10:231–55.
[3] Asirvatham SJ. Electrocardiogram interpretation
with biventricular pacing devices. In: Hayes DL,
Wang PJ, Sackner-Bernstein J, Asirvatham SJ, edi-
tors. Resynchronization and defibrillation for heart
488 BAROLD
failure. A practical approach. Oxford (UK): Black-
well-Futura; 2004. p. 73–97.
[4] Kay GN. Troubleshooting and programming of car-
diac resynchronization therapy. In: Ellenbogen KA,
Kay GN, Wilkoff BL, editors. Device therapy for
congestive heart failure. Philadelphia (PA): WB
Saunders; 2004. p. 232–93.
[5] Steinberg JS, Maniar PB, Higgins SL, et al. Nonin-
vasive assessment of the biventricular pacing system.
Ann Noninvasive Electrocardiol 2004;9:58–70.
[6] Lau CP, Barold S, Tse HF, et al. Advances in devices
for cardiac resynchronization in heart failure.
J Interv Card Electrophysiol 2003;9:167–81.
[7] Garrigue S, Barold SS, Clémenty J. Electrocardiog-
raphy of multisite ventricular pacing. In: Barold SS,
Mugica J, editors. The fifth decade of cardiac pacing.
Elmsford (NY): Blackwell-Futura; 2004. p. 84–100.
[8] Barold SS. Normal and abnormal patterns of ven-
tricular depolarization during cardiac pacing. In:
Barold SS, editor. Modern cardiac pacing. Mt. Kis-
co (NY): Futura; 1985. p. 545–69.
[9] Barold SS, Falkoff MD, Ong LS, et al. Electrocar-
diographic analysis of normal and abnormal pace-
maker function. In: Dreifus LS, editor. Pacemaker
therapy, cardiovascular clinics. Philadelphia (PA):
F.A.Davis; 1983. p. 97–134.
[10] Barold SS, Falkoff MD, Ong LS, et al. Electrocar-
diographic diagnosis of myocardial infarction dur-
ing ventricular pacing. Cardiol Clin 1987;5:403–17.
[11] Klein HO, Becker B, Sareli P, et al. Unusual QRS
morphology associated with transvenous pace-
makers. The pseudo RBBB pattern. Chest 1985;87:
517–21.
[12] Yang YN, Yin WH, Young MS. Safe right bundle
branch block pattern during permanent right ven-
tricular pacing. J Electrocardiol 2003;36:67–71.
[13] Coman JA, Trohman RG. Incidence and electrocar-
diographic localization of safe right bundle branch
block configurations during permanent ventricular
pacing. Am J Cardiol 1995;76:781–4.
[14] Ciolli A, Trambaiolo P, Lo Sardo G, et al. Asymp-
tomatic malposition of a pacing lead in the left ven-
tricle: the case of a woman untreated with
anticoagulant therapy for eight years. Ital Heart J
2003;4:562–4.
[15] Paravolidakis KE, Hamodraka ES, Kolettis TM,
et al. Management of inadvertent left ventricular
permanent pacing. J Interv Card Electrophysiol
2004;10:237–40.
[16] Ergun K, Tufekcioglu O, Karabal O, et al. An un-
usual cause of stroke in a patient with permanent
transvenous pacemaker. Jpn Heart J 2004;45:873–5.
[17] Arnar DO, Kerber RE. Cerebral embolism resulting
from a transvenous pacemaker catheter inadver-
tently placed in the left ventricle: a report of two
cases confirmed by echocardiography. Echocardiog-
raphy 2001;18:681–4.
[18] Sharafi M, Sorkin R, Sharifi V, et al. Inadvertent
malposition of a transvenous inserted pacing lead
et al
in the left ventricular chamber. Am J Cardiol 1995;
76:92–5.
[19] Burkart TA, Lewis JF, Conti JB, et al. Malposi-
tioned ventricular pacing l lead in the left ventricle.
Clin Cardiol 2000;23:123–4.
[20] Agnelli D, Ferrari A, Saltafossi D, et al. A cardiac
embolic stroke due to malposition of the pacemaker
lead in the left ventricle. A case report. Ital Heart J
2000;1(1 Suppl):122–5.
[21] Van Gelder BM, Bracke FA, Oto A, et al. Diagnosis
and management of inadvertently placed pacing and
ICD leads in the left ventricle: a multicenter experi-
ence and review of the literature. PACE 2000;23:
877–83.
[22] Agarwal A, Kapoor A, Garg N. Inadvertent trans-
venous left ventricular pacing. A case report. Indian
Heart J 2000;52:331–4.
[23] Blommaert D, Mucumbitsi J, de Roy L. Images in
cardiology. Ventricular pacing and right bundle
branch block morphology: diagnosis and manage-
ment. Heart 2000;83:666.
[24] Trigano JA, Paganelli F, Fekhar S, et al. Pocket
infection complicating inadvertent transarterial
permanent pacing. Successful percutaneous explan-
tation. Clin Cardiol 1999;22:492–3.
[25] Chun JK, Bode F, Wiegand UK. Left ventricular
malposition of pacemaker lead in Chagas’ disease.
Pacing Clin Electrophysiol 2004;27:1682–5.
[26] Orlov MV, Messenger JC, Tobias S, et al. Transeso-
phageal echocardiographic visualization of left ven-
tricular malpositioned pacemaker electrodes:
implications for lead extraction procedures. Pacing
Clin Electrophysiol 1999;22:1407–9.
[27] AL-Dashti R, Huynh T, Rosengarten M, et al.
Transvenous pacemaker malposition in the systemic
circulation and pacemaker infection: a case report
and review of the literature. Can J Cardiol 2002;18:
887–90.
[28] Firschke C, Zrenner B. Images in clinical medicine.
Malposition of dual-chamber pacemaker lead.
N Engl J Med 2002;346:e2.
[29] Shettigar UR, Loungani RR, Smith CA. Inad-
vertent permanent ventricular pacing from the
coronary vein: an electrocardiographic, roentgeno-
graphic, and echocardiographic assessment. Clin
Cardiol 1989;12:267–74.
[30] Altmiks R, Nathan AW. Left ventricular pacing via
the great cardiac vein in a patient with tricuspid and
pulmonary valve replacement. Heart 2001;85:91.
[31] Barold SS, Banner R. Unusual electrocardiographic
pattern during transvenous pacing from the middle
cardiac vein. Pacing Clin Electrophysiol 1978;1:
31–4.
[32] Ammann P, Sticherling C, Kalusche D, et al. An
electrocardiogram-based algorithm to detect loss of
left ventricular capture during cardiac resynchroni-
zation therapy. Ann Intern Med 2005;142:968–73.
[33] Yong P, Duby C. A new and reliable method of in-
dividual ventricular capture identification during
 DIAGNOSTIC VALUE OF THE
biventricular pacing threshold testing. PACE 2000;
23:1735–7.
[34] Georger F, Scavee C, Collet B, et al. Specific electro-
cardiographic patterns may assess left ventricular
capture during biventricular pacing [abstract].
PACE 2002;25:56.
[35] Hart D, Luiza P, Arshad R, et al. Assessment of ven-
tricular capture in patients with cardiac resynchroni-
zation devices; a simple surface electrocardiographic
algorithm [abstract]. PACE 2003;26:1083.
[36] Leclercq C, Kass DA. Retiming the failing heart:
principles and current clinical status of cardiac
resynchronization. J Am Coll Cardiol 2002;39:
194–201.
[37] Kass DA. Predicting cardiac resynchronization re-
sponse by QRS duration: the long and short of it.
J Am Coll Cardiol 2003;42:2125–7.
[38] Leclercq C, Faris O, Tunin R, et al. Systolic im-
provement and mechanical resynchronization does
not require electrical synchrony in the dilated failing
heart with left bundle-branch block. Circulation
2002;106:1760–3.
[39] Kistler PM, Mond HG, Corcoran SJ. Biventricular
pacing: it isn’t always as it seems. PACE 2003;26:
2185–7.
[40] Alonso C, Goscinska K, Ritter P, et al. Upgrading to
triple-ventricular pacing guided by clinical outcomes
and echo assessment; a pilot study [abstract]. Euro-
pace 2004;6(Suppl 1):195.
[41] Ricci R, Pignalberi C, Ansalone G, et al. Early and
late QRS morphology and width in biventricular
pacing: relationship to lead site and electrical
remodeling. J Interv Card Electrophysiol 2002;6:
279–85.
[42] Van Gelder BM, Bracke FA, Pilmeyer A, et al. Tri-
ple-site ventricular pacing in a biventricular pacing
system. PACE 2001;24:1165–7.
[43] Bulava A, Ansalone G, Ricci R, et al. Triple-site
pacing with biventricular device. Incidence of the
phenomenon and cardiac resynchronization benefit.
J Interv Card Electrophysiol 2004;10:37–45.
[44] Herweg B, Barold SS. Anodal capture with second-
generation biventricular cardioverter-defibrillator.
Acta Cardiol 2003;58:435–6.
[45] Thibault B, Roy D, Guerra PG, et al. Anodal right
ventricular capture during left ventricular stimula-
tion in CRT-implantable cardioverter defibrillators.
Pacing Clin Electrophysiol 2005;28:613–9.
[46] van Gelder BM, Bracke FA, Meijer A. The effect of
anodal stimulation on V–V timing at varying V–V
intervals. Pacing Clin Electrophysiol 2005;28:771–6.
[47] Patel AM, Westveer DC, Man KC, et al. Treatment
of underlying atrial fibrillation: paced rhythm ob-
scures recognition. J Am Coll Cardiol 2000;36:784–7.
[48] McLellan CS, Abdollah H, Brennan FJ, et al. Atrial
fibrillation in the pacemaker clinic. Can J Cardiol
2003;19:492–4.
[49] Sparks PB, Mond HG, Kalman JM, et al. Atrial
fibrillation and anticoagulation in patients with
12-LEAD ECG IN CARDIAC PACING 489
permanent pacemakers: implications for stroke
prevention. Pacing Clin Electrophysiol 1998;21:
1258–67.
[50] Fananapazir L, McAreavey D. Therapeutic options
in patients with obstructive hypertrophic cardio-
myopathyand severe drug-refractory symptoms.
J Am Coll Cardiol 1998;31:259–64.
[51] Barold SS. New indications for cardiac pacing. In:
Singer I, Barold SS, Camm AJ, editors. Nonpharma-
cological therapy of arrhythmias for the 21st
century. The state of the art. Armonk (NY): Futura;
1998. p. 775–95.
[52] Kleinfeld M, Barold SS, Rozanski JJ. Pacemaker
alternans. A review. Pacing Clin Electrophysio
1987;10:924–33.
[53] Daubert JC, Pavin D, Jauvert G, et al. Intra- and
interatrial conduction delay: implications for
cardiac pacing. PACE 2004;27:507–25.
[54] Parravicimi U, Mezzani A, Bielli M, et al. DDD pac-
ing and interatrial conduction block:importance of
optimal interval setting. PACE 2000;23:1448–50.
[55] D’Allones GR, Pavin D, Leclercq C, et al. Long-
term effects of biatrial synchronous pacing to pre-
vent drug-refractory atrial tachyarrhythmia:
a nine-year experience. J Cardiovasc Electrophysiol
2000;11:1081–91.
[56] Kistler PM, Mond HG, Vohra JK. Pacemaker ven-
tricular block. Pacing Clin Electrophysiol 2003;26:
1997–9.
[57] Klein HO, Di Segni E, Kaplinsky E, et al. The
Wenckebach phenomenon between electric
pacemaker and ventricle. Br Heart J 1976;38:961–5.
[58] Peter T, Harper R, Hunt D, et al. Wenckebach
phenomenon in the exit area from a transvenous
pacing electrode. Br Heart J 1976;38:201–352.
[59] Mehta J, Khan AH. Pacemaker Wenckebach
phenomenon due to antiarrhythmic drug toxicity.
Cardiology 1976;61(3):189–94.
[60] Varriale P, Manolis A. Pacemaker Wenckebach
secondary to variable latency: an unusual form of
hyperkalemic pacemaker exit block. Am Heart J
1987;114:189–92.
[61] Barold SS. Loss of atrial capture during DDD
pacing: what is the mechanism? Pacing Clin Electro-
physiol 1998;21:1988–9.
[62] Barold SS, Falkoff MD, Ong LS, et al. Hyperkale-
mia-induced failure of atrial capture during dual-
chamber cardiac pacing. J Am Coll Cardiol 1987;
10:467–9.
[63] Ortega-Carnicer J, Benezet J, Benezet-Mazuecos J.
Hyperkalaemia causing loss of atrial capture and
extremely wide QRS complex during DDD pacing.
Resuscitation 2004;62:119–20.
[64] Varriale P, Manolis A. Pacemaker Wenckebach
secondary to variable latency: an unusual form of
hyperkalemic pacemaker exit block. Am Heart J
1987;114:189–92.
[65] Dohrmann ML, Goldschlager NF. Myocardial
stimulation threshold in patients with cardiac
490 BAROLD
pacemakers: effect of physiologic variables, pharma-
cologic agents, and lead electrodes. Cardiol Clin
1985;3:527–37.
[66] Sweeney MO, Hellkamp AS, Lee KL, et al and
Mode Selection Trial (MOST) Investigators. Associ-
ation of prolonged QRS duration with death in
a clinical trial of pacemaker therapy for sinus node
dysfunction. Circulation 2005;111:2418–23.
[67] Miyoshi F, Kobayashi Y, Itou H, et al. Prolonged
paced QRS duration as a predictor for congestive
heart failure in patients with right ventricular api-
cal pacing. Pacing Clin Electrophysiol 2005;28:
1182–8.
et al
[68] Barold SS, Gallardo I, Sayad D. The DVI mode of
cardiac pacing: a second coming? Am J Cardiol
2002;90:521–3.
[69] Kahan S, Miller CW, Hayes DL, et al. Triboelectric
simulation of pacemaker malfunction. Europace
2002;4:325–7.
[70] Barold SS. Images is cardiology: initiation of pace-
maker endless loop tachycardia by triboelectricity.
Heart 2001;85:248.
[71] Barold SS, Falkoff MD, Ong LS, et al. Differential
diagnosis of pacemaker pauses. In: Barold SS, edi-
tor. Modern cardiac pacing. Mt Kisco (NY): Futura;
1985. p. 587–613.
 Cardiol Clin 24 (2006) 491–504

Status of Computerized Electrocardiography

Richard H. Hongo, MDa,*, Nora Goldschlager, MDb
a
Division of Cardiology, California Pacific Medical Center, 2100 Webster Street, Suite 516,
San Francisco, CA 94115, USA
b
Division of Cardiology, San Francisco General Hospital, 1001 Potrero Avenue, Room 5G1,
San Francisco, CA 94110, USA

In the 1950s, the first step toward computer-
ized electrocardiogram (ECG) analysis was taken
with the successful development of the analog-to-
digital converter. Electrical ECG signals could
now be transformed into digital information that
could then be processed by computer. The first
published reports on computerized ECG analysis
emerged in the early 1960s from two separate
laboratories [1,2]. Pipberger and colleagues [1]
championed the use of the Frank XYZ lead sys-
tem, and Caceres and colleagues [2] developed
the first analysis program for 12-lead ECGs. Ini-
tially, the simultaneously obtained three-orthogo-
nal lead system had an advantage over the 12-lead
system that was recorded sequentially. Eventually,
the 12-lead system became capable of recording
groups of three or more leads simultaneously.
The familiar lead configuration (I–II–III, aVR–
aVL–aVF, V1–V2–V3, V4–V5–V6) was introduced
by Bonner and colleagues in 1972 [3]. Clinician
preference ultimately led to the predominance
of the 12-lead system over the orthogonal Frank
XYZ lead system. By the mid-1970s, European
and Japanese investigators were actively contri-
buting to this new field.
The development of signal filtering in the
1960–1970s was a crucial step in advancing
computerized ECG analysis. It was recognized
that by changing the filtering, the same raw data
could be manipulated into any number of differ-
ent ECG signal tracings. The ECG signal was first
There was no funding support in the writing of this
article. There are no relationships that would pose
a conflict of interest for either coauthor.
* Corresponding author.
E-mail address: rhongo@cpcmg.com (R.H. Hongo).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.03.005
processed using a 50-Hz low-pass filter (passage of
lower frequency component with attenuation of
frequencies above the cutoff frequency); this
facilitated the blocking of electrical interference
by 60-Hz alternating current (AC) power lines. It
became apparent, however, that there was impor-
tant information in the ECG signal in the 100-Hz
region, especially in patients with heart disease,
making higher cutoff frequencies necessary. AC
noise was independently blocked with band-stop,
or notch filters. Determination of the ideal cutoff
frequency, in turn, influenced other processing pa-
rameters such as the sampling rate of the analog-
to-digital converters.
Early computer analysis programs used ‘‘de-
terministic’’ algorithms that processed ECG data
though multiple logical decision rules created by
expert ECG readers. The algorithms used sharply
defined ‘‘cutoff points’’ that determined whether
or not a decision rule was fulfilled. With this type
of analysis, ‘‘long-QT syndrome’’ could be di-
agnosed in one person but not another when the
respective QT intervals differed by only 1 milli-
second. Since the 1980s, probability theories and
statistics have been incorporated into computer-
ized ECG analysis. The use of Bayesian analysis
better accounts for the complexities of ECG
diagnosis, and introduces influences from ele-
ments such as pretest probabilities. Multivariate
statistics, in turn, are used to analyze demo-
graphic parameters that affect pretest probabili-
ties. More recently, artificial neural networks have
also been looked at as a way to better mimic how
expert ECG readers arrive at a diagnosis; that is,
through pattern recognition [4,5].
The lack of standardization of technical
parameters, logic systems, and diagnostic
cardiology.theclinics.com
492 HONGO & GOLDSCHLAGER
classifications has led to variability in perfor-
mance among the different computer programs
[6]. In 1975, the American Heart Association
made comprehensive technical recommendations
that were further expounded upon in 1990 [7,8].
In Europe, the ‘‘Common Standards for Quanti-
tative Electrocardiography’’ project was launched
in 1980. In collaboration with six North American
and one Japanese center, and with follow-up stud-
ies performed in 1985 and 1989, this project was
an effort to set an international standard for com-
puterized ECGs [9]. An expanding reference data-
base was created from ECGs obtained from actual
patients with corresponding clinical data (ie, car-
diac catheterizations, echocardiograms, and so
on), as well as from computer-generated ECGs.
Despite these efforts and the advances resulting
from them, an international standard is still
unrealized.
Despite the significant limitations of the initial
systems, computerized ECG analysis was recog-
nized early on as having the potential to make an
important clinical contribution [10], and develop-
ment gradually shifted from academic centers to
industry. With the advent of the microprocessor
in the 1970s, ECG analysis moved from main-
frames housed in large rooms to portable carts
at the bedside. The first commercial 12-lead
ECG analysis program was made available in
the early 1980s by Marquette Electronics, fol-
lowed closely by Siemens-Elema. In an important
and popular advance, Hewlett Packett and Mar-
quette made computer analysis available on their
portable ECG carts. In 1975, it was estimated
that a mere 4 million ECGs were being processed
by computer in the United States [11]. This esti-
mate had increased to 15 million by 1983 [12].
The last comprehensive survey was performed in
1988 by Drazen and colleagues [11], at which
time 52 million ECGs were being processed by
computer, with revenues exceeding $70 million.
It is estimated today that over 100 million ECGs
are analyzed by computer annually in the United
States, with a similar number in Europe and in the
rest of the world. The ECG is still the most com-
monly performed cardiac test [13], and in the
United States, virtually 100% are analyzed by
computer.
Technical aspects of computer analysis
Computerized ECG analysis can be divided
into three distinct stages: (1) signal processing,
(2) measurements, and (3) interpretation. In the
first stage, electrical data are acquired and con-
verted from analog to digital signals. The fidelity of
the signal is better preserved with faster sampling
rates. The signal is then filtered to eliminate noise.
Data compression, transmission, and archiving
are additional important aspects of digital pro-
cessing. In the second stage, the processed signal
undergoes wave recognition. This involves the
computer first selecting the signal (or ECG
complex) to be measured. The underlying basic
rhythm is distinguished from premature com-
plexes. The computer can also make a template
from the composite of several complexes as a way
to minimize the effects of small beat-to-beat
variations. Simultaneous acquisition of ECG
leads aids in wave recognition. Once there is
appropriate labeling of wave onset and offset,
interval and amplitude parameters are measured.
In the final stage, the measured parameters are
subjected to multiple criteria to arrive at the
computer diagnosis.
Acquisition, conversion, and filtering of data
Simultaneous signal acquisition is today’s stan-
dard. The ability to cross reference multiple signals
improves the determination of wave onset and
offset, especially of the QRS complex [14], and
yields more accurate interval measurements
and wave recognition. The conversion of the analog
electrical signal to digital data inevitably results in
a loss of information. The fidelity of the signal is
better preserved when a higher bit resolution and
a faster sampling rate are used. These parameters
are limited by both microprocessing capability
and also by storage capacity. A higher fidelity is re-
quired for computer analysis compared with visual
reading. For the former, expert opinion recom-
mends a maximum resolution of 10 mV and at least
a 500-Hz sampling rate [8,9].
Filtering is performed before wave recognition
to eliminate extraneous noise from the signal of
interest. An example of high-frequency noise is
myopotentials; movement artifact and baseline
wander from respiration are common examples
of low-frequency noise. Electrical interference
from local power circuits, or ‘‘60-Hz AC noise,’’
is another important source of unwanted signal
that is frequently encountered in medical office
and hospital environments. Low-pass filters (pas-
sage of lower frequency component, with attenu-
ation of frequencies above the cutoff frequency)
are effective in minimizing high-frequency noise.
 STATUS OF COMPUTERIZED ELECTROCARDIOGRAPHY 493

Initial cutoff frequencies of 50 Hz were designed
to also filter out ‘‘AC noise.’’ As mentioned
above, it has since been recognized that there is
clinically significant information in the 100-Hz
region, and a cutoff of 150 Hz is now recom-
mended [8]. A band-stop, or notch filter (attenua-
tion of component within a frequency band), is
used to blanket 60-Hz noise.
High-pass filters (passage of higher frequency
components, with attenuation of frequencies be-
low the cutoff frequency) with a 0.5 Hz cutoff
frequency will eliminate most baseline wander and
motion artifact while allowing heart rates of more
than 30 beats per minute to be detected. This
cutoff, however, encroaches on the frequencies of
ventricular repolarization (1–3 Hz) and distortion
in the ST-segment and T-wave can occur. A lower
cutoff of 0.05 Hz was the initial (1975) recom-
mendation by the American Heart Association [7]
that effectively avoided repolarization distortion,
but some baseline wander would still be present.
As sophisticated filters that specifically target
low-frequency distortions have become available,
the low-frequency cutoff can be relaxed back up
into the 0.5-Hz range.
The type of filtering and the choice of each
cutoff frequency can dramatically alter the final
processed signal. For instance, higher cutoff
frequencies for the low-pass filter can result in
uncovering high-frequency notches in the QRS
complex that have been suggested by some to have
clinical significance [15–17]. Higher cutoff fre-
quencies also allow for more accurate amplitude
measurements because less high-frequency signal
is filtered out; a difference between 40 Hz and
100 Hz results in significant changes in measured
amplitude [18]. Diagnoses that rely heavily on
accurate amplitude measurements, such as

Fig. 1. (A) Significant Q waves in the inferior leads are not recognized by the computer, and the diagnosis of an inferior
wall myocardial infarction is missed. (B) The QRS complex width is grossly undermeasured to be 100 milliseconds (man-
ually measured to be 180 milliseconds), and the diagnosis of right bundle branch block is not made. (C) Ventricular pre-
excitation (delta waves) is not recognized by the computer, and is misinterpreted as Q waves in the lateral leads, leading
to a diagnosis of lateral wall infarction. The prominent anterior forces, also due to ventricular preexcitation, lead to
a misdiagnosis of right ventricular hypertrophy versus posterior myocardial infarction.
494 HONGO & GOLDSCHLAGER
ventricular hypertrophy, are susceptible to these
types of parameter changes. Filtering parameters
also affects the limits on sampling rate, as the sam-
pling frequency should be several times faster than
the ‘‘high-frequency cutoff’’ [8].
Wave recognition and measurements
Before components of the ECG signal can be
measured, the waves that make up the ECG signal
must be identified. When the signal is taken
during normal sinus rhythm, and patterns are
regular and predictable, this process is relatively
reliable. On the other hand, in the setting of atrial
arrhythmias, this step can be challenging. Much
of the accuracy of rhythm diagnosis depends on
the proper recognition of irregular wave activity.
The difficulties in diagnosing premature atrial
complexes, atrial flutter, and atrial fibrillation
stem from the inadequacies of atrial wave recog-
nition. Techniques such as time-domain analysis
of QRST-wave-subtracted electrograms have been
used to unmask underlying atrial activity so that
Fig. 1 (continued )
ventricular electrical activity does not drown out
the atrial signal. Spectral analysis can be used to
detect regularity in the atrial activity, and can help
distinguish atrial flutter from atrial fibrillation
[19].
Once the waves are properly identified, dura-
tion and amplitude measurements are made.
Although this step is seemingly straightforward,
there are factors that can influence the precision
and accuracy of these measurements. Despite
significant improvements in the ability to analyze
simultaneous ECG leads, identification of wave
onset and offset varies among different computer
programs. Identical raw digital data have been
shown to result in differences in QRS duration
and QT interval measurements, with mean differ-
ences of up to 18 and 24 milliseconds, respectively
[20]. This is partly explained by the different ways
various algorithms treat wave components such as
isoelectric segments involving the Q or QS waves,
or the U waves within the QT interval [9]. Impre-
cise measurement of the QRS complex and incor-
rect identification of wave components can lead to
 STATUS OF COMPUTERIZED ELECTROCARDIOGRAPHY
misinterpretation of important ECG diagnoses
such as myocardial infarctions, bundle branch
blocks, and ventricular preexcitation (Fig. 1).
Variability in amplitude measurement has been
found to be less than that of interval measurements
[20], but there can be day-to-day variability in volt-
age measurement of up to 20% that can result in re-
classification and consequent differences in the
computer diagnosis [21]. Some of the known factors
that contribute to variability in voltage include fluc-
tuation in metabolic and fluid status, variable lung
and torso impedances, and changes in patient posi-
tion that lead to change in the position of the heart
within the chest. There are also procedural and
equipment-related factors such as accuracy and
consistency of electrode placement, low-pass filter
settings, digital sample error, and variable perfor-
mance among machines [22].
Interpretation and diagnostic classification
The foundation for computerized ECG inter-
pretation has been a ‘‘deterministic’’ approach, or
Fig. 1 (continued)
495
one that uses an algorithm with predictable
behavior. This method relies on diagnostic criteria
constructed by human experts. These criteria, in
turn, are either met or not met, and do not allow
for gradations. External factors that influence the
likelihood of a diagnosis, such as pretest proba-
bility, are ignored. Bayesian probability theory
has introduced a ‘‘statistical’’ approach. Consid-
erations such as pretest probability, or clinical
suspicion, can now be factored into the computer
interpretation. For example, ST-segment eleva-
tion is much more likely be a manifestation of
acute myocardial infarction if the patient is male,
65 years old, has a history of coronary disease,
and is experiencing severe substernal chest pain,
as opposed to an asymptomatic 30-year-old
woman undergoing a routine screening ECG.
Multivariate statistical methods are used to ana-
lyze the demographic parameters. The effective-
ness of this type of statistical analysis, however,
depends on the reliability of available ECG
databases that become the basis for determining
key determinants such as the prevalence of
496 HONGO & GOLDSCHLAGER

a particular diagnosis. Artificial neural networks,
which use pattern recognition, have been incor-
porated into some programs [4,5].
The inability to know the specific details of the
computer algorithms is a reality of industry-
driven technology development, and the logic
systems for ECG interpretation have become
proprietary information of manufacturers. To
a certain degree, programs are developed by trial
and error, or by retesting against ECG databases
[23]. These databases are created from real patient
data, and from ‘‘electronically generated ECG sig-
nals’’ or ‘‘electronic test patients’’ [24,25]. Manu-
facturers report a high sensitivity and specificity
for the major diagnoses when tested in this fash-
ion [23]. A relatively small number of studies
have evaluated the performance of computer anal-
ysis in the clinical setting. The rate of computer-
generated diagnostic errors reported in these
studies, however, is much larger than what would
be expected from the performance indices re-
ported by the manufacturers. This suggests that
the ECG databases used in testing these computer
programs do not yet sufficiently represent the clin-
ical population.
Current status of computerized
electrocardiography
Willems and colleagues [26] compared the di-
agnostic abilities of nine computer programs
with that of eight experienced cardiologists, using
a database comprised of 1220 ECGs. The data-
base was limited to patients with hypertrophy
and infarction, and a control group. The final di-
agnoses were verified by nonelectrocardiographic
clinical data. The combined overall accuracy of
the computer programs was 76.3% compared
with an accuracy of 79.2% for the cardiologists

Fig. 2. (A) The computer misinterprets the 12-lead ECG to be a ‘‘normal ECG.’’ Right axis deviation, qR complex in
lead V1, and T-wave inversions in multiple leads are not detected by the computer algorithm. (B) The computer misin-
terprets the presence of ‘‘multiple atrial premature complexes’’ in this 12-lead ECG, missing the diagnosis of normal si-
nus rhythm. It is not clear from the tracing what is being misidentified as atrial premature complexes.
 STATUS OF COMPUTERIZED ELECTROCARDIOGRAPHY
as a group. When the accuracy of the computer
programs and cardiologists were assessed sepa-
rately within each group, the median accuracy of
computer programs was lower (69.7%) than that
of the cardiologists (76.3%). Although some
computer programs were clearly inferior to the
interpretation of cardiologists, others were compa-
rably accurate. In the detection of acute myocar-
dial injury by ST-segment assessment, computer
ECG analysis has been found to have lower sensi-
tivity compared with expert readers (52% versus
66%, P ! 0.001), but higher specificity (98% ver-
sus 95%, P ! 0.001) [23].
Although computer ECG analysis can be
comparable to interpretations by expert ECG
readers, misinterpretations do occur. Bogan and
colleagues [27] reported that out of 2298 computer
diagnoses of atrial fibrillation, 19% were in error.
Of particular concern was that of the inaccurate
interpretations, 24% were not corrected by the or-
dering physician, and 10% received inappropriate
therapy that included initiation of anticoagulation
therapy and hospitalization. Misinterpretation of
Fig. 2 (continued)
497
the ECG, however, is not unique to computerized
analysis. In a comprehensive review of the current
literature on ECG interpretation, Salero and col-
leagues [28] found that 4% to 33% of physician
interpretations contained errors of major impor-
tance, and inappropriate management occurred in
up to 11% of patients. Adverse clinical outcomes,
however, including preventable death, was ob-
served in only 0.1% to 1.4% of cases.
Perhaps the major reason that computerized
misinterpretations lead to clinical mismanagement
is an overreliance on the accuracy of the computer
algorithms. Primary care physicians have been
observed to change up to 45% of their initial ECG
interpretations after seeing a computerized in-
terpretation [29], and the computer reading has
been shown to mislead clinical decision making
[30]. The computerized interpretation was found
to not reduce major errors in ECG reading by se-
nior house officers [31]. Many house officers fold
over the computer reading at the top of the
ECG tracing so it cannot influence their interpre-
tation. Although this is done partly so that they
498 HONGO & GOLDSCHLAGER

can practice their own interpretive skills, there is Strengths of computerized ECG analysis
also a general mistrust of the computer reading.
One of the most obvious strengths of computer
Because computerized ECG analysis is seldom re-
analysis is the automated measurement of key
viewed during medical training, most physicians
parameters such as the heart rate, PR interval,
do not gain a sense of how reliable the computer
QRS duration, and QT interval. Calculation of
reading is unless they become ECG readers.
the corrected QT interval and axes are also
As is true with any diagnostic test, computer-
performed automatically. This is an important
ized ECG analysis should be an adjunct to, not
component of ECG reading, but one that is
a substitute for, physician interpretation. When
extremely tedious if done manually. One should
the limitations of the computer program are not
be aware of potential variability in the measured
known (which is usually the case), there is a real
intervals, however, that stems from variability in
danger of physicians accepting the interpretation
inscribing wave onset and offset. The QT interval
either without overreading, or without under-
is the most difficult measurement because of a lack
standing the findings. A certain level of interpre-
of agreement in the way T-wave offset is de-
tive skill is needed by physicians overreading the
termined [32,33]; additional errors are made if TU
ECG if they are expected to recognize misinter-
waves are present. For this reason, when serial QT
pretations by the computer. A firm understanding
or QTU intervals are being monitored during an-
of the strengths and limitations of computer
tiarrhythmic drug therapy, it is not sufficient to
analysis is necessary if one is to use this diagnostic
follow the computerized measurement without
tool effectively.

Fig. 3. Concentric left ventricular hypertrophy with wall thickness of 15 mm by echocardiography was not detected by
the 12-lead ECG. The QRS complex amplitudes do not reach criteria for an ECG diagnosis. The accompanying repo-
larization changes with characteristic ‘‘strain’’ are seen, but not read by the computer.
 STATUS OF COMPUTERIZED ELECTROCARDIOGRAPHY 499

verifying the intervals manually in multiple ECG
leads.
Although a rare report can be found [34] that
argues that computer interpretation slows readers
down, the consensus of multiple studies [30,31,35–
37] is that it saves time. Hillson and colleagues [30]
performed a study in which 40 primary care phy-
sicians (family physicians and general internists)
interpreted 10 ECGs with accompanying clinical
vignettes. There was an average decrease in the
time it took to read each tracing of 15 seconds
(24%) when a computer reading was made avail-
able to the readers. Brailer and colleagues [35]
evaluated 22 cardiologists that as a group read
1760 ECGs. Computerized analysis cut the time
spent reading a tracing from an average of 81 sec-
onds to 64 seconds (28%), while improving the
accuracy of the readings.
The more straightforward diagnoses appear to
be better detected by the computer, and most time
is saved when the ECG is simple but tedious
(multiple diagnoses) to read [35,38]. The comput-
erized reading also offers a real-time second opin-
ion that may prevent ‘‘inadvertent oversight.’’
When the diagnoses are ambiguous or controver-
sial, computerized interpretations can make over
reading more time consuming [35].
The anecdotal experience of many ECG
readers is that a computer interpretation of
‘‘normal ECG’’ is usually, but not always, correct
(Fig. 2). Poon and colleagues [39] found the com-
puter to have high sensitivity of 98.7% (3531/
3579), but somewhat lower specificity of 90.1%
(338/375) in diagnosing sinus rhythm. The au-
thors cautioned against routinely accepting a com-
puter interpretation of ‘‘sinus rhythm’’ because
close to 10% of nonsinus rhythms were errone-
ously interpreted as sinus rhythm. Because the
prevalence of sinus rhythm was high in this study
(3579/3954), however, the positive predictive

Fig. 4. (A) The computer identifies P-waves in error and atrial fibrillation is not detected. (B) Atrial fibrillation with
a relatively regular ventricular rhythm is not recognized by the computer algorithm. The computer erroneously detects
an organized atrial rhythm ‘‘with unusual P axis.’’ (C) The premature atrial complexes are not recognized, and the var-
iability in the ventricular rate lead to the misdiagnosis of atrial fibrillation. (D) Significant artifact that potentially could
lead a reader to misdiagnose atrial fibrillation does not mislead the computer.
500 HONGO & GOLDSCHLAGER
value of a computer diagnosis of ‘‘sinus rhythm’’
was 99.0%, supporting the sentiment that a com-
puter statement of ‘‘sinus rhythm’’ is generally
reliable.
Limitations of computerized ECG analysis
Limitations of electrocardiography as a diag-
nostic modality carry over to computerized elec-
trocardiography. One of the most important
limitations of electrocardiography is the inability
to effectively detect certain anatomic and patho-
physiologic conditions. Despite the development
of multiple diagnostic ECG criteria, electrocardi-
ography has only around a 50% sensitivity in
detecting left ventricular hypertrophy that has
been established by echocardiography (Fig. 3)
[40–42]. It has been recognized for years that elec-
trocardiography is better at diagnosing arrhyth-
mias and conduction disturbances than
structural cardiac or metabolic abnormalities.
Ironically, rhythm interpretation is recognized to
be one of the most difficult tasks for the computer.
Fig. 4 (continued )
Only a handful of studies have specifically
evaluated the accuracy of computerized rhythm
diagnosis. Shirataka and colleagues [24] found
that although sinus rhythm with first-degree AV
block had 100% agreement with the reference
standard across five different computer programs,
the accuracy of diagnosing second-degree AV
block varied between 0% and 100%. In a study
that assessed 11,610 consecutive computerized
ECG interpretations, Varriale and colleagues
[43] reported that multifocal atrial tachycardia
was universally misdiagnosed as atrial fibrillation,
accounting for 14% of ECGs interpreted as atrial
fibrillation. Bogun and colleagues [27] found that
of 2298 ECGs interpreted to be atrial fibrillation,
19% were misinterpreted by the computer. The
difficulties in diagnosing atrial fibrillation is one
of the most commonly encountered problems
with computerized rhythm analysis (Fig. 4).
More recently, Poon and colleagues [39] examined
computerized rhythm interpretation by analyzing
4297 consecutive ECGs using one of the more
advanced programs currently available (GE
 STATUS OF COMPUTERIZED ELECTROCARDIOGRAPHY
Marquette, version 19). The computerized inter-
pretation was assessed against a consensus
interpretation by reading cardiologists. Overall,
13.2% of the rhythm interpretations made by
the computer were incorrect, and 7.8% if cardiac
pacemaker activity misinterpretation was
excluded.
The correct identification of pacemaker activ-
ity is still a major problem for today’s computer
software [44]. Poon and colleagues [39] reported
that 75.2% of all pacemaker rhythms were misin-
terpreted. Because pacemaker stimulus outputs
are very short in duration (generally 0.4–0.6 milli-
seconds), high sampling rates (O1000 Hz) are
needed for adequate detection. In addition, cur-
rent bipolar pacemaker systems are able to cap-
ture the myocardium with smaller amounts of
energy, in turn resulting in progressively smaller
pacemaker stimulus outputs that can be undetect-
able on the surface ECG. This reduction in stimu-
lus output is facilitated by autocapture algorithms
that allow the device to deliver its output just
above the capture threshold, achieving pacemaker
battery conservation while maintaining patient
Fig. 4 (continued )
501
safety. Current algorithms ‘‘write in’’ the pacing
output stimulus to make pacing activity apparent.
High-frequency components of the ECG, such as
pacemaker output stimuli, can sometimes disap-
pear when the tracing is reprinted after data has
been compressed for storage; the noncompressed
ECG has the best signal fidelity and is the most
appropriate for determining pacemaker rhythm.
Considerable improvement has been made [45],
however, especially in the detection of ventricular
pacemaker activity, and newer iterations of the
available algorithms are expected to have even
further enhanced ability to define ventricular
and atrial pacemaker stimuli.
There does appear to be progressive improve-
ment in rhythm analysis software. The Marquette
12SL ECG Analysis software program (GE
Health Care Technologies, Waukesha, Wisconsin)
has been the single most studied commercial
computer algorithm through its many versions
[22,23,29,31,38,39]. The most current version 20
has been found to need physician correction of
a rhythm interpretation in only 4.1% of cases
[44] compared with 7.8% reported with version
502 HONGO & GOLDSCHLAGER
19 [39]. Although interpretive software from ma-
jor manufacturers, such as GE Marquette and
HP Phillips, include more sophisticated program-
ming that specifically address difficult aspects of
rhythm analysis, such as pacemaker activity rec-
ognition and analysis of pediatric ECGs, contin-
ued development is still needed [38].
Future directions
There are many challenges if current comput-
erized ECG analysis programs are to be used to
their fullest capabilities. On an individual physi-
cian level, there needs to be a fuller understanding
of the strengths and limitations of the programs, so
that overreliance on the computer reading is
avoided. Continued improvement in the physi-
cian’s own ECG interpretative skills will aid in
recognition of computer misinterpretations, and
will help steer clear of inappropriate management
of patients. Teaching institutions are needed in this
process, not only by encouraging higher standards
Fig. 4 (continued )
in ECG reading skills, but also by incorporating
into their curricula instruction on the current
status of computerized electrocardiography.
On the level of institutions and industry,
medical centers are faced with the need to better
integrate computer ECG analysis with expert
physician overreading. Not only should comput-
erized interpretations be verified in a timely man-
ner that actually impacts clinical decision making,
but also expert ECG readers should be made
readily available for physicians at the point of
patient care, both in the inpatient and outpatient
setting [46]. Finally, developers are faced with the
continued challenge of improving the accuracy of
computer algorithms. ECG databases, used to test
newer programs, should be sufficiently large, di-
verse, and contain all clinical diagnoses. True
standardization within the industry should help
facilitate this process, and efforts should continue
toward this goal.
The reliability of computerized ECG analysis
programs will only continue to improve, and the
degree to which expert readers will continue to be
 STATUS OF COMPUTERIZED ELECTROCARDIOGRAPHY
needed remains to be seen. The importance of
understanding the technical aspects of computer-
ized electrocardiography may, therefore, wane as
computers become more accurate and physicians
find that computerized interpretations can be
trusted. In the meantime, until there is true
automated ECG analysis, computerized interpre-
tations should be viewed as an adjunct to, not
a substitute for, interpretation by a competent and
experienced physician.
Acknowledgments
The authors thank Dr. E. William Hancock,
Professor of Medicine at Stanford University, and
Dr. Paul D. Kligfield, Professor of Medicine at
Cornell University, for their expert insights during
the preparation of this manuscript.
References
[1] Pipberger HV, Arms RJ, Stallman FW. Automatic
screening of normal and abnormal electrocardio-
grams by means of a digital electronic computer.
Proc Soc Exp Biol Med 1961;106:130–2.
[2] Caceres CA, Steinberg CA, Abraham S, et al. Com-
puter extraction of electrocardiographic parameters.
Circulation 1962;25(2):356–62.
[3] Bonner RE, Crevasse L, Ferrer MI, et al. A new
computer program for analysis of scalar electrocar-
diograms. Comput Biomed Res 1972;5(6):629–53.
[4] Holst H, Ohlsson M, Peterson C, et al. A confident
decision support system for interpreting electrocar-
diograms. Clin Physiol 1999;19(5):410–8.
[5] Kaiser W, Faber TS, Findeis M. Automatic learning
of rules: a practical example of using artificial intel-
ligence to improve computer-based detection of
myocardial infarction and left ventricular hypertro-
phy in the 12-lead ECG. J Electrocardiol 1996;
29(Suppl):17–20.
[6] Willems JL, Abreu-Lima C, Arnaud P, et al. Evalu-
ation of ECG interpretation results obtained by
computer and cardiologists. Methods Inf Med
1990;29(4):308–16.
[7] Pipberger HV, Arzbaecher RC, Berson AS, et al.
Recommendations for standardization of leads and
of specifications for instruments in electrocardiogra-
phy and vectorcardiography: report of the Commit-
tee on Electrocardiography. Circulation 1975;52:
11–31.
[8] Bailey JJ, Berson AS, Garson A Jr, et al. Recommen-
dations for standardization and specifications in au-
tomated electrocardiography: bandwidth and digital
signal processing. A report for health professionals
by an ad hoc writing group of the Committee on Elec-
trocardiography and Cardiac Electrophysiology of
503
the Council on Clinical Cardiology, American Heart
Association. Circulation 1990;81(2):730–9.
[9] Willems JL, Arnaud P, van Bemmel JH, et al. Com-
mon standards for quantitative electrocardiography:
goals and main results. Methods Inf Med 1990;29(4):
263–71.
[10] Burchell HB, Reed J. A test experience with a
machine-processed electrocardiography diagnosis:
the recognition of ‘‘normal’’ and some specific pat-
terns. Am Heart J 1976;92(6):773–80.
[11] Drazen E, Mann N, Borun R, et al. Survey of com-
puter-assisted electrocardiography in the United
States. J Electrocardiol 1988;21(Suppl):S98–104.
[12] Grauer K, Kravitz L, Curry R Jr, et al. Computer-
ized electrocardiogram interpretations: are they use-
ful for the family physician? J Fam Pract 1987;24(1):
39–43.
[13] Kadish AH, Buxton AE, Kennedy HL, et al. ACC/
AHA clinical competence statement on electrophys-
iology and ambulatory electrocardiography. Circu-
lation 2001;104(25):3169–78.
[14] Willems JL, Arnaud P, van Bemmel JH, et al.
A reference database for multilead electrocardiogra-
phic computer measurement programs. J Am Coll
Cardiol 1987;10(6):1313–21.
[15] Goldberger AL, Bhargava V, Froelicher V, et al. Ef-
fect of myocardial infarction on high-frequency
QRS potentials. Circulation 1981;64(1):34–42.
[16] Pettersson J, Warren S, Mehta N, et al. Changes in
high-frequency QRS components during prolonged
coronary artery occlusion in humans. J Electro-
cardiol 1995;28(Suppl):225–7.
[17] Pettersson J, Carro E, Edenbrandt L, et al. Spatial,
individual, and temporal variation of the high-fre-
quency QRS amplitudes in the 12 standard electro-
cardiographic leads. Am Heart J 2000;139(2 Pt 1):
352–8.
[18] Garson A Jr. Clinically significant differences be-
tween the ‘‘old’’ analog and the ‘‘new’’ digital elec-
trocardiograms. Am Heart J 1987;114(1 Pt 1):194–7.
[19] Taha B, Reddy S, Xue Q, et al. Automated discrim-
ination between atrial fibrillation and atrial flutter in
the resting 12-lead electrocardiogram. J Electrocar-
diol 2000;33(Suppl):123–5.
[20] Willems JL. A plea for common standards in com-
puter aided ECG analysis. Comput Biomed Res
1980;13(2):120–31.
[21] McLaughlin SC, Aitchison TC, Macfarlane PW.
The value of the coefficient of variation in assessing
repeat variation in ECG measurements. Eur Heart
J 1998;19(2):342–51.
[22] Farb A, Devereux RB, Kligfield P. Day-to-day var-
iability of voltage measurement used in electrocar-
diographic criteria for left ventricular hypertrophy.
J Am Coll Cardiol 1990;15(3):618–23.
[23] Kudenchuk PJ, Ho MT, Weaver WD, et al. Accu-
racy of computer-interpreted electrocardiography
in selecting patients for thrombolytic therapy.
J Am Coll Cardiol 1991;17(7):1486–91.
504 HONGO & GOLDSCHLAGER
[24] Shirataka M, Miyahara H, Ikeda N, et al. Evalua-
tion of five computer programs in the diagnosis of
second-degree AV block. J Electrocardiol 1992;
25(3):185–95.
[25] Teppner U, Lobodzinski S, Neuberg D, et al. A tech-
nique to evaluate the performance of computerized
ECG analysis systems. J Electrocardiol 1987;
20(Suppl):S68–72.
[26] Willems JL, Abreu-Lima C, Arnaud P, et al. The di-
agnostic performance of computer programs for the
interpretation of electrocardiograms. N Engl J Med
1991;325(25):1767–73.
[27] Bogun F, Anh D, Kalahasty G, et al. Misdiagnosis
of atrial fibrillation and its consequences. Am
J Med 2004;117(9):636–42.
[28] Salerno SM, Alguire PC, Waxman HS. Competency
in interpretation of 12-lead electrocardiograms:
a summary and appraisal of published evidence.
Ann Intern Med 2003;138(9):751–60.
[29] Grauer K, Kravitz L, Ariet M, et al. Potential bene-
fits of a computer ECG interpretation system for pri-
mary care physicians in a community hospital. J Am
Board Fam Pract 1989;2(1):17–24.
[30] Hillson SD, Connelly DP, Liu Y. The effects of com-
puter-assisted electrocardiographic interpretation
on physicians’ diagnostic decisions. Med Decis
Making 1995;15(2):107–12.
[31] Goodacre S, Webster A, Morris F. Do computer
generated ECG reports improve interpretation by
accident and emergency senior house officers? Post-
grad Med J 2001;77(909):455–7.
[32] Xue Q, Reddy S. Algorithms for computerized QT
analysis. J Electrocardiol 1998;30(Suppl):181–6.
[33] Azie NE, Adams G, Darpo B, et al. Comparing
methods of measurements for detecting drug-
induced changes in the QT interval: implications
for thoroughly conducted ECG studies. Ann Nonin-
vasive Electrocardiol 2004;9(4):166–74.
[34] Phibbs BP, Marriott HJL. Computer ECG pro-
grams: a critical evaluation. Primary Cardiol 1981;
7:49–60.
[35] Brailer DJ, Kroch E, Pauly MV. The impact of com-
puter assisted test interpretation on physician deci-
sion making: the case of electrocardiograms. Med
Decis Making 1997;17(1):80–6.
[36] Salerno SM, Alguire PC, Waxman HS. Training and
competency evaluation for interpretation of 12-lead
electrocardiograms: recommendation from the
American College of Physicians. Ann Intern Med
2003;138(9):747–50.
[37] Sridharan MRL, Flowers NC. Computerized elec-
trocardiographic analysis. Mod Concepts Cardio-
vasc Dis 1984;53:37–41.
[38] Snyder CS, Fenrich AL, Friedman RA, et al. The
emergency department versus the computer: which
is the better electrocardiographer? Pediatr Cardiol
2003;24(4):364–8.
[39] Poon K, Okin PM, Kligfield P. Diagnostic perfor-
mance of a computer-based ECG rhythm algorithm.
J Electrocardiol 2005;38(3):235–8.
[40] Romhilt DW, Bove KE, Norris RJ, et al. A critical
appraisal of the electrocardiographic criteria for
the diagnosis of left ventricular hypertrophy. Circu-
lation 1969;40(2):185–95.
[41] Devereux RB, Casale PN, Eisenberg RR, et al. Elec-
trocardiographic detection of left ventricular hyper-
trophy using echocardiographic determination of
left ventricular mass as the reference standard. Com-
parison of standard criteria, computer diagnosis and
physician interpretation. J Am Coll Cardiol 1984;
3(1):82–7.
[42] Casale PN, Devereux RB, Alonso DR, et al. Im-
proved sex-specific criteria of left ventricular hyper-
trophy for clinical and computer interpretation of
electrocardiograms: validation with autopsy find-
ings. Circulation 1987;75(3):565–72.
[43] Varriale P, David W, Chryssos BE. Multifocal atrial
arrhythmiada frequent misdiagnosis? A correlative
study using the computerized ECG. Clin Cardiol
1992;15(5):343–6.
[44] Farrell RM, Xue JQ, Young BJ. Enhanced rhythm
analysis for resting ECG using spectral and time-do-
main techniques. Comput Cardiol 2003;30:733–6.
[45] Helfenbein ED, Lindauer JM, Zhou SH, et al.
A software-based pacemaker pulse detection and
paced rhythm classification algorithm. J Electrocar-
diol 2002;35(Suppl):95–103.
[46] Hongo RH, Goldschlager N. Overreliance on com-
puterized algorithms to interpret electrocardio-
grams. Am J Med 2004;117(9):706–8.
 Cardiol Clin 24 (2006) ix

Foreword

Michael H. Crawford, MD
Consulting Editor

Dr. Webb has organized a unique issue on the
adult with congenital heart disease (CHD). Dr.
Webb runs a multidisciplinary CHD unit in Phil-
adelphia that is a model for other such units.
Thus, I was delighted that he had agreed to guest
edit this issue. His unique perspective is evident in
the first article, which presents the patient’s
perspective on the care of adults with CHD. Al-
though this may seem strange, patient involve-
ment is a critical part of most adult CHD units.
Our unit at University of California, San Fran-
cisco and the one at Stanford University recently
held a continuing medical education conference
on adult CHD and invited our patients. More
patients showed up than caregivers.
Much of this issue explores the concept of the
multidisciplinary approach that these patients
require. There are articles on anesthesia, surgical,
interventional, genetic, obstetric, and psychologi-
cal issues surrounding the care of these patients.
One major issue covered is the orderly transition
from pediatric to adult cardiologist. In my expe-
rience, this is one of the more contentious issues
with these patients. There is an understandable
tendency for pediatric cardiologists to hold onto
these patients well into adulthood, which in some
circumstances, is in the patient’s best interest. On
the other hand, many adult cardiologists believe
that pediatric patients should be transferred to
them at the first signs of puberty. This also can
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.09.001
make sense because of the new risks of venereal
disease, pregnancy, drug abuse, and so forth, that
adolescence brings.
Finally, the last articles are on two important
topics. Tetralogy of Fallot is the most common
cyanotic heart disease and is usually repaired in
early childhood. Currently, post repair pulmonic
regurgitation is a major issue that is fully dis-
cussed by Redington. Besides survival, the other
major goal of care in CHD patients is their ability
to lead normal lives. The latter requires an ability
to be active. Thus, the last article, on exercise
capacity in CHD patients, is very topical.
I believe that the readers will agree that this is
an outstanding issue by noted experts in the field
from both sides of the Atlantic that will provide
those caring for adult CHD patients with consid-
erable practical knowledge and that will help them
organize their own adult CHD units.
Michael H. Crawford, MD
Division of Cardiology
Department of Medicine
University of California
San Francisco Medical Center
505 Parnassus Avenue, Box 0124
San Francisco, CA 94143-0124, USA
E-mail address: crawfordm@medicine.ucsf.edu
cardiology.theclinics.com
 Cardiol Clin 24 (2006) xi–xii

Preface

Gary D. Webb, MD
Guest Editor

This issue of Cardiology Clinics is devoted to
the growing field of adult congenital heart disease
(ACHD). The central roles of cardiologists, car-
diovascular surgeons, and nurses are well under-
stood. In recent years in the United States, there
has been a major growth in the number of centers
caring for adult patients with congenital heart de-
fects. There are now approximately 60 ACHD
centers in the United States, many of which have
just started up in the past few years. What role
should these centers play in the care of ACHD
patients?
The first article offers an important and unique
perspective, that of the ACHD patient. The lead
author is Amy Verstappen, president of the Adult
Congenital Heart Association, an important
ACHD advocacy organization. For the first time
in the medical literature, she and her colleagues
articulate the perspectives of the ACHD patient.
Issues discussed include the impact of overly
pessimistic and overly optimistic prognoses, com-
mon patient misperceptions and knowledge gaps,
frustrations and dangerous encounters in the med-
ical system, and living with invisible disabilities.
One of the implications of establishing an
ACHD clinic is that these patients, many of
whom have quite complex anatomy and physiol-
ogy, require multidisciplinary care. How does an
ACHD center develop multidisciplinary capabil-
ity? What are the elements of multidisciplinary
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.09.002
care that need to be considered? What can
consultants and team members be expected to
contribute to the care of these patients?
The second article defines the role of cardiac
catheterization in a modern ACHD facility. The
authors discuss the planning of the procedure, the
potential pitfalls to be anticipated, and the equip-
ment needed for the procedure. They then discuss
the performance of the cardiac catheterization.
They go on to outline the current role of heart
catheterization in the era of advanced noninvasive
imaging, with a focus on interventional proce-
dures. Finally, the authors look at new and
emerging interventions and speculate on the
future of diagnostic heart catheterization in the
ACHD patient.
The next multidisciplinary topic considers the
role of the geneticist in an ACHD center, sum-
marizing the clinical and molecular advances
relevant to the care and genetic counseling of
ACHD patients and exploring the role of genetic
care providers in such clinics.
Although the key role of the cardiovascular
surgeon is well understood, we are reminded that
a strong surgical program requires a multidisci-
plinary team of its own. The next articles focus on
the role of the anesthesiologist in caring for
ACHD patients. The authors review the preoper-
ative assessment of the ACHD patient; the
pharmacology of anesthetic agents; and the
cardiology.theclinics.com
xii
anesthesiologist’s role in noncardiac surgical pro-
cedures, diagnostic and interventional procedures,
and congenital heart surgery on adult patients.
ACHD patients are relatively young, so the
management of the pregnant patient is an impor-
tant part of ACHD patient care. The group from
the Royal Brompton Hospital describes their
model for meeting this challenge using a compre-
hensive team approach that includes evaluation
and communication with a team of trained and
experienced specialists, including cardiologists,
obstetricians, anesthetists, pediatricians, clinical
nurse specialists, and clinical geneticists.
The next article focuses on the role of the
psychologist in ACHD patients. The authors
remind us that patients benefit from having their
caregivers recognize and manage the potential
psychosocial consequences of growing up with
congenital heart disease. The role of the psychol-
ogist integrated into multidisciplinary ACHD care
teams is reviewed in relation to the provision of
clinical services, multidisciplinary research, and
professional education.
The transition process prepares children and
adolescents with congenital heart defects to learn
how to manage their own health care as they enter
adult life. The transfer of adolescent patients to
adult care facilities should follow such a transition
process wherever competent adult care is avail-
able. The next article focuses on these issues. It
outlines the key elements of a successful transition
program, and provides a general curriculum
appropriate for the young adult with congenital
heart disease. It also identifies barriers to the
transfer of care, discusses the importance of
a policy on the timing of transfer, and reviews
the steps to an orderly transfer process.
The last two articles depart from the prevailing
theme. The first deals with the critically important
PREFACE
issue of pulmonary regurgitation in patients who
have had repairs of tetralogy of Fallot. Andrew
Redington takes a fresh and critical look at this
very important topic. He describes the primary
determinants of pulmonary regurgitation, the
methods by which pulmonary regurgitation is
measured, and some of the pitfalls in the assess-
ment of this condition. He points out that
although our ability to measure pulmonary in-
competence has improved significantly, there re-
main important pitfalls to its assessment in the
individual patient, and that our understanding of
its implications remains rather unsophisticated.
The final article represents a unique and
important synthesis of our knowledge about
exercise intolerance in ACHD patients. The au-
thors discuss the assessment of exercise intoler-
ance, the mechanisms of reduced exercise
capacity, neurohormonal activation, the prognos-
tic value of parameters of exercise intolerance,
and the means of improving the exercise capacity
of ACHD patients.
It has been a pleasure for me to have worked
with these excellent colleagues to create an issue
presenting important and unique new perspectives
that can be used to improve the care of adult
patients with congenital heart defects. We antic-
ipate considerable progress in the next decade,
and hope that the information provided here will
help guide the growth and development of high-
quality ACHD programs.
Gary D. Webb, MD
Philadelphia Adult Congenital Heart Center
6 Penn Tower
3400 Spruce Street
Philadelphia, PA 19104-4283, USA
E-mail address: Gary.webb@uphs.upenn.edu
 Cardiol Clin 24 (2006) 515–529

Adult Congenital Heart

Disease: the Patient’s Perspective
Amy Verstappen, MEda,*, Disty Pearson, PA-Cb,
Adrienne H. Kovacs, PhDc
a
Adult Congenital Heart Association, 6757 Greene Street, Philadelphia, PA 19119, USA
b
Boston Adult Congenital Heart (BACH) Service, Brigham and Woman’s Hospital,
Departments of Cardiology, 300 Longwood Avenue, Boston, MA 02115, USA
c
Cardiac Psychology, Division of Cardiology, University Health Network,
399 Bathurst Street, 1-West-414, Toronto, ON M5T 2N2, Canada

In addressing the patient’s perspective in con-
genital heart disease (CHD), we must first ac-
knowledge that there are many patient
perspectives. The wide variability in diagnoses,
age at diagnosis, treatment, and outcomes that
makes CHD medically complex also results in
a highly heterogeneous patient population and set
of life experiences. Although there have been
attempts to correlate specific defects with specific
psychosocial outcomes, patients born with the
same anatomy can have widely varied experiences.
For example, consider three 40-year-old patients
who have atrial septal defects. One patient,
surgically repaired as an infant, considers herself
cured, has not seen a cardiologist in 35 years, and
is asymptomatic. The second is newly diagnosed,
having developed symptoms after the birth of her
second child, and is now awaiting repair. The
third, born in a rural area and not referred for
closure, has experienced life-long disability, de-
veloped Eisenmenger’s syndrome, and has limited
treatment options. Although the underlying defect
is the same, the experience and impact of this
anatomy is profoundly different in each case.
Amid this broad diversity, however, certain
patterns of experience and perspective emerge,
and these are known to the Adult Congenital
Heart Association (ACHA), the only adult
* Corresponding author.
E-mail address: amyv@achaheart.org
(A. Verstappen).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.004
congenital heart disease (ACHD) patient advo-
cacy group in the United States. Although the
existing educational, psychosocial, and quality-of-
life research is referenced in this article, the primary
focus is on patient reports, drawn primarily from
ACHA participant experiences. ACHD literature
contains little that is written by the patients
themselves. When ACHD patients share experi-
ences with each other, many of the issues and
challenges identified fall outside of the existing
published literature.
To illustrate these perspectives, the authors
provide quotations drawn from the ACHA
discussion board at the Web site achaheart.org,
which currently has over 900 registered partici-
pants with over 14,000 posted discussion topics.
The authors also draw from their own experi-
ences working with the adult congenital heart
community as ACHD health advocate and peer
support leader, midlevel health professional, and
psychologist; the lead author is also informed by
her own experiences as a survivor of complex
CHD. The authors recognize that this approach
is largely subjective and anecdotal; however, by
offering this alternate perspective in which the
primary concerns identified by the patients are
addressed, the authors hope to prompt new
research and new solutions.
The primary goals of this article are (1) to
provide readers with ‘‘patients’-eye views’’ of the
challenges of living with CHD and (2) to suggest
implications for health care professionals. In
cardiology.theclinics.com
516 VERSTAPPEN
selecting topics, the authors particularly focused
on areas in which the gap between the patient and
medical perspective is most likely to negatively
impact health care delivery and patient well-being.
The focus initially is on themes with a more
medical focus, including the impact of being told
one was ‘‘cured,’’ the impact of growing up as
a ‘‘miracle baby,’’ medical misinformation among
patients, and commonly reported frustrations
when navigating the medical system. The authors
also address concerns raised by living with an
invisible disability and confronting new health
challenges. In addition to these more medical
themes, the question of normalcy and the ways
in which CHD can be a ‘‘gift’’ are also addressed.
The article concludes with a discussion of the
importance of ACHD patient associations. Each
section identifies recommended best practices for
the health care providers, and these recommen-
dations are summarized in Box 1.
Issue 1: the myth of ‘‘the cure’’
[Patient 1] In 1973, when I was 17 years old, my
cardiologist ended my annual visit by saying, ‘‘I
don’t need to see you anymore.’’ While I don’t
recall his exact words, the essence was, ‘‘We’ve
seen you through adolescence. That was the
worrisome part. Now go live your life.’’
[Patient 2] I had a follow-up cath when I was five
years old and the doctors told my parents that the
result of the surgery was excellent, so I was
‘‘fixed.’’ The terms total correction and fixed
were used frequently by the doctors and they
were used both synonymously and interchangea-
bly.I saw an adult cardiologist for an evaluation
at puberty [and].did not see a cardiologist for
between twenty-five and thirty years.I had an
EKG annually. When I started with my current
primary care practitioner in 1996, he started do-
ing echoes every two years to ‘‘monitor’’ my heart
as I was getting older.. But, like everyone before
him, he said that I was ‘‘fixed.’’
Denial, lack of compliance, transition difficul-
ties, and lack of education about one’s defect are
barriers to care commonly referenced in the
cardiac literature. An additional commonly re-
ported reason for no longer seeing a cardiologist,
however, is being explicitly graduated from car-
diac care. Many complex patients, particularly
those who underwent surgery before 1980, report
being told that they no longer needed to see
a cardiologist, and many adults who have CHD
report believing themselves ‘‘cured’’ following
childhood or adolescent surgeries.
et al
Box 1. Recommended best practices
for interactions with adults who have
congenital heart disease
Practices to avoid
 Using terms such as fixed, cured, or
complete repair
 Withholding known negative
information
 Making predictions unsupported by
data (eg, ‘‘This is your last surgery,’’
‘‘You won’t live to see thirty’’)
 Relying on patient reports of normal
functional capacity without objective
testing
 Asking patients to compare
themselves to ‘‘normal people’’
 Confusing quality of life with health
status
Practices to promote
 Educating for life-long self-advocacy
and self-care
 Sharing data or acknowledging lack of
data
 Encouraging optimism and planning
for the future
 Providing copies of medical records
 Assessing function through serial
testing
 Soliciting and respecting the patient’s
point of view
 Providing mental health resources
 Providing access to peer support
These reports should not be surprising given
the lack of knowledge about long-term ACHD
outcomes at the time and the commonly held
belief that these surgeries were curative. What is
more surprising is that many adult patients report
that the misperception that they are ‘‘fixed’’ has
been reinforced rather than corrected by their
current health providers. The 32nd Bethesda
Conference (‘‘Care of the Adult with Congenital
Heart Disease’’) categorized cardiac defects as
being of mild, moderate, or great complexity
(Appendix 1) [1]. Patients of moderate and great
complexity should receive regular care at a spe-
cialty adult congenital cardiology program. For
the purposes of this article, the authors refer to
patients who have defects of moderate and great
complexity as ‘‘complex.’’ When complex patients
 PATIENT’S PERSPECTIVE
are cared for by primary care providers without
being referred to even a community-level cardiol-
ogist, this lack of referral can itself provide further
evidence to the patient that they are ‘‘cured.’’
Another issue that bears further examination is
the terminology of many congenital cardiac sur-
geries. Adult patients often refer to phrases such as
‘‘complete surgical correction’’ or ‘‘total surgical
repair’’ to explain their perception of being fixed.
To those unfamiliar with the complexities of CHD,
these phrases can imply that the heart is ‘‘totally
corrected’’ (ie, cured). ACHA’s current experience
with the pediatric congenital heart community
suggests that despite the extensive data on residua
and sequelae in complex CHD now available [2–4],
today’s children may still be at risk from language-
based misperception. For example, in ACHA
encounters with survivors of the arterial switch
for transposition of the great arteries, parent state-
ments that the child now has ‘‘normal heart anat-
omy’’ because she or he underwent an ‘‘anatomic
repair’’ are common; few understand that an arte-
rial switch, like virtually all complex CHD surgery,
leaves behind abnormal hemodynamics and anat-
omy [5]. Such phrases may also contribute to the
general community provider’s lack of awareness
of ACHD as a chronic disease. Modifying the
terminology in CHD surgeries might reduce the
misperception of cure and more accurately reflect
current knowledge of complex CHD. As indicated
in Appendix 1, few surgeries are considered
curative.
The possibility that many complex ACHD
patients perceive themselves as graduates of car-
diac care is supported by data suggesting that only
a minority of adults living with CHD continue to
seek appropriate levels of care. For example, in the
Natural History Study, 40% of patients studied
had not had a cardiac examination in over 10 years
[6]. Similarly, a study from the Canadian Adult
Congenital Heart Network Consortium reported
that only 37% to 47% of patients had successfully
transitioned from pediatric cardiology care to
adult medicine; 27% had not been evaluated since
turning 18 years old [7]. This rate of transition is
expected to be lower in the United States given
the more organized nature of the Canadian care
system, the frequent transfer of patients to adult
CHD clinics at age 18, and the lack of financial
barriers to medical care. In 2005, 59 self-described
American ACHD clinics responding to an ACHA/
International Society for Adult Congenital Car-
diac Disease (ISACCD) survey reported a com-
bined ACHD patient population of fewer than
517
34,000 per year; this number is likely an overesti-
mate given that patient numbers were self-
reported, often estimated, and included patients
treated at more than one center. This survey indi-
cates that a very small percentage of the estimated
500,000 American complex ACHD patients are
being treated at an ACHD clinic.
In addition, until a more representative sample
of the patient population is available for study, the
generalization of research conclusions is limited
due to biases in sample selection. The experiences
of the minority of ACHD patients followed in
pediatric and specialized ACHD facilities might
not reflect the entire ACHD patient population.
For example, these patients may be more likely to
have health issues than those who perceive them-
selves as fixed and have left cardiac care. Popula-
tion-based studies are essential to developing an
accurate understanding of the long-term outcomes
in ACHD, and ACHA strongly endorses the
development of a national ACHD registry.
Implications for health care providers
Individuals working with pediatric and adult
patients must avoid misleading language and
clearly communicate to patients and families
that those who have CHD, by definition, do not
have normal hearts and those who have more
complex CHD need life-long specialized cardiac
care. Individuals working with adult patients must
also acknowledge the possibility that many adults
who have complex CHD consider themselves
graduated from cardiac care due to misinforma-
tion provided by past and present health care
providers. This acknowledgment calls for different
action than if one sees ACHD patients as ‘‘in
denial’’ because denial implies that the patient
understands on some level that she or he needs
cardiac care. Providing the best possible care to
congenital heart survivors demands reaching out
beyond the existing patient base and working with
patient groups and others to promote community
outreach, patient identification, and education
regarding appropriate referral. The emerging
consensus in congenital heart care is that health
surveillance matters; delays in care can result in
needless disability and loss of life.
Issue 2: ‘‘miracle babies’’ and ‘‘lost causes’’
[Patient 1] When I was three months old, I was
diagnosed. The doctors looked at my parents and
said, ‘‘Take her home and love her, she won’t live
to be one.’’
518 VERSTAPPEN
For ACHD patients born in the era when
congenital heart surgery was in its infancy, the
experience of having been a ‘‘miracle baby,’’
although rarely referenced in the ACHD litera-
ture, can also be prominent in many patients’
self-identity. When ACHD patients meet, one
common topic of conversation is sharing these
early experiences. For many patients too young
to remember these encounters when they oc-
curred, these stories are not memories but rather
retellings of frequently recounted family lore.
These stories typically begin with the statement
that the child was not expected to survive and
then include the extraordinary efforts the family
went through to gain their child access to what
was then rare, difficult to access, and sometimes
extremely risky surgery. Even patients born
when interventions had become more routine
are often able to recount the moment that their
parents were told their child would not live
without surgery. For those who experienced
‘‘complete repairs,’’ these stories typically end
with the miracle of their surgery and successful
survival.
In contrast, ACHD patients ineligible for
surgery or only ‘‘partially repaired’’ often report
the opposite experience: on-going, explicit state-
ments by medical professionals that they would
not live long.
[Patient 2] After my Mustard, the surgeon said
I’d never live to see my fifth birthday. When I was
diagnosed with CHF, I was given three years to
live (‘‘five, if you’re lucky’’). I was sixteen, and
my pediatric cardiologist thought I wouldn’t
make it through my sophomore year in college.
I didn’t find out until years later that he called my
mom that afternoon and apologized for not being
able to save me.
Not surprisingly, these kinds of predictions
can result in a fear of death and difficulty
planning for the future. Decisions such as pursu-
ing higher education, working toward a career,
establishing savings, or even developing long-term
relationships may be avoided if one does not
anticipate living into adulthood. At the extreme,
some adults who have CHD reported engaging in
highly risky behavior such as illegal drug use
because they believed they did not have a future
to put at risk. When these individuals do not die
as expected, they may struggle with the migration
into an adulthood for which they did not prepare.
More surprisingly, many ACHD patients re-
port that rather than bringing on feelings of
et al
despair or passivity, statements of expected non-
survival functioned as a challenge to prove the
experts wrong.
[Patient 3] I adapted the ‘‘I’ll show them’’ attitude.
I can do whatever I want no matter what anyone
else tells me because the only person who limits me
is me. So I celebrate every birthday with the
knowledge that I am not supposed to be here and
with the determination to be here to celebrate the
next one.
ACHA participants with life-long significant
cardiac disability frequently report taking on
professional training and education despite oppo-
sition from family or medical professionals, and
many achieve high levels of independence and
career success.
Implications for health care providers
For patients and families who perceive pre-
vious surgeries as evidence of a ‘‘medical miracle,’’
learning that they have new health problems can
be particularly difficult. Conversely, ACHD pa-
tients raised with the expectation that they would
not live long may have an inappropriately nega-
tive view of their expected life span: many are
surprised and encouraged to learn that the re-
ported median age of survival in patients who
have Eisenmenger’s syndrome is into the sixth
decade [8]. Those who have repeatedly heard neg-
ative health predictions that did not come true
may be justifiably skeptical of new information
provided by medical professionals. As a result,
health care providers should avoid assuming
that patients who appear reluctant to accept
health information are in denial or unable to
comprehend health risks. Care must be taken to
establish a supportive and collaborative patient–
physician relationship in which patient experi-
ences and medical advice are valued.
Issue 3: facing the facts and the unknown: patient/
physician communication and education
[Patient 1] The reason that I thought that I was
‘‘fixed’’ came more from what was not said and
because of the old phrase ‘‘no news is good
news’’.. It was well known at the time of surgery
at the age of six that I may need future surgery,
but I was never told, even when I became an
adult. I was also never told that things may get
worse at all. I was treated as if my heart was not
perfect, but there was nothing to worry about.
What was said at doctor’s appointments was that
 PATIENT’S PERSPECTIVE
everything was fine and that I would need to
return in a year or two.
[Patient 2] I wish someone had sat down with me
and said, ‘‘This is what your heart defect is. These
are the consequences and what we are watching
for. This is what we know and this is what we
don’t know. This is what we are doing to help
and this is what you can do to optimize your
life.’’
The provision of overly optimistic or overly
pessimistic outlooks results in suboptimal knowl-
edge among ACHD patients. ACHA participant
reports confirm the existing educational research
suggesting that very few adults who have CHD
have the knowledge and understanding necessary
to ensure appropriate health surveillance behav-
ior and to optimize their health. In their 2001
overview of existing research in patient educa-
tion, Moons and colleagues [9] reported that be-
tween 54% and 76% of patients are able to
name their cardiac defect and many ACHA
members recall being ‘‘drilled’’ on the name of
their diagnosis and previous surgeries by their
pediatric care providers, particularly as they
reached their teens. Very few, however, can
draw or describe their anatomy, and even fewer
can identify their specific cardiac issues and the
reasons they should return for follow-up care.
One study reported that 84% of patients cannot
identify the symptoms of heart failure and 92%
cannot identify symptoms of endocarditis. Al-
though many patients understand they are ex-
pected to return for follow-up care, Moons and
colleagues [9] noted that less than half identify the
potential for deterioration as a reason for their
follow-up visits.
Many factors contribute to the lack of com-
plete knowledge among ACHD patients. Al-
though formal data are lacking, ACHA
encounters with cardiology professionals suggest
that some health care providers have attitudinal
barriers to fully educating CHD survivors about
their risks and the need for life-long health
surveillance. An often-identified competing goal,
particularly in pediatric cardiology, is preventing
the potentially detrimental impact of overprotec-
tion and limited expectations. Many adults who
have complex CHD report strong positive at-
tachment to their pediatric care team based in
part on the positive messages they were given:
that they were ‘‘fine’’ and they should go out and
live a normal life. Many of these same patients,
however, took this message so literally that they
519
left cardiac caredin some cases for decades at
a time. Not all survived to return to care.
In working with pediatric and adult cardiolo-
gists, ACHA’s suggestions to refrain from using
words like ‘‘fixed’’ have been met with the
assertion that the need to avoid undue anxiety
outweighs the need to prepare patients for possi-
ble problems that may lie well in the future, and
physicians state that information should be given
on an as-needed basis. These responses, particu-
larly in the adult context, raise serious ethical
concerns. According to the Code of Medical
Ethics of the American Medical Association [10]:
It is a fundamental ethical requirement that
a physician should at all times deal honestly
and openly with patients. Patients have a right to
know their past and present medical status and to
be free of any mistaken beliefs concerning their
conditions.
Although the preferences of ACHD patients
have not been studied, patient preference for full
disclosure of information regarding other health
conditions such as an Alzheimer’s disease di-
agnosis is well documented [11]. It is an accepted
concept in bioethics that except in cases in which
the patient indicates that she or he does not want
to know, respecting the dignity of the patient de-
mands fully sharing health information.
An additional factor directly inhibiting the
conveyance of accurate health information is the
absence of data on long-term outcomes. Few
issues engender as much passion in the ACHD
community as the need for more large-scale
studies examining long-term outcomes in CHD.
ACHD patients report on-going frustration at the
number of health questions to which the best
answer is ‘‘we don’t know,’’ particularly when
ACHA interactions make clear that the popula-
tion of complex CHD survivors available for
study is large and growing. It is essential that
current ACHD patients be identified and studied
so that patients and families can benefit from
evidence-based practice rather than hopes and
suppositions.
Parental responsibility for health care decision
making can also influence patient education.
During childhood and adolescence, parents are
usually considered the ‘‘holders of the informa-
tion,’’ and few patients recall a specific transition
of primary medical responsibility from parent to
patient. In some families, this transition may not
have occurred, and parental fear of risk and loss
520 VERSTAPPEN
of control, issues of guilt related to the etiology of
the defect, or the patient’s own desire to continue
within the ‘‘safety net’’ of parental responsibility
can result in a continued lack of independence.
Frequently, individuals grow up with a ‘‘child’s
understanding’’ of their cardiac condition (eg,
‘‘backwards heart,’’ ‘‘hole in the heart’’), which
may or may not be supplemented with more adult
terminology. Some ACHD patients report that
their parents’ knowledge was also minimal, which
may reflect the lack of CHD information at the
time or an earlier era of more paternalistic
doctor–patient relationships. When ACHD pa-
tients ask their parents why they did not tell them
more about their defect, typical responses include,
‘‘I didn’t know,’’ ‘‘I did not want to burden you,’’
or ‘‘well, since you were fixed, I didn’t think the
details were important.’’
Finally, patient behaviors can also contribute
to a lack of knowledge. Visits to health care
providers can be fraught with worry, and such
anxiety interferes with concentration and mem-
ory. Concerns that the worst fears will be realized
can interfere with the ability to focus on the
conversation at hand or to ask for information.
In the absence of accurate information re-
garding risks and symptoms specific to CHD,
many adults who have CHD focus on more widely
available information about coronary artery dis-
ease symptoms. Many ACHD patients report
relying on the absence or presence of chest pain
or pressure as their primary indicator of heart
health; some believe they are at very high risk of
heart attack. Concurrently, symptoms such as
fluid retention, racing or pounding heart, and
shortness of breath may be ignored or ascribed to
aging, weight gain, and anxiety. Discussing chil-
dren and young adults, Veldtman and colleagues
[12] stated, ‘‘Illness understanding is poor.and
many have an entirely wrong concept of their dis-
ease.’’ This lack of understanding also appears
prominently in the older ACHD community.
Implications for health care providers
Appropriately educating complex CHD survi-
vors requires honesty and humility on the part of
the health care providers. The myth of the ‘‘fix’’
and the emphasis on being ‘‘normal’’ must be
replaced with an acknowledgment of a life-long
health condition and the goal of empowering
patients to successfully negotiate life with complex
CHD. Although optimism and encouragement are
et al
essential, specific assurances such as statements
that future surgery is unlikely should be offered
only when there are strong supporting data, and
those who have undergone recently developed
procedures should understand that long-term out-
comes, by definition, are as yet unknown. Con-
versely, negative predictions regarding long-term
outcomes should also be made with an acknowl-
edgment that future medical progress is unknown.
Past experience in the treatment of CHD suggests
that on-going progress will continue to be made in
response to today’s known problems.
Appropriate ACHD patient education also
demands extensive, on-going time and resources,
a scarce commodity at many clinics in which
existing resources are already stretched thin by the
complex medical needs of these highly challenging
patients. Having an adequate number of midlevel
providers specifically trained in patient education,
such as nurses, nurse practitioners, or physician
assistants, should be considered a priority.
Although high-level technology and medical ex-
pertise is essential, equally essential to patient
well-being is the availability of relevant informa-
tion to ensure that they have the knowledge and
skills needed to optimize their health and to access
care appropriately.
ACHA has defined a ‘‘tool kit’’ of information
with which all ACHD patients should be equip-
ped (Box 2). All CHD survivors are required to
undertake life-long health surveillance behavior
and should be able to identify and describe their
defects and previous interventions, observe endo-
carditis prophylaxis guidelines if appropriate,
and obtain cardiac care and testing at recommen-
ded intervals. Other specific topics for education
might include how to identify and respond to
heart failure, arrhythmias, or endocarditis and
the known likelihood of future interventions (eg,
restenting or valve replacement). It is essential
that complex congenital heart defect survivors
be instructed on the importance of seeking spe-
cialized ACHD care and be given specific strate-
gies and criteria for finding such care. Unless
these instructions are made explicit, ACHA expe-
rience shows that many complex ACHD patients
will seek care from nonpediatric community-level
cardiologists or choose cardiologists based on
media ‘‘health report cards,’’ cardiac center adver-
tising, or name recognition. Patients should also
be provided with copies of essential health records
such as catheterization and surgical reports and
understand that because hospital systems are
 PATIENT’S PERSPECTIVE
Box 2. Adult Congenital Heart
Association educational ‘‘tool kit’’
 Ability to name and describe defect
and interventions
 Ability to recognize cardiac symptoms
 Understanding of need for on-going
care
 Information on how to find appropriate
care
 Understanding of risks particular to
defect(s)
 Understanding of risk of CHD
recurrence
 Knowledge of appropriate birth control
options
 Understanding of pregnancy risks/
special needs
 Understanding of appropriate exercise
activities
 Access to medical records
 Access to appropriate vocational
education
 Access to support and on-going
information
legally obligated to keep information for only
a limited number of years, it is in their best inter-
est to keep copies of their own health records.
Congenital heart defects are complex and
difficult to understand, and a single educational
discussion is rarely adequate; patient education
must be revisited as many times as needed until
a stable understanding is achieved. Patient fears
and anxieties should be directly solicited and
addressed. There must also be patience and
acknowledgment that individuals bring their
own limitations for mastering the information.
Families in which a parent is still managing health
care may require assistance from the health care
team to allow the adult patients to take control of
their own health care needs. Mastery may also be
limited by other barriers such as learning disabil-
ities, impaired cognitive function, and mental
illness.
When patients are fully educated about their
CHD, not only are they more likely to engage in
appropriate health surveillance behavior but they
can also use the information to guide life decisions
such as when and whether to have children and
what career to pursue. When the potential for
future problems is significant, patients may wish
521
to engage as early as possible in activities best
undertaken with optimal heart health, such as
third-world travel or applying for life insurance.
As stated by Feudtner [13], the benefits of appro-
priate patient education include improved quality
of life by increased knowledge and skills, im-
proved self-care, increased adherence to and effec-
tiveness of therapy, better monitoring for possible
complications, reduced physical morbidity and
risk of mortality, reduced psychosocial stress
and anxiety, and enhanced self-esteem, decision-
making capacity, and satisfaction with care.
Issue 4: negotiating the medical system
[Patient 1] I told them my left and right ventricles
were reversed. The cardiologist stated that this
did not matter since ‘‘the left and right ventricles
are the same.’’
[Patient 2] This is a recurrent conversation that I
have with medical staff, especially in the emer-
gency room:
‘‘Why are you here today’’
‘‘I have a congenital heart defect.’’
‘‘Okay, what is it?’’
‘‘A vascular ring.’’
‘‘Oh.okay (pause).so is it metal? When did
you have it put in?’’
[Patient 3], ‘‘How long have you had a congenital
heart defect, Mrs. J’’?
When adults who have complex CHD meet
and exchange stories, one recurring theme is the
misunderstanding and ignorance frequently en-
countered in the ‘‘regular’’ medical system, espe-
cially in the emergency room. Although patients
often recount these stories with humor, the re-
alization that one’s medical care is in the hands of
those who do not understand one’s medical
condition is terrifying. For many who have
complex CHD, establishing safe and effective
relationships with their noncardiac care providers
involves on-going education and conversations
with the ACHD specialists and the patient. For
example, adult patients commonly report the need
to be vigilant in rejecting inappropriate medica-
tions offered by primary care providers. In the
context of a medical emergency in the emergency
room, patients are faced with medical staff likely
to know little or nothing about CHD, and insist-
ing on special care can result in an adversarial
situation. For example, one cyanotic patient
reported that her increasingly agitated requests
for an air filter on her IV line did not result in
522 VERSTAPPEN
a filter but did result in the medical staff calling for
assistance from the hospital’s mental health team.
Implications for the health care provider
Patient reports support the importance of the
32nd Bethesda Conference recommendation that
all CHD patients be equipped to advocate for their
medical needs in urgent situations [14]. Given the
rarity and complexity of many CHD diagnoses,
ACHD patients should understand that outside
of the pediatric and specialized ACHD settings,
health care providers are likely to have little or no
knowledge of their anatomy and special needs. Al-
though adult patients may wish for a health care
system in which they could trust medical profes-
sionals to recognize and understand their diagno-
ses, this is not currently a realistic expectation.
Many patients who have CHD have specific
medical ‘‘red flags’’ unique to their defect. Two
common examples are the need for filters on
intravenous lines for those who have right to left
shunting across a defect, and the need for endo-
carditis testing before antibiotics are given when
there is fever of unknown origin. Patients ideally
should not only understand these needs but also
have the ability to explain them with confidence to
medical providers. Patients must serve as their
own advocates, at times cajoling health care
providers to take appropriate action. Ideally,
patients should have a cell phone number or
pager number to ensure 24-hour contact with
a member of their ACHD team. Even when health
records are available, they may not be consulted
or the diagnoses and interventions referenced may
not be recognized or understood. Reports of
resistance or reluctance by the emergency room
staff to contacting the ACHD program directly
are common. Handing a surprised emergency
room physician a cell phone connected to an
ACHD specialist can be quite effective.
In addition to primary care and emergency
room providers, ACHD clinic staff should be
prepared to educate the teams of providers during
hospitalizations. At times, escorting the patient
from site to site through hospitalizations is neces-
sary to ensure no errors occur at any step of the
way. In the words of Dr. Jane Somerville [15],
These are a precious group of patients.. Half the
avoidable deaths occurred in those who were
well, leading normal lives without symptoms or
with mild disability.errors of management are
being made at all levels in the care of the GUCH
[ACHD] patient.
et al
Issue 5: invisible disabilities
[Patient 1] What’s frustrating is that because you
don’t have some visible handicap like the loss of
limbs, people seem to think you’re faking, or they
forget you have it, or something. And who of us
wants to keep reminding people they have heart
problems? I hate saying it, period.
[Patient 2] When I have a bad day, my husband
accuses me of making it up.
Like other invisible disabilities, cardiac health
problems such as heart failure and arrhythmias
can produce a particular set of issues with
family and the larger community. ACHD pa-
tients commonly report family or friends who
imply that their problems are ‘‘in their head’’ or
insist that changes to work, childcare, or house-
hold responsibilities are not necessary. Reports
of difficulties at work related to misunderstand-
ings about a patient’s physical limitations are
also common. Because ACHD patients may
appear to be well, requests for time off or other
accommodations, such as adjustments in work
schedule or responsibilities, may be met with
skepticism or accusations of dishonesty. Many
ACHD patients using handicapped parking
stickers report being challenged by bystanders
who believe they are parking illegitimately. Re-
sponses to this particular situation can be
creative: one patient developed a false limp to
use in the parking lot, another started lifting her
shirt and showing her scars to all challengers.
Implications for health care providers
Health care providers are encouraged to ex-
plicitly ask patients about their support systems
and the level of understanding these people have
about the patients’ limitations and to address the
stresses these issues can cause. Patients report that
direct communication between spouse or family
members and care providers can be extremely
helpful, particularly when family members are
encouraged to ask questions and raise indepen-
dent concerns. Questions about workplace diffi-
culties should also be included; when appropriate,
supports such as workplace letters and documen-
tation of health status can be offered.
Issue 6: confronting mortality
[Patient 1] I found out how serious things were
when I changed doctors at the age of twenty-five
and found out I needed heart surgery again.I was
 PATIENT’S PERSPECTIVE
scared, angry, and felt completely alone. Since my
mother knew all along that I may need more
surgery and did not tell me, I did not trust her to
tell me the truth.. At the time I was seeing new
doctors and didn’t know exactly what to ask them.
[Patient 2] I was furiousdI kept saying, ‘‘but they
told me I would be fine!’’ and they kept saying,
‘‘You knew you had this condition. You’re
actually lucky to have lived this long.’’
For many ACHD patients, new health informa-
tion comes not in the context of a planned transi-
tion or educational process but at the onset of new
health problems. Despite the fact that most patients
have always known that they have CHD, the onset
of new problems can be analogous to the onset of
any adult-onset, new, potentially fatal condition.
Those who discover that information on future
health risks was withheld may have an additional
layer of anger and a sense of having been deceived
or betrayed. Particularly difficult situations include
facing difficult treatment decisions, referral for
surgery, or being told that ‘‘nothing more can be
done’’ [16]. Psychosocial research has found that
the manner in which adults who have CHD cope
with their disease is impacted by environmental fac-
tors (eg, friends and family), experiences with med-
ical treatment and health care providers, and
personal attributes (eg, whether they believe out-
comes reflect personal actions or external factors)
[17]. For participants in ACHA, a primary coping
strategy is often getting peer support from those
who have faced similar challenges, and many report
these interactions to be profoundly helpful.
Although CHD is best considered a chronic
condition, new problems can dramatically change
the patient’s experience and outlook. It is unfortu-
nate that the gap between patient and care provider
may become particularly acute at this time. Health
care providers may indicate annoyance with the
patient’s distress and lack of understanding or
imply that the patient ‘‘should have known’’ that
future health problems were likely.
Implications for health care providers
The congenital heart patient experiencing new,
serious health problems should be offered the same
compassion, education, and support resources that
would be offered adults newly diagnosed with
other life-threatening conditions. Medical profes-
sionals working with adults who have CHD are
encouraged to be particularly respectful and sen-
sitive during these times. Although health care
providers may be surprised by a patient’s lack of
523
previous knowledge, it is never appropriate to
imply that the patient has been in denial. At no
time should patients facing life-threatening health
problems be told by their health care team that
they are lucky or they should be grateful. It is
appropriate to refer patients who have significant
difficulties managing these situations to mental
health professionals who are able to provide
emotional support in addition to specific coping
strategies. Opportunities for peer support should
also be offered.
Issue 7: what is ‘‘normal’’?
[Patient 1] I grew up knowing I had brown eyes,
brown hair, and a backwards heart.
[Patient 2] I hate when they ask, ‘‘How do you
feel compared with a normal person?’’ I always
want to say, ‘‘How the hell would I know?’’
ACHD survivors report widely variable child-
hood experiences that were influenced by many
factors including health status, family styles and
structure, and personalities. Multiple psychosocial
studies, however, have suggested that feeling
‘‘different’’ is a common experience among pa-
tients who have CHD [16–20]. This feeling is not
surprising because the experiences of most of
these individuals are different from those of chil-
dren born without congenital heart defects. Exam-
ples of the experiences that many ‘‘heart kids’’
share include cardiology visits, cardiac testing,
taking medications, school activity restrictions,
hospitalizations, interventions, and having a scar.
Depending on health status, reported challenges
may also include frequent school absences, aca-
demic struggles due to absence or illness, or the
inability to participate in childhood activities
such as riding a bike or playing tag. For the small
minority of patients who experience on-going sig-
nificant cyanosis, reports of teasing and negative
attention brought on by having blue lips and club-
bed fingers are common; research by Horner and
colleagues [16] reported that most boys who
have cyanotic CHD describe teasing, rejection,
and distressing nicknames.
Research suggests that approximately one
fourth of adults who have CHD recall parental
overprotection during childhood and adolescence
[19,21,22], and the theme of parental overprotec-
tion is echoed in the recollections of some
ACHA members. It must be emphasized, however,
that this behavior does not reflect that of all or
even most parents. Congenital heart survivors
524 VERSTAPPEN
report a wide range of parenting styles, including
parents who were risk-taking, casual, or neglectful.
[Patient 3] Even when I was the ‘‘bluest kid in
class,’’ I played baseball, basketball, football, and
most other kids’ games. My mother was de-
termined that if I had a short time to live, she
would allow me to live it doing the things I
enjoyed doing.
[Patient 4] I was raised by narcissistic parents
who had little regard for anyone other than
themselves and what they felt like doing in the
moment. [I had] heart and lung problems in
a house full of heavy smokers and drinkers.
Such neglect can have devastating conse-
quences; for example, some patients who have
Eisenmenger’s syndrome struggle with the knowl-
edge that their parents’ failure to obtain prompt
medical care was a central factor in their compro-
mised health and limited options. ACHA partici-
pant experiences also suggest that given the right
education and support, even the most overpro-
tected ‘‘heart kid’’ can become an independent and
effective self-advocate.
Although many interviewees in qualitative
research studies describe striving for normality
[17,18], many of the ACHD patients encountered
by the ACHA express irritation with the word
normal, particularly in the context of their health
and experience. One of the first questions par-
ents of children who have heart defects typically
ask ACHD patients is, ‘‘Are you able to have
a normal life?’’ For those diagnosed as children,
however, having a heart defect is normal. For
some patients, their CHD has or will involve
many limitations and challenges; for others,
very few. If normal is defined as typical, then be-
ing normal is not an appropriate goal for those
who have complex CHD. Part of the matura-
tional process identified by many adults who
have CHD is coming to terms with the chal-
lenges presented by their heart defect and realiz-
ing that the concept of normal is an artifice and
having limitations is the human condition.
[Patient 5] We all are ‘‘dealt a hand’’ in lifed
heart defect, dyslexia, quick temper, tone deaf,
drug-addicted mother, shy, bossy, born poor,
diabetic, etcetera. Some of us just know sooner
than others what our cards are.
[Patient 6] I can’t.get caught up in the idea that
life has been ‘‘unfair’’ to me. Especially as there are
so many other ways life could have been so much
more unfair.. Do I wonder what it would be like
et al
to spring out of bed and go jogging? Sure! But we
all wonder about the things others have that we
don’t.
Although adults in the CHD cohort showed
significant physical dysfunction on objective test-
ing, Moons [23] found that adults who have CHD
assess their health status to be equal to that of
their heart-healthy peers. For some, this may be
because their ‘‘normal function’’ has always been
subnormal; for others, it may be that the dysfunc-
tion they are now experiencing developed over
time, with the patient unknowingly adjusting his
or her behavior to the new level of function. Over-
all, it suggests that most adults who have CHD
are content with their abilities and do not experi-
ence diminished physical capabilities as a loss.
Examining the language that CHD survivors
use to describe their condition also illuminates the
way that many see CHD as part of their self-
identity. Many adults who have CHD strongly
object to the use of the word disease to describe their
anatomy because it implies that they were ‘‘born
diseased.’’ For the layperson, disease often con-
notes contagion, and ‘‘heart disease’’ is often used
as a synonym for coronary artery disease. One of
the most common complaints of heart defect survi-
vors is being confused with elderly heart attack vic-
tims. For some, the term defect is also seen as
offensive because it implies that they themselves
are ‘‘defective.’’ It would be the rare patient who
has cancer or acquired heart disease who would
take criticism of their disease personally, but this
generalization from defective heart to defective
self illustrates that many of those who grew up
with CHD see their heart as an essential component
of themselves.
The most common term used by patients to
describe their heart is the more neutral phrase heart
condition, thereby referring to the heart’s static, an-
atomic reality. When patients encounter health
problems, a distinction is often made between the
heart defect itself and the consequences of the de-
fect (ie, heart failure, pulmonary hypertension, ar-
rhythmia, and endocarditis).
Patients who have experienced life-long signif-
icant cardiac disability may struggle with surgical
decisions that offer the potential of significant
functional improvement, often making comments
such as, ‘‘but I do fine’’ or ‘‘but I’m not that
blue.’’ Patients who experience significant func-
tional improvement (eg, after Fontan revision)
often comment that they had no idea that they
could feel this well.
 PATIENT’S PERSPECTIVE
Implications for health care providers
One challenge for health care providers and
educators is to use language that does not alienate
the patients but accurately reflects the reality of
their conditions and on-going health care needs.
A component of some patients’ rejection of the
term congenital heart disease may be lack of under-
standing or denial of the serious health risks and the
often progressive dysfunction that can come with
CHD. The usefulness of the term congenital heart
disease is limited by its generality; each defect has
its own unique anatomy and concerns, and most
health problems involved in CHD can be best dis-
cussed in the context of a specific defect. Explaining
things in defect-specific language (ie, ‘‘people born
with tetralogy of Fallot’’) reinforces a patient’s
knowledge of his or her own diagnosis and helps
support the key concept that all congenital heart
defects are different. In printed patient materials,
ACHA uses the term congenital heart defect, which
avoids the negative overtones potentially associ-
ated with the word disease and the vagueness and
potential euphemism of the word condition.
Given the well-documented inability of ACHD
patients to accurately assess their own physical
function, objective measurements of functional
capacity such as exercise testing with maximum
VO2 determination is essential. Discussions of
function should focus on relative changes or sta-
bility for the individual and should inform deci-
sions about the need for therapeutic changes.
Results discussed and compared with normal
function may be useful for some as general
markers but may raise significant issues for the in-
dividual and do not necessarily inform the discus-
sion in a frame of reference that enables the
individual to consider his or her stability or de-
cline in functional capacity. The importance of
reaching maximum potential rather than attaining
‘‘normality’’ must be emphasized. When discus-
sing potential interventions such as surgery, there
must be an open exchange between the provider
and the individual regarding reasonable expecta-
tions for results and improvement in functional
capacity. Conversations with others who have un-
dergone similar surgeries can help patients under-
stand potential functional outcomes.
Issue 8: the gift of congenital heart disease
[Patient 1] I like who I am and the people I’ve met
along the way. I am far less shallow and way more
empathetic than I would have been without CHD.
Last month when I went snowshoeing in the
525
mountains, the sky was bluer, the mountains
grander, and the company finer because I spent
last summer chained to the flatlands with arrhyth-
mia. I live in gratitude more because I know some
days are not so good. Life is far more precious
knowing the alternatives that I’ve faced at times.
[Patient 2] I know every day when I look in the
mirror and see my scars that I am a living miracle
and have been blessed by God more times than I
can count.
[Patient 3] My father.told me at an early age
that.‘‘you are different. The demands on you
throughout childhood and adulthood will be
different than everyone else’s. Therefore, this is
what we’re going to do.’’ He taught me that I
couldn’t do blue collar work. And because of
that, I needed to be better than the next person at
whatever I did. He taught me to invest in the
stock market at age seven. He taught me ac-
counting at age nine. I attended my first share-
holder’s meeting at age ten. He taught me to
exploit my assets and not dwell on my limitations.
He gave me an extra skill set that he did not give
my brother or sister because he felt the demands
on me would be higher.
Although we have focused on the concerns and
frustrations faced by many adults with CHD, it is
also important to highlight the fact that many
long-term survivors report benefits from their
experiences with CHD. Adults who have CHD
often report positive and negative childhood
associations: for example, yearly cardiology visits
might bring not only time alone with parents but
also rare treats such as train rides, hotel visits, new
toys, and meals out. For those uninterested in
athletics, being excused from running laps can be
a privilege and make one the envy of one’s fellow
nonathletes. Many adult patients report jealousy
from brothers and sisters because they received
‘‘special’’ treatment growing up. The experience of
facing fear and overcoming medical adversity can
yield life-long resilience and the ability to pursue
long-term goals. It is common for ACHD patients
to report that they were the first in their family to
go to college, some using vocational services
funding offered because of their cardiac disability.
In recent years, disability researchers have
identified a large gap between how those who
are nondisabled perceive the quality of life of
those who have disabling conditions (which is
uniformly negative) and how those who have such
conditions perceive themselves. Moons’ [23] find-
ing of positive health assessment in the CHD com-
munity is echoed in numerous studies that find
526 VERSTAPPEN
that patient perception of personal health status
and abilities in a wide variety of disabling condi-
tions is at odds with objective health status. Mul-
tiple studies have found that most of those living
with significant disabilities report that their qual-
ity of life is good or excellent [24,25]. One study
examining this phenomenon identified the follow-
ing factors as key to perceived high quality of life
with disability: acknowledgment of one’s impair-
ment, preserved control over mind and body, the
ability to perform expected roles, a ‘‘can-do’’ atti-
tude, finding a purpose and meaning in life, hav-
ing a spiritual foundation, social connections,
and feeling satisfied with one’s capabilities in the
context of one’s particular health condition [26].
These same key factors are often referenced by
CHD survivors.
Other ‘‘benefit finding’’ studies report similar
findings. For example, in a qualitative study of 16
stroke survivors, 63% identified positive conse-
quences in the domains of increased social re-
lationships, increased health awareness, change in
religious life, personal growth, and altruism [27].
Moons and colleagues [28] reported that ACHD
patients indicated they placed less emphasis on fi-
nancial means and material well-being than their
heart-healthy peers. Many ACHD survivors
echo the opinions expressed in What’s Your Ex-
piry Date?, the recent book by Canadian ACHD
advocate Patrick Mathieu [29], which describes
how the on-going awareness of one’s mortality
can lead to clarity of purpose, better decision-
making, and a deep appreciation for life.
Implications for health care providers
Because most health care providers see their
patients only within a medical setting, opportuni-
ties to observe the gifts of CHD and the resilience
of CHD survivors may be limited. ACHD pa-
tients often comment that because of the anxiety
and stress raised in the medical setting, their
health care providers tend to see them at ‘‘their
worst.’’ The appropriate goal of medicine is to
identify health problems, cure disease, and relieve
symptoms; therefore, there are few opportunities
to discuss contentment in the face of limitations.
As noted by Moons and colleagues [30], however,
an essential component of assessing long-term
CHD outcomes is creating quality-of-life indica-
tors that accurately assess the patient’s own con-
tentment and experience rather than his or her
ability to function normally. Health care pro-
viders working with the pediatric community are
et al
encouraged to remind parents that health status
and functional capacity are different from quality
of life. Health care providers can help avoid inap-
propriately negative projections by helping par-
ents understand that significant health challenges
in no way preclude a high quality of life and can
build resilience, optimism, and contentment.
For patients reporting a poor quality of life,
potential interventions that address the factors
listed earlier might include increasing patient
understanding of their disease to improve a sense
of mastery, helping to identify core functional
goals and strategies to achieve these goals, or
providing opportunities to meet others facing
similar health challenges. When appropriate, re-
ferral to mental health professionals for assessment
and treatment of potential depression and for
therapeutic intervention should be made. It should
not be assumed that poor quality of life is a neces-
sary corollary of significant cardiac disability.
Issue 9: the importance of patient associations
In the last 15 years, numerous associations for
adult survivors of CHD have been formed
throughout Europe, Australia, Canada, and the
United States: an interactive world map listing
international ACHD groups is available at
www.worldcongenitalheart.org. The psychosocial
benefits of patients interacting with others who
share their health challenges should not be under-
estimated; opportunities for peer support help ad-
dress many of the issues identified in this article.
In addition to offering peer support, ACHD
patient associations can play a seminal role in
addressing the many existing challenges facing
long-term survivors of CHD. Patient organiza-
tions in other diseases, such as the National
Marfan Foundation, the Cystic Fibrosis Founda-
tion, and the Pulmonary Hypertension Associa-
tion, demonstrate the efficacy of such groups in
promoting research, establishing care guidelines,
and providing resources to their communities.
When patients speak directly about their own
experiences and advocate for their own life-long
needs, they have unique power and impact.
Participation in advocacy can be a transformative
experience for the patients and family members
involved: for example, many patient and family
participants in ACHA’s 2005 National ACHD
Lobby Day commented that in educating law-
makers about the unique needs of the ACHD
community, life-long frustrations were translated
into effective action.
 PATIENT’S PERSPECTIVE
Table 1
Adult congenital heart disease patient association challenges and contributions
Challenges facing ACHD patients
Social isolation and lack of support
Public awareness and finding ‘‘the lost’’
Lack of patient education
Access to medical records
Access to appropriate care
Lack of ACHD research
As an example of what patient associations can
undertake, Table 1 lists the key challenges identi-
fied by ACHA and the initiatives addressing these
challenges. A number of these are collaborative
efforts: for example, the National ACHD Clinic
Directory is a joint project between ACHA and
ISACCD, and ACHA’s 2005 National ACHD
Lobby Day involved numerous partners including
the American College of Cardiology, the Congen-
ital Heart Information Network, and the Chil-
dren’s Heart Foundation. Such collaboration is
directly addressed in the 32nd Bethesda Confer-
ence Report, which calls for the establishment of
a national patient advocacy organization and col-
laborations between patients, families, health care
providers, federal agencies, and professional orga-
nizations to promote the provision of long-term
care to the CHD community.
A major triumph of twentieth century medicine
was the success of childhood interventions for
CHD and the creation of the first large cohort of
ACHD survivors. A major challenge for the
twenty-first century is how to address the long-
term surveillance and health care needs created by
the transformation of complex CHD from a fatal
to a life-long condition. As the primary life-long
stakeholders, the patients themselves must con-
tinue to be central to this effort, and their perspec-
tive must continue to inform the development of
a cardiac care system that meets the needs of the
CHD community.
Just as ACHD patients as a group must remain
central in the design and provision of care systems
for CHD survivors, each individual ACHD
527
ACHA program
Online discussion board
Opportunities to meet other ACHD patients
Media campaigns
National ACHD Lobby Day
Patient/family conferences
Patient/family Newsletter
Online information
Print and Electronic Personal Health Passport
ACHD Clinic Directory
Dissemination of existing ACHD care guidelines
ACHA-sponsored Continuing Medical Education programs
Promotion of national ACHD registry
Sponsorship of national ACHD research symposium
patient must be equipped to play a central role in
his or her own life-long cardiac care. Due to the
novelty and rarity of these conditions among
adults, ACHD patients are called on to self-
advocate with medical professionals who have
little knowledge of their disease and to negotiate
adult roles in a community that has little aware-
ness of congenital heart defects. As described
earlier, many ACHA participants report barriers
to care and well-being that might have been
avoided or ameliorated with better and more
honest communication, more accurate informa-
tion, or more insight and assistance from previous
health care providers. For each issue identified in
this article, the authors have outlined implications
for the ACHD health professional and made
recommendations for best practices. Table 1 sum-
marizes specific behaviors to avoid and best prac-
tices to promote. Overall, the authors’ goal is to
help ensure that every ACHD survivor has the
tools she or he needs to maximize his or her life
span, functional status, and quality of life.
Appendix 1
Classification of congenital heart defects
Group 1: congenital heart disease of mild
complexity
Native conditions:
Isolated congenital aortic valve disease
Isolated congenital mitral valve disease
(except parachute valve, cleft leaflet)
528 VERSTAPPEN
Isolated PFO/small atrial septal defect
Isolated small ventricular septal defect
Mild pulmonic stenosis
Repaired conditions:
Previously ligated or occluded patent ductus
arteriosus
Repaired secundum or sinus venosus atrial
septal defect without residua
Repaired ventricular septal defect without
residua
Group 2: congenital heart disease of moderate
complexity
Aorto-left ventricular fistulae
Anomalous pulmonary venous drainage
Atrioventricular canal/septal defects
Ostium primum atrial septal defect
Coarctation of the aorta
Ebstein’s anomaly
Infundibular right ventricular outflow obstruction
Patent ductus arteriosus (not closed)
Pulmonary valve regurgitation (moderate to
severe)
Pulmonary valve stenosis (moderate to severe)
Sinus of Valsalva fistula/aneurysm
Sinus venosus atrial septal defects
Subvalvular or supravalvar aortic stenosis
Tetralogy of Fallot
Ventricular septal defect with
Absent valve or valves
Coarctation of the aorta
Mitral disease
Right ventricular outflow tract obstruction
Straddling tricuspid/mitral valve
Subaortic stenosis
Group 3: congenital heart disease of great
complexity
Conduits, valved, or nonvalved
Cyanotic CHD (all forms)
Double-outlet ventricle
Eisenmenger’s syndrome
Fontan procedure
Mitral atresia
Functionally single ventricle
Pulmonary atresia (all forms)
Pulmonary vascular obstructive disease
Transposition of the great arteries
Congenitally corrected transposition of the
great arteries
et al
Tricuspid atresia
Truncus arteriosus/hemitruncus
Other abnormalities of atrioventricular or ven-
triculoarterial connection not included
above (ie, crisscross heart, isomerism, heter-
otaxy syndromes)
References
[1] Webb GD, Williams R. 32nd Bethesda Conference:
‘‘care of the adult with congenital heart disease.
J Am Coll Cardiol 2001;37(5):1161–98.
[2] Warnes CA. The adult with congenital heart disease:
born to be bad? J Am Coll Cardiol 2005;46(1):1–8.
[3] Deanfield J, Thaulow E, Warnes CA, et al. Manage-
ment of grown up congenital heart disease: the Task
Force on the Management of Grown Up Congenital
Heart Disease of the European Society of Cardiol-
ogy. Eur Heart J 2003;24(11):1035–84.
[4] Connelly MS, Webb GD, Somerville J, et al.
Canadian consensus conference on adult congenital
heart disease 1996. Can J Cardiol 1998;14(3):
395–452.
[5] Wernovsky G, Rome JJ, Tabbutt S, et al. Guide-
lines for outpatient management of complex con-
genital heart disease. Congenit Heart Dis 2006;
1(1–2):10–26.
[6] Kidd L, Driscoll D, Gersony W. Second natural his-
tory study of congenital heart defects: results of
treatment of patients with ventricular septal defects.
Circulation 1993;87(2 Suppl):138–51.
[7] Reid GJ, Irvine MJ, McCrindle BW, et al. Preva-
lence and correlates of successful transfer from pedi-
atric to adult health care among a cohort of young
adults with complex congenital heart defects. Pediat-
rics 2004;113(3 Pt 1):e197–205.
[8] Cantor WJ, Harrison DA, Moussadji JS, et al. De-
terminants of survival and length of survival in
adults with Eisenmenger syndrome. Am J Cardiol
1999;84(6):677–81.
[9] Moons P, De Volder E, Budts W, et al. What do
adult patients with congenital heart disease know
about their disease, treatment, and prevention of
complications? A call for structured patient educa-
tion. Heart 2001;86(1):74–80.
[10] American Medical Association. E-8.12. Patient
information page. Available at: http://www.
ama-assn.org/ama/pub/category/8497.html. Accessed
May 9, 2006.
[11] Erde EL, Nadal EC, Scholl TO. On truth-telling and
the diagnosis of Alzheimer’s disease. J Fam Pract
1988;24(4):401–6.
[12] Veldtman GR, Matley SL, Kendall L, et al. Illness
understanding in children and adolescents with heart
disease. West J Med 2001;174(3):173–4.
 PATIENT’S PERSPECTIVE
[13] Feudtner C. What are the goals of patient education?
Heart 2000;84:395–7.
[14] Landzberg MJ, Murphy DJ Jr, Davidson WR, et al.
Task force 4: organization of delivery systems for
adults with congenital heart disease. J Am Coll Car-
diol 2001;37(5):1161–98.
[15] Somerville J. Near misses and disasters in the treat-
ment of grown-up congenital heart patients. J R
Soc Med 1997;90:124–7.
[16] Horner T, Lieberthson R, Jellinek MS. Psychosocial
profile of adults with complex congenital heart dis-
ease. Mayo Clin Proc 2000;75(1):31–6.
[17] Tong EM, Sparacino PS, Messias DK, et al. Grow-
ing up with congenital heart disease: the dilemmas of
adolescents and young adults. Cardiol Young 1998;
8(3):303–9.
[18] Claessens P, Moons P, de Casterle BD, et al. What
does it mean to live with a congenital heart disease?
A qualitative study on the lived experiences of adult
patients. Eur J Cardiovasc Nurs 2005;4(1):3–10.
[19] McMurray R, Kendall L, Parson JM, et al. A life less
ordinary: growing up and coping with congenital
heart disease. Coron Health Care 2001;5(1):51–7.
[20] Gantt LT. Growing up heartsick: the experiences of
young women with congenital heart disease. Health
Care Women Int 1992;13(3):241–8.
[21] Brandhagen DJ, Feldt RH, Williams DE. Long-
term psychologic implications of congenital heart
disease: a 25-year follow-up. Mayo Clin Proc 1991;
66(5):474–9.
529
[22] Arnett JJ. Emerging adulthood: a theory of develop-
ment form the late teens through the twenties. Am
Psychol 2000;55(5):469–80.
[23] Moons P, Van Deyk K, De Bleser L, et al. Links
quality of life and health status in adults with con-
genital heart disease: a direct comparison with
healthy counterparts. Eur J Cardiovasc Prev Reha-
bil 2006;13(3):407–13.
[24] Albrecht GL, Higgins P. Rehabilitation success: the
interrelationships of multiple criteria. J Health Soc
Behav 1997;18:36–45.
[25] Weinberg N. Another perspective: attitudes of
people with disabilities. In: Attitudes toward per-
sons with disabilities. New York: Springer; 1998.
p. 141–53.
[26] Albrecht GL, Devlieger PJ. The disability paradox:
high quality of life against all odds. Soc Sci Med
1999;48:977–88.
[27] Gillen G. Positive consequences of surviving
a stroke. Am J Occup Ther 2005;59(3):346–50.
[28] Moons P, Van Deyk K, De Geest S, et al. Is the se-
verity of congenital heart disease associated with
the quality of life and perceived health of adult pa-
tients? Heart 2005;91(9):1193–8.
[29] Mathieu P. What’s Your Expiry Date? Ontario,
Canada: Patrick Mathieu Unlimited; 2005.
[30] Moons P, Van Deyk K, Marquet K. Individual
quality of life in adults with congenital heart dis-
ease: a paradigm shift. Eur Heart J 2005;26(3):
298–307.
 Cardiol Clin 24 (2006) 531–556
The Role of Cardiac Catheterization in Adult
Congenital Heart Disease
Peter McLaughlin, MD, FRCP(C)a,*, Lee Benson, MD, FRCP(C)b,
Eric Horlick, MD, FRCP(C)c
a
Peterborough Regional Health Centre, Peterborough, ON, Canada
b
Division of Cardiology, Hospital for Sick Children, Toronto, ON, Canada
c
University Health Network, Toronto, ON, Canada
The purpose of this article is to define the role
of cardiac catheterization in modern adult con-
genital heart disease (ACHD) facilities. The role
of heart catheterization continues to evolve as the
sophistication of cardiac MRI and CT improves
and the breadth of interventional catheter tech-
niques widens dramatically. This task is ap-
proached from four perspectives. The first is the
planning of the procedure, including consider-
ation of the information required, the potential
pitfalls to be anticipated, and the equipment
needed for the procedure. The second perspective
is the performance of the procedure, including the
essential points related to the sample run, coro-
nary arteriography, chamber angiography, and
angiography of selected specific lesions. The third
perspective is the current role of heart catheteri-
zation with a consideration of the impact of echo,
MRI, and CT on indications for catheterization
procedures and a brief look at today’s interven-
tional procedures. Finally, new and emerging in-
terventions are considered and speculation is
given to the future role of diagnostic heart cathe-
terization in patients who have ACHD.
Planning the procedure
The quality and usefulness of the diagnostic
information yielded by a catheterization proce-
dure can be related directly to the quality of the
planning and preparation undertaken before the
* Corresponding author. 1 Hospital Drive, Peterbor-
ough, ON, Canada K9J 7C6.
E-mail address: pmclaugh@prhc.on.ca
(P. McLaughlin).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.008
catheterization study and the knowledge base and
experience of the operators. Operators must have
a thorough understanding of the anatomy and
physiology of congenital cardiac defects, the
potential defects associated with the primary de-
fect, the therapeutic options for the defect under
investigation, and the information surgeons or
interventionalists require if patients are to be
referred for treatment. They need to know
What information is absolutely essential to
establish a diagnosis or plan treatment;
What information is useful to obtain but is not
critical; and
What information is redundant and already
available from other imaging studies.
This information, combined with a discussion
with adult congenital cardiac clinicians who are
leading patient care, provides the questions that
must be answered. When such thought and prepa-
ration has been done before the procedure, cathe-
terization can be performed in the most efficient
manner, minimizing radiation exposure, length of
the procedure, and volume of contrast media.
Preprocedural preparation includes a review of
the echo Doppler information, CT scanning,
cardiac MRI, and functional cardiac testing,
such as perfusion imaging, if available. Operators
then understand the anatomy, what is known, and
what information must be obtained for the
assessment of particular patients.
What information is essential?
Not uncommonly, studies in some complex
cases become unnecessarily long or, perhaps worse,
cardiology.theclinics.com
532 MCLAUGHLIN
are completed without a key piece of information
having been obtained. The most common essential
information concerns the pulmonary vasculature:
the pulmonary artery pressures and resistance, the
reactivity of the pulmonary vasculature, and shunt
calculations. If access to the pulmonary artery is
difficult, this may mean prolongation of the pro-
cedure, but time invested here may be far more
valuable than recapturing other data already clear
from other diagnostic testing. Similarly, the temp-
tation may occur to defer coronary angiography
after a difficult procedure in young adults in the
precoronary age group, missing the opportunity to
detect a relevant congenital anomaly of the coro-
nary circulation.
What catheters to use and the sequence of events
Prepared operators go into a cardiac catheter
laboratory with a clear idea of which catheters are
likely to be most helpful and the sequence they wish
to follow to obtain the information. For example,
it may be useful to begin the right heart catheter-
ization with a steerable catheter, such as a Good-
ale-Lubin catheter to sample for oxygen saturation
and probe for atrial septal defects (ASDs) and
anomalous venous drainage, then change to a bal-
loon-tipped angiocatheter to reach the pulmonary
artery through a difficult right ventricular outflow
tract. Operators are chagrined if they realize they
have reached the pulmonary artery but have
forgotten to obtain all the oximetry samples and
pressures on the way up. They then face the choice
of compromising the saturation run and pressures
with some delayed postangiographic measure-
ments or having to withdraw the catheter, obtain
the measurements, and re-enter the pulmonary
artery, adding to fluoroscopy and procedure time.
To summarize, make a mental checklist at the
beginning of the procedure outlining which he-
modynamic information must be obtained, which
chamber and great vessel angiography must be
done, whether or not the coronary anatomy must
be determined, which catheters are most useful,
and the sequence to be followed throughout the
procedure.
What can go wrong
The common problems with cardiac catheter-
ization studies of adult patients who have con-
genital cardiac disease relate to
Prolonged catheterization time and contrast
used
et al
Inadequate, missing, or nondiagnostic infor-
mation obtained
Redundant information obtained
Catheter complications
Many of these problems can be minimized with
planning and consultation with clinicians and
surgeons involved in patients’ care. Most of these
procedures can be expected to take substantially
more time to complete than the usual right and
left heart procedures in patients who have coro-
nary or valvular heart disease, particularly if an
unexpected finding arises during the procedure.
Keep in mind, however, the key questions as the
catheterization progresses.
1. Have the diagnosis and best form of treat-
ment been established?
2. Do I have the essential information required
to establish the diagnosis, define the anat-
omy, define the physiology, define the pres-
ence or absence of associated anomalies,
and provide the surgeon/interventionalist
with the information required?
3. If not, must this information be obtained at
this procedure, or can noninvasive testing ob-
tain equivalent or better information?
4. How much contrast has been used so far in
this procedure? Has the patient been prehy-
drated adequately?
5. How is the patient tolerating the procedured
if poorly, should the procedure be terminated
with plans for a second procedure to obtain
any further essential data?
6. Are there comorbidities that might add to the
risk of a complication and, if so, what are the
potential preventive measures (eg, erythrocy-
tosis increases the risk of catheter clotting, so
more frequent flushing is required)?
7. Am I collecting the information in the right
sequence (eg, Have I acquired all the pressure
and oximetry samples on the way up to the
pulmonary artery? Have I acquired all the
hemodynamic measurements before proceed-
ing to angiography?)?
Although no procedure can ever be entirely
risk-free, taking these precautions lessens the
chance of problems occurring.
Equipment for interventional cardiac
catheterization
As in adult coronary interventions, a consider-
able amount of equipment is required to address
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
the varied lesions that present in the catheteriza-
tion laboratory. Unlike coronary angioplasty and
stenting, however, the different types of devices,
wires, catheters, embolization, and retrieval
equipment can be vast when addressing congenital
heart lesions. It is challenging especially in pedi-
atric laboratories, where the inventory has to
further address varied patient sizes. If money is
not an issue (at the rare institution), then it is
possible to keep everything in stock. If finances
are limited, then rational stocking of the labora-
tory is needed, as a laboratory cannot have
everything but must have the equipment to cover
most of the procedures. There are three principles
of inventory management: PLANNING, PLAN-
NING, and PLANNING.
The types of equipment needed to cover most
interventional cases include sheaths, guide wires,
catheters, guiding catheters, balloons, coils, oc-
clusion devices, stents, covered stents, and re-
trieval kits.
Sheaths
A selection of sheaths from 4- to 12-Fr is
required. If the technique of coarctation stenting
is offered, a 14-Fr system is needed (for covered
stent implantation). Rarely do operators need an
18-Fr sheath. Short percutaneous sheaths to long
transseptal lengths should be available. There are
various Mullins-type sheaths, and they should be
purchased with radio-opaque tip markers (Fig. 1).
Additionally, some operators find it helpful to
have kink-resistant long sheaths, but this may
not be an issue in adult patients, as the intravascu-
lar curves are gentler.
Guide wires
A selection of guide wires, with sizes from 0.01
in to 0.03 in, in lengths from 50 cm to 260 cm, is
needed (Fig. 2). They should be super-floppy,
Fig. 1. The left panel is a photo of the side-arm bleed-back tap on a Mullins-type long sheath, whereas the right panel
shows the radio-opaque marker tip, which is most useful when initially positioning the sheath and when directing a bal-
loon stent toward the target lesion.
533
ordinary, super-stiff, and glide wire (eg, Terumo)
varieties. The so-called ‘‘Amplatzer stiff’’ and ‘‘ul-
tra-stiff’’ guide wires (0.025 in to 0.038 in) are
found invaluable for stabilizing balloons across
high flow lesions and during stent implantation.
Catheters
A variety of catheters is required. Basic con-
figurations, such as the multipurpose (Goodale-
Lubin), pigtails, NIH, cobra, Judkins coronary,
and Multitrack (NuMed, Hopkinton, New York)
should be available. Infusion catheters (such as
the Tracker-18) should be available. Tapered
and nontapered catheters, guiding catheters,
and balloon wedge (end-hole) and angiographic
(side-hole) catheters, such as the Berman cathe-
ters (Critikon), are the foundation of any inter-
ventional laboratory (Fig. 3).
A comprehensive stock of catheters is not
needed. Operators should choose an appropriate
selection and become familiar with their use. The
more complex the case mix, the greater the variety
of catheters needed.
Miscellaneous equipment
Transseptal needles, using the Mullins tech-
nique, occasionally are required to enter the left
heart or perforate an atretic vascular structure. For
adults, a single length of transseptal needle can be
stocked, but the needle must fit the long sheath
dilator. Additionally, the hub of the dilator should
be such that when the needle and hub are engaged,
only 2 or 3 cm of the needle are exposed from the tip.
Balloons
Although adult interventional cardiologists are
experienced with a variety of coronary balloons
for primary angioplasty and as a platform for
stent delivery, the form and function of the bal-
loons used in interventions that focus on
534 MCLAUGHLIN
Fig. 2. Wires of various curves.
a congenital heart patient population are different
regarding size, length, design, and material. Fur-
thermore, many of the balloons used initially
were produced not for intracardiac or pulmonary
applications but for peripheral angioplasty.
Low-pressure balloons (eg, Tyshak II, Z-Med
[II], NuMed) are available in a range of sizes,
many on 4- to 6-Fr shafts. They are advantageous
especially because of their rapid deflation rates,
which limit the time an outflow tract is occluded
during a procedure. High-pressure balloons also
come in a range of sizes and lengths from several
manufacturers (NuMed-Mullins high pressure
balloons, Cordis Johnson & Johnson). Most bal-
loons can be used as platforms for stent delivery.
The NuMed BIB (balloon in a balloon) is popular
Fig. 3. The variety of catheter curves is endless. The au-
thors find the most useful to be the Judkins right coro-
nary, multipurpose, and cobra shapes. Operators must
determine their own preferences for each type of vascu-
lar structure that must be traversed.
et al
for controlled expansion and is useful especially
for stent delivery. Despite the enormous selection
that is available, a large number of different
makes of balloons is not needed (Figs. 4 and 5).
Embolization equipment
In patients who have congenital heart disor-
ders, standard embolization devices, such as coils,
have, in many instances, been supplanted by the
application of device implants designed for other
locations and indications, such as ductal, atrial,
and patent foramen ovale (PFO) defect occluders.
Coils
For several years, the Gianturco free release
coil (Cook) has been the primary device for pe-
ripheral embolization, and it remains the single
most used implant in congenital lesions (patent
arterial duct, aortopulmonary collaterals, and ac-
quired venovenous collaterals). Controlled release
coils (where release is dependent on an active ma-
neuver from the operator), however, offer a safer
implant, especially in higher flow lesions or areas
where precise coil implantation is critical. A large
variety of sizes, lengths, and shapes is available,
from various suppliers. Additionally, a selection
of guiding and delivery catheters (eg, Tracker-
18; Cook) to allow coil embolization of very tortu-
ous vessels should be available (Fig. 6).
Controlled release coils offer an additional
advantage for they can be retrieved and reposi-
tioned before release. Some coil formats use an
electrolytic detachment, whereas others a mechan-
ical release mechanism. Such coils are atraumatic
and result in no damage to the vessel. They offer
low radial friction within the delivery catheter
lumen for easy placement. For an effective occlu-
sion, a dense mass of coils is needed. As such, they
take a longer time to form a thrombus than
feathered steel coils. The detachable controlled re-
lease coils are made of platinum 0.018-in and
0.011-in wire. They are introduced through
a Tracker-18 catheter with a simple introducer
system. Several coil shapes are available.
Atrial and ventricular septal devices
There are several devices clinically available for
closure of secundum atrial and ventricular septal
defects (primary muscular) (Fig. 7). The Amplat-
zer (AGA Medical, Golden Valley, Minnesota),
CardioSEAL or Starflex (NMT Medical), and He-
lex (Gore Medical) are a few of the devices ap-
proved for clinical use. What is kept in inventory
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
Fig. 4. Various balloons used for stent delivery and vessel/valve angioplasty.
depends on operator experience, and the range and
number of defects that are to be addressed. The
Amplatzer Septal Occluder has the advantage of
being applicable to a wide range of defect diame-
ters but, therefore, requires a large stock (sizes 4
to 20 mm in 1-mm increments; sizes 22 to 40 in
2-mm increments). The most commonly used im-
plant for muscular ventricular septal defects is
that designed by AGA Medical, the Amplatzer
Septal Occluder. It is easy to deliver to the target
lesion, retrievable, and comes in several sizes
(4- through 18-mm diameter central plugs). These
devices also can be used for other occlusions (eg,
atrial implants can be used in ventricular defects,
in persistently patent arterial ducts, or in fistulas).
Fig. 5. The BIB balloon has an inner balloon, con-
structed from the same material as the so-called ‘‘Tyshak
II’’ balloon, whereas the outer balloon is constructed
from a heavier gauge material as used in the Z-Med
higher pressure balloon. The BIB balloon allows con-
trolled delivery of stents and adjustment of stent posi-
tion after inflation of the inner balloon. This prevents
stent migration and the balloon design prevents flaring
of the stent.
535
Endovascular stents
There are several endovascular stents that are
useful in interventional management of patients
who have a congenital heart lesion. Operators
should be familiar with one or two types, noting
their advantages and limitations. In the adult
setting, stocking a range of sizes and lengths can
be rationalized, particularly for use in aortic co-
arctation. In this case, it seems that a covered stent
(see later discussion) is the safest form of implant.
There are several choices, including but not limited
to the Palmaz Genesis or XL (Johnson & Johnson,
Warren, New Jersey), Cheatham-Platinum (CP)
(NuMed), Jomed (Jomed NV), Wallstent (Boston
Scientific), Corinthian (Johnson & Johnson).
There is a limited number of covered stents appli-
cable to adults who have congenital heart lesions,
primarily because of available expansion ratios
(eg, Jomed, Gore, Medtronic, Wallstent, and CP
stents).
In addition to their use in aortic coarctation,
they have an important role as standby or bailout.
The CP stent is the covered stent used most
commonly for congenital lesions. It is made of
platinum and iridium, with gold used to
strengthen the 0.013-inch–thick wire solders. Un-
like other implants, they have rounded leading
and trailing edges that reduce the risk of balloon
rupture during inflation and of vessel trauma. Its
visibility is good; it can be expanded to large
diameters (up to 25 mm); and delivery can be
through 12-Fr sheaths. Importantly, the implant
is MRI compatible, a significant issue when
managing adults (Fig. 8).
In summary, careful thought in stocking a lab-
oratory with a wide variety of equipment for all
types of interventional procedures is required for
536 MCLAUGHLIN
Fig. 6. The top three panels from the left depict controlled release implants; the remaining panels show a few of the
many shapes that can be useful in particular locations.
an effective program. Although considerable var-
iations in kinds of equipment are required, a ra-
tional inventory with a focus on adult applications
easily is achievable and at a reasonable expense.
Flows and shunts
The sample run
In the decision-making process for patients
who have a congenital heart lesion, a great deal of
significance is placed on oxygen saturation data.
As an isolated measurement, the determination of
blood oxygen saturation can provide important
information about patients early in the catheter-
ization procedure. Arterial desaturation may reflect
a right-to-left shunt or a ventilation-perfusion
mismatch. Systemic venous saturation less than
Fig. 7. A few of the available atrial defect implants are shown. Top panels: the CardioSEAL (left) and Starflex (right). In
the lower panel, the Amplatzer Septal Occluder (left), Cardia (middle), and Helex (right).
et al
50% indicates a low cardiac output; a high
saturation indicates either a left-to-right shunt
or high output state. The calculation of cardiac
outputs, shunts, and resistances is dependent on
an accurate determination of oxygen saturation.
There are several assumptions in using oxygen
as an indicator (see later discussion), however,
and some practical considerations in obtaining
oxygen saturation data, and potential errors can
be introduced. As such, oxygen saturation data
is the least sensitive and most prone to error of
all the physiologic data obtained in a catheter
laboratory.
Assumptions when using oxygen as an indica-
tor are:
All measurements are made during steady-
state blood flow. In other words, there are
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
Fig. 8. Various stents are available. As an example, the Genesis balloon expandable stent is seen in the left upper panel,
and a self-expanding Wallstent in the left lower panel. The bare metal and covered CP stent are shown in the right upper
and lower panels, respectively.
no changes in blood flow, respiratory rate,
heart rate, or level of consciousness.
Two or more samples are obtained from at
least three sites in rapid succession, which,
in patients who have complex congenital
heart disorders, can be difficult to achieve.
For flow determinations (not discussed in
this article), the assumption is that the sam-
ples are taken at the same time that oxygen
consumption is measured.
The mixed venous saturation
There is no single uniform source for mixed
venous blood, as the ‘‘mixed’’ venous blood
sample has three variable sources, that is, the
inferior vena cava (IVC), the coronary sinus, and
the superior vena cava (SVC), with each caval vein
having multiple sources of blood with different
saturations.
The SVC oxygen saturation may vary by 10%,
as it receives blood from the jugular, subclavian,
and azygous systems, each with different satura-
tions and flows (the subclavian and azygous veins
have higher saturations than the jugular vein).
The IVC also has variable oxygen saturations, as
the components that make up its flow may vary by
up to 10% to 20%. For example, more saturated
blood originates from the renal veins, whereas less
saturated blood comes from gastrocolic and
hepatic sources. The net mixed sample from the
IVC generally is 5% to 10% greater then the SVC.
Coronary sinus blood also contributes to the total
pool of mixed systemic return. Despite making up
only 5% to 7% of total venous return, the very
low saturation in this sample (25%–45%) can
have an impact on the total mixed saturation.
537
As there are multiple contributions to the so-
called ‘‘mixed’’ venous sample, there is no prac-
tical way to measure each and account for the
variations in flow. Not even a sample from the
right atrium can adjust for streaming completely.
In the absence of a shunt lesion, a sample
downstream from the right atrium, such as from
the main pulmonary artery, can provide for
a thoroughly ‘‘mixed’’ sample. Also, it has been
noted that the SVC blood saturation is close to
that in the main pulmonary artery and can be
used as representative of the mixed venous sample
unless patients have a low cardiac output state.
Some investigators use a weighted average of SVC
and IVC blood (see later discussion) as a calcu-
lated mixed venous sample. In the presence of
a downstream shunt lesion, several samples,
obtained in rapid sequence and found to be near
or equal in value, should be used.
Similarly, the mixed pulmonary venous satu-
ration is a combination of all the pulmonary
veins, each reflecting different ventilation-to-per-
fusion ratios. As such, a pulmonary vein sample
may be 50% to 100% disparate from the ‘‘true’’
mixed venous sample. In the absence of a right-to-
left shunt, a downstream sample is preferable (left
ventricle or aorta) to using a single pulmonary
vein saturation. In the presence of a right-to-left
shunt, the assumptions are even more difficult. If
there is a distal right-to-left shunt, then the most
distal left atrial sample, at the orifice of the mitral
valve, should be used.
Sampling
When blood is drawn for shunt calculations,
the samples should be obtained proximally and
distally to the lesion. Note must be taken of the
538 MCLAUGHLIN et al

influence of streaming, where a saturation gradi- The pulmonary-to-systemic-blood-flow ratio

ent may exist. Samples must be drawn in rapid
This calculation is based on the Fick principle;
temporal sequence, no more than 1 to 2 minutes
that is, factors, such as oxygen-carrying capacity
for the sampling run. Duplicate samples should be
and oxygen consumption, that are used for in-
obtained when possible and should differ by no
dividual calculations (for pulmonary and systemic
more then 1% or 2%. Operators must be aware of
flows) cancel out when only the ratio of the two
potential equipment malfunctions as a source of
flows is estimated. This is convenient as it removes
differing sample values. When a sample is drawn,
the more difficult and time-consuming parts of the
all flush solution and blood must be cleared from
calculation. The resulting equation (after remov-
the catheter and the catheter filled with the sample
ing the factors that cancel out) is pleasingly
blood by a further withdraw. If there is a poor
simple: Qp:Qs ¼ (Sat AoSat MV)/(Sat PVSat
connection between the catheter and syringe or
PA), where Sat Ao is aortic saturation, Sat MV is
significant negative pressure is applied, then micro-
mixed venous saturation, Sat PV is pulmonary
bubbles can be drawn into the sample, resulting in
vein saturation, and Sat PA is pulmonary artery
oxygenation. Samples should not be drawn from
saturation.
the side arm of a bleed-back tap, as they contain
As the aortic saturation and pulmonary artery
a chamber where the sample can be contaminated.
saturation are measured routinely, the only com-
This also applies to stopcock valves, where a small
ponents that may present any problem are the
amount of blood remains in the connecting cham-
pulmonary vein saturation and the mixed venous
ber and contaminates the sample.
saturation (see previous discussion). If a pulmo-
nary vein has not been entered, an assumed value
Clinical applications of 98% may be used (note the potential error).
The left atrial saturation can be substituted pro-
In patients who have congenital heart disease
vided there is no right-to-left shunt at atrial level.
in whom a communication between the two sides
Similarly, left ventricular or aortic saturation may
of the heart, or between the aorta and the
be substituted, provided there is no right-to-left
pulmonary artery, allows a shunt to exist, several
shunt. For mixed venous saturation, the tradition
calculations may be made, namely: (1) the mag-
is to use the most distal right heart location if
nitude of a left-to-right shunt, (2) the magnitude
there is no left-to-right shunt. Thus, the pulmo-
of a right-to-left shunt, (3) effective pulmonary
nary artery sample should be used if there is no
blood flow (PBF), and (4) pulmonary-to-systemic-
shunt at atrial or ventricular level.
blood-flow ratio (Qp:Qs).
In practice, the SVC saturation often is used,
Of these, the only calculation that is of
although some prefer to use a value intermediate
practical value is Qp:Qs. This provides a simple
between the SVC and IVC (see previous discus-
and reliable estimate of the extent to which PBF is
sion). It has been demonstrated, however, that the
increased or reduced and provides a useful insight
mixed venous saturation approximates the SVC
into the severity of the hemodynamic disturbance
more closely than the IVC. The following formula
in most cases. It also is simple to perform, using
often is used: mixed venous saturation (MV) ¼
solely the oxygen saturation data from systemic
(3 times the Sat SVC þ the Sat IVC)/4. A simple way
arterial blood, left atrial/pulmonary venous
of calculating this (‘‘in your head’’) is to use the for-
blood, pulmonary artery, and vena caval/right
mula, MV ¼ Sat SVC(Sat SVCSat IVC)/4.
heart samples.
Thus, if SVC saturation is 78% and IVC saturation
The samples need to be acquired in (or be
is 70%, MV should be 76% (7870 ¼ 8; 8/4 ¼ 2;
ventilated with) room air or a gas mixture con-
782 ¼ 76).
taining no more than a maximum of 30% oxygen.
If oxygen-enriched gas (O30% oxygen) is being
The usefulness of the shunt ratio in practice
given, the saturation data may not provide accu-
rate information regarding PBF, as a significant The Qp:Qs is useful, for instance, in making
amount of oxygen may be present in dissolved decisions about surgery for patients who have
form in the pulmonary venous sample (which is not a ventricular defect. Beyond infancy, a Qp:Qs
factored into the calculation if saturations alone greater than 1.8:1 likely requires intervention,
are used). Under such circumstances, pulmonary whereas less than 1.5:1 does not. Qp:Qs also is
flow tends to be overestimated and the Qp:Qs helpful in assessing the hemodynamics of many
exaggerated correspondingly. more complex defects, but it should be recognized
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
that under some circumstances it is of limited
practical help. In atrial defects, if there is evidence
of a significant shunt on clinical and noninvasive
testing(rightventriculardilatation,paradoxicseptal
motion, cardiomegaly on radiograph, or right
ventricular hypertrophy on ECG), the shunt ratio
should not be used to decide about treatment.
This is because atrial shunts depend on right
ventricular filling characteristics, which can vary
depending on conditions (sympathetic tone or cat-
echolamine concentrations). It is not uncommon
for a measured shunt to be small (for example,
!1.5:1) despite other evidence of a significant defect.
Coronary arteriography
This section does not attempt to discuss all
anomalies of the coronary circulation in any
detail, as excellent reviews may be found in the
literature [1,2]. Instead it focuses on the more
common anomalies faced by adult congenital an-
giographers and some possible approaches to con-
sider. First, it is useful to consider a working
classification of the type of coronary anomalies
seen in structurally normal and abnormal hearts.
Freedom and Culham [3] present an excellent re-
view of these anomalies and divide the possibilities
into four groups as outlined in Box 1.
For adult angiographers, one of the more
common and often frustrating presentations is
‘‘the missing coronary artery.’’ A frequent mistake
is to assume that the origin of the artery is indeed
in its expected position in or just above the
midpoint of its facing sinus. An operator then
may persist for inordinate lengths of time with the
usual Judkins-shape catheter, believing the vessel
is there and that further catheter manipulation
will identify its origin. If the usual shape catheter
does not identify the origin quickly, consider the
alternatives. Coronary arteries may connect to the
aorta immediately adjacent to a commissure, to
the ascending aorta well above the sinotubular
junction, or to the contralateral facing sinus. In
addition, the coronary circulation may have
a single main coronary artery, with the right
coronary, circumflex, and left anterior descending
arteries all arising from the trunk, with the main
trunk itself having an anomalous aortic wall or
sinus origin. Similarly, it is not uncommon to find
an individual artery arising from another coro-
nary artery, for example the circumflex or left
anterior descending artery from the right coro-
nary, or the right coronary from the left coronary
artery. If the left coronary artery cannot be found
539
Box 1. Classification of coronary artery
anomalies
1. Anomalies of origin
A. Ostial anomalies
B. Ectopic origin
i. Anomalous origin from the
aortic wall or sinus
ii. Anomalous origin from
a coronary artery
iii. Abnormal connection to
a pulmonary artery
iv. Origin from a vessel other than
the pulmonary artery or aorta
v. Origin from a ventricular cavity
2. Anomalies of course
A. Intramural course
B. Aberrant course of proximal
coronary artery
C. Myocardial bridge
D. Epicardial crossing
3. Anomalies of termination or
connection
A. Connections to cardiac structures
B. Connections to extracardiac
structures
4. Anomalies of coronary size
with the left Judkins catheter in the left coronary
sinus, or if the circumflex or left anterior descending
is ‘‘missing,’’ the next step is to proceed to the right
coronary artery, which usually identifies the miss-
ing artery arising from the right coronary trunk. If
the right coronary artery is the missing artery, then
an aortogram or review of the left ventricular
angiogram often identifies the anomalous origin.
The operator then must select from the variety of
catheter shapes available the best fit for the location
in the aortic wall of the origin. Although there are
no hard and fast rules, often the Amplatzer shapes
and multipurpose catheter are first choices to reach
anomalous origins.
When a proximal coronary artery, in particular
the circumflex or left anterior descending, has an
aberrant course from the anterior facing sinus, it
is important to define which of four courses the
vessel pursues to reach the left ventricle: retro-
aortic, interarterial, right ventricular free wall, or
infundibular septum. Some criteria are available
and, combined with careful angiography, help
540 MCLAUGHLIN et al

make the correct diagnosis [4,5]. This becomes im- Table 1

portant if a cardiac surgical procedure is planned. Angiographic projections
The other not infrequent finding that arises for Projection Angles
adult congenital angiographers is one or more Single plane projections
coronary arteriovenous fistulae. These may con-
nect from either coronary artery, be quite small or Conventional RAO 40 RAO
Frontal 0
very large, be single or multiple, and connect to
Shallow LAO 1 to 30
a chamber, usually right-sided, or to a coronary Straight LAO 31 to 60
vein or coronary sinus. Most are small, exit in Steep LAO 61 to 89
a mediastinal vessel, do not require any interven- Left lateral 90 left
tion, and are of passing interest. A few are large Cranially tilted RAO 30 RAO þ 30 cranial
and associated with symptoms or signs of volume Cranially tilted 30 or 45 cranial
overload and lead to the question of catheter or frontal (sitting-up view)
surgical intervention. In these few cases, angiog- Cranially tilted 25 LAO þ 30 cranial
raphers should spend the time and make addi- shallow LAO
tional contrast injections in multiple projections Cranially tilted 60 LAO þ 20 –30 cranial
mid LAO
to define the exact origin of the fistula carefully
(long axis oblique)
and the anatomy of the exit of the fistula. These Cranially tilted 45 to 70 LAO þ 30
are important in deciding if catheter occlusion is steep LAO cranial
possible, if surgery is necessary, and the best (hepatoclavicular view)
interventional or surgical approach to closure. Caudally tilted frontal 45 caudal
Biplane
combinations A plane B plane
Chamber angiography 
AP and LAT 0 Left lateral
Accurate anatomic and physiologic diagnosis Long axial 30 RAO 60 LAO þ 20 to
is the foundation of a successful catheter-based oblique 30 cranial
therapeutic procedure. This section includes a dis- Hepatoclavicular 45 LAO þ 30 120 LAO þ 15
cussion of standard angiographic approaches and view cranial cranial
Specific lesions
how to achieve them. Emphasis is placed on the
RVOT-MPA 10 LAO þ 40 Left lateral
application of these projections as applied to
(sitting up) cranial
interventional procedures. A detailed description Long axial 30 RAO 60 LAO þ 30
of the physical principles of image formation is for LPA cranial
beyond the scope of this article and interested (biplane)
readers are referred to other sources for more LPA long 60 LAO þ 20
detailed information [6]. axis (single cranial
plane)
Angiographic projections ASD 30 LAO þ 30
cranial
In the therapeutic management of patients who PA bifurcation 30 caudal þ 10 20 caudal
have a congenital heart lesion, the spatial orien- and branches RAO
tation and detailed morphology of the heart and Primary projections are in italics.
great vessels are of critical importance (Table 1). Abbreviations: LPA, left pulmonary artery; MPA,
As an operator enters a laboratory, an under- main pulmonary artery; RVOT, right ventricular out-
standing of the anatomy should have been synthe- flow tract; PA, pulmonary artery.
sized, based on information from other imaging
modalities, such as chest roentgenography, echo-
cardiography, CT, and MRI. As such, the angio-
graphic projections used in the procedure are 3-D structure that takes an S-curve from apex
tailored to outline the lesion to allow appropriate to base, the so-called ‘‘sigmoid septum.’’ From
measurements and guide the intervention. caudal to cranial, the interventricular septum
The heart is oriented obliquely, with the left curves through an arc of 100 to 120 , and the
ventricular apex leftward, anterior, and inferior, right ventricle appears as an appliqué or overlay
in relation to the base of the heart. The in- on the left ventricle. To address this topology,
terventricular septum is a complex geometric today’s angiographic equipment allows a wide
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
range of projections, incorporating caudocranial
or craniocaudal angulations. The up-to-date lab-
oratory consists of independent biplane imaging
chains, which, with the proper selection of views,
minimizes overlapping and foreshortening of
structures.
Terminology
Angiographic projections are designated either
according to the position of the recording detector
(image intensifier or flat panel detector) or the
direction of the x-ray beam toward the recording
device. In cardiology, the convention usually is
the former. For example, when the detector is
directly above a supine patient, the x-ray beam
travels from posterior to anterior and the angio-
graphic projection is designated posteroanterior
(PA), but based on detector position, it is called
frontal, and the position of the detector by
convention is at 0 . Similarly, when the detector
is moved through 90 , to a position beside and to
the left of a patient, a lateral (LAT) projection
results. Between 0 and 90 , there are a multitude
of projections, termed left anterior oblique (LAO),
and when the detector is moved to the right of
a patient, a right anterior oblique (RAO) pro-
jection is achieved. Standard detectors mounted
on a C-arm or parallelogram not only allow these
positions but also the detectors can be rotated
around the transverse axis, toward the feet or
head, caudally or cranially (Fig. 9).
Biplane angiography
A dedicated interventional catheterization lab-
oratory addressing congenital heart defects re-
quires biplane facilities [7,8]. Biplane angiography
has the advantage of limiting contrast exposure
and evaluating the cardiac structures in real-time
in two projections simultaneously. This is at
a cost, however, as these facilities are expensive,
Fig. 9. Angiographic projections in biplane. L, left; R, right.
541
and with existing image intensifiers and newer
flat panel detectors, extreme simultaneous angula-
tions can be compromised. Standard biplane con-
figurations include RAO/LAO and frontal/lateral
projections, with additional cranial or caudal tilt.
The possible combinations are endless (see Box 1;
see Fig. 1).
Cranial–left anterior oblique projections
A clear working understanding of these pro-
jections is of critical importance in developing a
flexible approach to congenital heart defect angi-
ography and intervention. The practice of using
‘‘cookbook’’ projections for each case may allow
acceptable diagnostic studies but falls short of the
detail required to accomplish an interventional
procedure. A comprehensive understanding of
normal cardiac anatomy, especially the interven-
tricular septum, allows operators to adjust the
projection to profile the region of interest
optimally.
There are several rules of thumb that allow
operators to judge the steepness or shallowness of
an LAO projection. Of importance is the relation-
ship of the cardiac silhouette to the spine, the
ventricular catheter, and the ventricular apex.
To optimize the profile of the midpoint of the
membranous ventricular septum (thus, the major-
ity of perimembranous defects), two thirds of the
cardiac silhouette should be to the right of the
vertebral bodies (Figs. 10 and 11). This results in
a cranially tilted left ventriculogram showing the
left ventricular septal wall, with the apex (denoted
by the ventricular catheter) pointing toward the
bottom of the image. A shallower projection has
more of the cardiac silhouette over toward the
left of the spine and profiles the inferobasal com-
ponent of the septum, ideal for inlet type ventric-
ular defects. This projection allows for evaluation
of atrioventricular valve relationships, inlet exten-
sion of perimembranous defects, and posterior
542 MCLAUGHLIN et al
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE 543

muscular defects. A steeper LAO projection can
be used to profile the outlet extension of a peri-
membranous defect and anterior muscular and
apical defects. As noted in Fig. 10, the ventricular
catheter in the cardiac apex can be used to help
guide the projection, but only if it enters the
chamber through the mitral valve. If catheter en-
try is through the ventricular defect or retrograde,
it tends to be more basal and left lateral.
Modification of the cranial LAO projection
has to be made if there is a discrepancy in
chamber sizes, and the septum is rotated such
that a steeper or shallower projection may be
required. Also, it is assumed that a patient is lying
flat on the examining table, but if the head is
turned to the right or there is a pad under the
buttocks, it rotates the thorax such that the LAO
projection is steeper and the detector caudal. This
has to be compensated for during the set-up for
the angiogram. The clue in the former case is that
the more of the heart silhouette is over the spine.
The first step in setting up a cranial-LAO
projection is to achieve the correct degree of
steepness or shallowness. After that, the degree
of cranial tilt has to be confirmed, so that the
basal-apical septum is elongated. This can be
estimated by seeing how much of the hemi-
diaphragm is superimposed over the cardiac
silhouette; the greater the superimposition, the
greater the cranial tilt. Additionally, the degree
of cranial tilt can be determined by looking at
the course of the ventricular catheter, which
appears to be foreshortened or coming directly
at the viewer as the degree of cranial angulation
is decreased (Fig. 12).
Specific lesions
Ventricular septal defect
The imaging of specific ventricular defects
(Fig. 13) is beyond the scope of this review but
is commented on in detail by various investigators
[9]. The injections to outline the septum and the
margins that circumscribe the defects are per-
formed best in the left ventricle using a power in-
jector. Two orthogonal (right-angle) projections
give the best chance of profiling the lesion. Table
1 lists single and biplane angulations for the vari-
ous projections. For the perimembranous defect,
the midcranial LAO projection, at approximately
50 to 60 LAO, and as much cranial tilt as the
equipment and patient position allow (see
Fig. 10) should be attempted. Additional projec-
tions can include a shallow LAO with cranial tilt
(so-called ‘‘four-chamber’’ or hepatoclavicular
view) to outline the basal septum or inlet exten-
sion of a perimembranous defect. The RAO
view outlines the high anterior and infundibular
(outlet) defects [10].
Coarctation of the aorta
Biplane angiography should be used to outline
the aortic arch lesion (Fig. 14). Projections that
can be used include LAO/RAO, frontal and
LAT, or a shallow or steep LAO. The authors’
preference is a 30 LAO and left LAT, with 10
to 15 caudal tilt to minimize any overlapping
structures, such as a ductal bump or diverticulum.
Modifications to accommodate a right arch gener-
ally are mirror-image projections (ie, 30 RAO
and left LAT). Operators must be cautious to ex-
amine the transverse arch for associated hypopla-
sia, and this may be foreshortened in the straight
left-LAT projection. In such an instance, for a left
arch, a left posterior oblique projection may elon-
gate the arch. This is important particularly if an
endovascular stent is deployed near the head and
neck vessels.
Aortic valve angiography
In the setting of normally related great arteries
with ventriculoarterial concordance, assessment
of the diameter of the aortic valve for balloon
dilation is performed best using biplane in the long

=
Fig. 10. Setting up a standard LAO projection. To achieve the LAO projection, attempt to adjust the detector angle so
that two thirds of the cardiac silhouette is to the left of the spine as in (E). If a catheter is through the mitral valve in
the left ventricular apex, it points to the floor, as in (F). In this view, the intraventricular septal margin points toward the
floor. The so-called ‘‘4-chamber’’ or hepatoclavicular view is achieved by having half the cardiac silhouette over
the spine, as in (C). A catheter across the mitral valve appears as in (D). A steep LAO projection has the cardiac silhou-
ette as in (G), and a transmitral catheter in the left ventricle appears as in (H). (A) and (B) show the frontal projection.
(Modified from Culham JAG. Physical principles of image formation and projections in angiocardiography. In: Freedom
RM, Mawson JB, Yoo SJ, et al, editors. Congenital heart disease textbook of angiocardiography. Armonk, NY: Futura
Publishing; 1997. p. 9–93; with permission.)
544 MCLAUGHLIN et al

Fig. 11. Achieving an LAO projection. (A) For a hepatoclavicular view, half of the cardiac silhouette is over or just left
of the spine, with the catheter pointing toward the left of the image. During the injection, the apex and catheter (arrow)
point toward the bottom and left of the image. In this example, the basal (inlet) portion of the septum in intact. In (B)
multiple midmuscular septal defects are not well profiled (arrowheads). In (C), the LAO projection is achieved with the
catheter pointing toward the bottom (arrow) of the frame and the cardiac silhouette well over the spine. During the con-
trast injection (D), the midmuscular defects (arrow) are profiled better. (Modified from Culham JAG. Physical principles
of image formation and projections in angiocardiography. In: Freedom RM, Mawson JB, Yoo SJ, et al, editors. Con-
genital heart disease textbook of angiocardiography. Armonk, NY: Futura Publishing; 1997. p. 39–93; with permission.)

axis and RAO projections (Fig. 15; see Table 1).
The authors’ preference is to obtain the diameter
of the aortic valve from a ventriculogram, which
profiles the hinge points of the leaflets. Caution
must be observed when using an ascending aorto-
gram, as one of the leaflets of the valve may ob-
scure the margins of attachment.
The Mustard baffle
Over time, patients who have had a Mustard
operation may develop obstruction to one or both
limbs of the venous baffle (Fig. 16). As atrial
arrhythmias are not uncommon in such adult pa-
tients, pacing systems frequently are required for
management. To facilitate pacing catheter inser-
tion, enlargement of a stenotic, often asymptom-
atic, superior baffle frequently is required. The
optimum projection to outline superior baffle ob-
struction for potential stent implantation is a cra-
nially angulated LAO projection (30 LAO and
30 cranial). This view elongates the baffle path-
way, allowing accurate measurement before stent-
ing. For inferior baffle lesions, a frontal projection
allows adequate localization of the lesion. Leaks
along the baffle are more problematic and require
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE 545

Fig. 12. Obtaining the cranial tilt. In the standard RAO view (A), the left ventricular apex points caudally and to the left.
The LAO view opens the outflow from apex to base, as in diagram (C). If there is an upturned apex as in Fallot’s te-
tralogy the RAO view appears as in (B). Adding cranial tilt to a mid-LAO projection does not open the apex to base
projection effectively, and the appearance is as looking down the barrel of the ventricles, as in (D). (Modified from Cul-
ham JAG. Physical principles of image formation and projections in angiocardiography. In: Freedom RM, Mawson JB,
Yoo SJ, et al, editors. Congenital heart disease textbook of angiocardiography. Armonk, NY: Futura Publishing; 1997.
p. 39–93; with permission.)

modification of the projection. The initial ap-
proach should be a frontal projection, with mod-
ifications in angulation made thereafter to best
profile the lesion for device implantation.
The secundum atrial septal defect
and the fenestrated Fontan
Secundum ASDs are profiled best in the 30
LAO with 30 cranial tilt (Figs. 17 and 18). With
the injection made in the right upper pulmonary
vein, the sinus venosus portion of the septum
can be visualized, and anomalous pulmonary
venous return ruled out. Additionally, any associ-
ated septal aneurysm can be outlined. With the
application of transesophageal echocardiography
(TEE) or intracardiac echocardiography, there is
less reliance on fluoroscopic device positioning.
When balloon sizing is performed, this projection
elongates the axis of the balloon for proper
measurements.
The interventional management of patients
who have a fenestrated Fontan, either a lateral
tunnel or extracardiac connection, generally re-
quires selective studies of the SVC and IVC and
pulmonary circulations to determine the presence
or absence of obstructive or hypoplastic pathways
and whether or not venous collaterals have de-
veloped. If present, they must be addressed by
angioplasty, stenting, or embolization techniques
before fenestration closure. Venous collaterals
after an extracardiac Fontan generally develop
either from the innominate vein or from the right
upper hepatic/phrenic vein, toward the neoleft
atrium, less frequently from the right hepatic veins
to the pulmonary veins. The optimum projection
to outline these lesions is in the frontal and LAT
projections, with selective power injections in the
546 MCLAUGHLIN et al

Fig. 13. (Left) Long axis oblique projection of a left ventriculogram, defining a perimembranous ventricular septal
defect. (Right) A midmuscular defect outlined with a hepatoclavicular left ventricular injection.

appropriate vessel. The location and dimensions
of the fenestration also may be defined in these
views, but for ideal profiling, some degree of right
or left anterior obliquity may be required.
The bidirectional cavopulmonary connection
Second-stage palliation for several congenital
defects consists of a bidirectional cavopulmonary
connection (also known as the bidirectional Glenn
anastomosis). Because the caval to pulmonary
artery connection is toward the anterior surface of
the right pulmonary artery (rather than on the
upper surface), an anteroposterior (AP) projection
results in overlapping of the anastomotic site with
the pulmonary artery. Therefore, to determine
whether or not the anastomosis is obstructed,
a 30 caudal with 10 LAO projection generally
opens that region for better definition (Fig. 19).
Furthermore, this projection outlines the full ex-
tent of the right and left pulmonary arteries. The
left-LAT projection, with or without 10 caudal
angulation, profiles the anastomosis for its ante-
rior-posterior dimension. Contrast injection must
be made in the lower portion of the SVC. Exami-
nation of venous collaterals can be performed
from the AP and LAT projections in the innomi-
nate vein.
Pulmonary valve stenosis and Fallot’s tetralogy
Percutaneous intervention on isolated pulmo-
nary valve stenosis was the procedure that ushered
in the current era of catheter-based therapies
(Fig. 20). Although angiographic definition of
the right ventricular outflow tract and valve is
not complicated, several features must be kept in
mind when approaching angiography for an

Fig. 14. Left panel shows an ascending aortogram taken with a shallow-LAO projection without caudal angulation. The
catheter was placed through a transeptal entry to the left heart. Although the area of the coarctation can be seen, it is the
caudal angulation that identifies the details of the lesion, including a small ductal ampulla (right panel).
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
Fig. 15. Intervention on the aortic valve requires accurate definition of the hinge points of the leaflets. In the two left
panels, long axis oblique views from an ascending aortogram, the margins of the leaflets are not defined because of over-
lap of the cusps (bicuspid in these examples). In the two right panels, long axis oblique (left) and RAO views, the left
ventriculogram allows easier identification of the leaflet hinge points, where measurements can be made.
interventional procedure. In the case of isolated
pulmonary valve stenosis and other right ventric-
ular outflow tract lesions, because the outflow
tract can take a horizontal curve, a simple AP pro-
jection foreshortens the structure. Therefore, a 30
cranial with 15 LAO opens up the infundibulum,
allowing visualization of the valve and the main
and branch pulmonary arteries. The best defini-
tion of the hinge points of the valve, to choose
the correct balloon size, is from the left-LAT pro-
jection. Occasionally, 10 or 15 caudal angula-
tion of the LAT detector can be used to separate
the overlap of the branch vessels seen on a straight
left-LAT projection. This is not recommended,
however, as it also foreshortens the outflow tract
and the valve appears off plane, giving incorrect
valve diameters.
Branch pulmonary artery stenosis
Pulmonary artery interventions are common
and represent the most difficult angiographic
projections to separate out individual vessels for
assessment and potential intervention (Figs. 21
and 22). A cranially tilted frontal projection
with left-LAT or RAO/LAO projections fre-
quently is the first series of views that can be per-
formed as scout studies to map the proximal and
hilar regions of the pulmonary circulation. The in-
jection may be performed either in the ventricle or
main pulmonary artery. Because there is frequent
overlap in viewing the right ventricular outflow
tract (see previous discussion), these standard
views can be modified by increasing or decreasing
the degree of RAO or LAO and adding caudal or
cranial tilt. Selective branch artery injections are
best for detailed visualization to plan the interven-
tion. For the right pulmonary artery, a shallow
RAO projection with 10 or 15 cranial tilt
547
separates the upper and middle lobe branches,
whereas a left LAT with 15 caudal tilt opens up
all the anterior vessels. Similarly, to maximize
the elongated and posterior leftward directed left
pulmonary artery, a 60 LAO with 20 cranial is
effective, with a caudal tilt on the lateral detector.
The changing indications for cardiac
catheterization in adult congential heart disease
There has been no greater impact on the care
of patients who have congenital heart disease
than the new imaging modalities. The heart,
once accessible only by surgeons, angiocardiogra-
phers, and pathologists, now can be sliced safely
and accurately, rotated, and examined with mini-
mal discomfort to patients. The development of
better imaging has had a great impact on the abil-
ity to make decisions and plan percutaneous inter-
ventions and surgery.
Unfortunately, the clinical appreciation of
patient anatomy can become quite befuddled as
years go by and patients are cared for by different
pediatricians and adult cardiologists or even lost
to specialized follow-up. This is never so true as
when geographic migrations occur and patients
leave a pediatric hospital where they were cared
for initially and arrive in a new city without their
medical file.
Clinical information about patients has a hier-
archy in terms of relevance and importance. A
tattered 25-year-old surgical report may be the
Holy Grail of information. A seasoned surgeon’s
operative report may describe carefully native and
surgical anatomy in great detail. The report may
provide insight into what was repaired and how
and why. A close second in the hierarchy is
a good-quality CT scan or MRI by an experienced
548 MCLAUGHLIN et al

Fig. 16. Baffle obstruction after a Mustard operation is, as the population ages, an increasingly common event. This is
important particularly when such patients need transvenous pacing devices. In (A) (left panel), the presence of a superior
baffle obstruction can be identified from the left LAT projection. Only with cranial angulation (cranial-LAO view) (right
panel), however, is the full extent of the lesion detailed. This is critical (B), particularly where the frontal view (left panel),
does not show the full extent of the obstruction, and only from the angulated view are the length and diameter of the
lesion outlined (middle panel). A stent is placed, followed by a transvenous pacing system, as shown in the right panel
from a frontal projection. For an inferior baffle lesion, the frontal (PA) projection is optimal (C), before (left) and after
(right) a stent is placed.
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE 549

Fig. 17. Use of angiography for septal defect definition and device placement in the setting of a secundum ASD has been
supplanted by intracardiac and transesophageal techniques (A). Fluoroscopy still is required for initial device localiza-
tion, however, and in many laboratories, a short cine-run to record the diameter of the static balloon diameter to choose
device size. In this case, the authors find the 30 LAO with 30 cranial tilt to best elongate the balloon to avoid fore-
shortening (B).

imager. The latest generation of imaging equip-
ment from the major modalities of echo, MRI,
and CT permits the appreciation of the most
subtle anatomic detail.
More important than the newest technology
and costliest equipment is the availability of
imaging specialists to conduct the study, interpret
it expertly, report it insightfully, and caution if
there are limitations that should be known about.
Collaboration between image and clinical, surgi-
cal, and interventional physicians allows the in-
tegration of knowledge and facilitates the delivery
of excellent patient care. Little is gained from the
ability to produce beautiful images that are inter-
preted in a way that is not meaningful to clinicians.
It is as important for radiologists to know the
concerns of surgeons and interventionalists as is
the reverse. Choosing the right modality to answer
a particular question or set of questions is key, as is
providing imagers with the specific questions to be
addressed so that correct protocols are used to
obtain the information required.
Echocardiography
Echocardiography always will play an impor-
tant role in the care of patients who have ACHD.
Its availability, portability, and familiarity to
most recently trained cardiologists is a strength.
Echocardiographers trained to interpret complex
anatomy in congenital heart disease are a neces-
sity. The ability to assess patients rapidly at the
bedside in an emergency department, during off
hours, and in an ICU is a definite benefit,
especially when patients are critically ill. Echo
provides important structural and physiologic
information and is the noninvasive reference
standard for valvular assessment. TEE or, more
recently, intracardiac echocardiography is helpful
especially during interventional procedures. These
550 MCLAUGHLIN et al

Fig. 18. Left panel shows the appearance of a fenestrated extracardiac Fontan in the frontal projection, the right panel
its appearance after device closure. Generally, a frontal projection profiles the defect adequately, but at times some an-
gulation is required, where the defect is profiled best in a shallow RAO view. Also, note coils in the left SVC, which
developed after the Fontan procedure and required embolization. Occasionally, collateral vessels develop from the
hepatic/phrenic vein or innominate vein, the primary view being frontal and left LAT.

modalities allow the real-time monitoring and
detailed evaluation of device implantation or
valvular intervention in an unobtrusive and min-
imally limiting fashion.
Cross-sectional imaging
MRI and CT are critical imaging modalities for
clinicians. MRI provides tremendous anatomic
detail and functional information. Gradients
across valvular orifices, planimetry of valve areas,
relative pulmonary flow, and collateral vessel flow
are in the repertoire of experienced imagers. Hints
are provided to the presence of small ASDs or
baffle leaks. The ability to postprocess a data set
adds to its versatility and permits post hoc
evaluation of new questions immediately or years
in the future. The examinations are lengthy and
require significant time to reconstruct the data.
MRI, however, lacks global applicability to all
patients. Patients who have pacing devices or
similar implants, cerebral aneurysm clips, or for-
eign bodies in an eye may not be studied at this
time. Patients who have nonplatinum coils, stents,
or devices may be imaged but are not well suited to
this investigation. The artifact caused by many
devices distorts the magnetic field and renders the
area of interest inaccessible. As the examinations
are lengthy and many MRI bores still are closed,
narrow, and dark, it is not uncommon for patients
to be unable to tolerate an examination for reasons
of discomfort or claustrophobia. It is possible to
prepare patients properly with information or
medication to overcome this problem but if

Fig. 19. Because of an offset in the anastomosis between the SVC and right pulmonary artery, the optimal view to see
the anastomosis without overlap is a shallow onedwith caudal tilt as seen in the right panel. In the left panel, in the
frontal projection, there is overlap of the anastomosis which obscures a potential lesion, as seen in the angulated
view. The combination of an angulated frontal detector and caudal angulation of the lateral tube allows definition of
the anastomosis and the pulmonary artery confluence.
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE 551

Fig. 20. Left panel depicts a case of typical isolated pulmonary valve stenosis in a neonate. The outflow tract is profiled
in the cranially angulated frontal projection, with a slight degree of LAO angulation. The right ventriculogram outlines
the form of the ventricle, the main pulmonary artery (and ductal bump), and the pulmonary artery confluence and
branch dimensions. The LAT view (right panel) outlines the valve leaflets (thickened and doming) and allows accurate
delineation of the valve structures for balloon diameter determination.

unanticipated, this usually results in a cancelled or
partially performed examination that is unsatisfy-
ing for patients and institutions.
Because of new high-speed multidetector scan-
ners, CT scanning has experienced a revival. The
ability to gate these scans has allowed the pro-
duction of elegant images previously not possible.
At present, this requires sufficiently low heart
rates (60–70 bpm) and the ability to breath-hold.
CT offers the advantage of superior visualization
in and around metallic implants, such as stents
placed in the aorta for coarctation. CT also has
the ability to visualize the coronaries and is
beginning to approach the diagnostic sensitivity
of angiography. This milestone is anticipated,
although it is yet to come to fruition. The large
doses of radiation required to generate elegant
images make serial CT examinations unattractive
for the follow-up of young patients over the
course of a lifetime.
Cardiac catheterization
What is the role of catheterization? Is it only
a dinosaur in the face of such advanced imaging?
Is it barbaric to use catheters to measure pressures
and inject dye directly into a cardiac chamber for
fleeting seconds when a simple peripheral intrave-
nous injection suffices for detailed cross-sectional
imaging?

Fig. 21. Angiography for selective intervention on the branch pulmonary arteries can be most difficult because of over-
lapping of structures. No single projection is totally representative and multiple views frequently are required. In the left
panel, a scout film is taken in the main pulmonary artery, and in the right, the right ventricle. Both images are taken in
the cranial-LAO projection and in these examples clearly outline the outflow tracts and branch confluences. In the left
panel, the dilated main pulmonary artery would have obscured the branch pulmonary artery confluence, and this cranial
LAO (left upper panel) and caudal left LAT (right upper panel) nicely detail the anatomy for subsequent intervention.
552 MCLAUGHLIN
Fig. 22. The image is taken from a left-LAT projection
with caudal tilt. This separates the proximal right and
left pulmonary artery branches and details the main pul-
monary artery. The outflow tract is foreshortened, and
this view misleads operators when examining the diame-
ter of the valve and the infundibulum. When examining
the infundibulum and the diameter of the valve,
a straight left LAT should be performed. In the cau-
dal-LAT projection, the left pulmonary branch sweeps
superiorly and toward the upper right corner of the im-
age, whereas the left pulmonary artery appears more me-
dial and in the center of the image. Using the left-LAT
view, stents could be placed in each branch.
In counterpoint, catheterization is the only
modality that provides the gold standard of
pressure measurement in a vessel or chamber. In
stark contrast to the complexities of the newest
technology, the measurement of intracardiac pres-
sures is simple, reliable, and reproducible. The
assessment of the hemodynamic significance of
a lesion never should be left to cross-sectional
imaging and always verified by catheter. Some
argue that an aortic valve never should be
changed without a patient having had a catheter
examination [11,12]. This may be interpreted as an
old way of doing things by some, but by others it
is a refreshing confirmation of the value of inva-
sive imaging. Some send patients for operative
correction of a defect without catheterization,
but they may on occasion fall victim to this ap-
proach; an anomalous coronary may be missed
and transected or ligated or an opportunity to ad-
dress an unrecognized defect may be missed.
Catheterization is the only method to deter-
mine the pulmonary artery pressure and pulmo-
nary vascular resistance accurately. These values
et al
are of utmost importance in many patients who
have congenital heart disease. The time-honored
practice of oximetry and shunt determination is
a confirmatory piece of information and the
weight placed on it is reflected in our current
guidelines for intervention in congenital heart
disease [13].
There are several situations where noninvasive
imaging cannot provide the anatomic detail re-
quired for decision making. The recognition that
noninvasive imaging may not assess the lumen of
a pulmonary artery or collateral vessel after
stenting reliably may lead to a further intervention
that could improve patient quality of life.
Coronary angiography provides the only stan-
dard used to assess coronary lesions and their
suitability for revascularization. The addition of
invasive physiologic (coronary flow reserve) and
imaging modalities (intravascular ultrasound) can
make the resolution of a clinical question re-
garding lesion severity a straightforward issue.
Diagnostic catheterization is alive and well in
congenital heart disease.
A look at today’s interventions
Most adult congenital catheterization practices
consist of approximately 70% intracardiac device
implantation and approximately 30% other in-
terventions. The world of device closure for ASDs
is divided between ASD and PFO closure.
Atrial septal defect closure
ASDs are present in approximately 0.317 of
every 1000 live births [14]. A variety of devices has
been used to close these defects, generally for the
indication of right ventricular volume overload.
The guidelines for these interventions are drawn
from the Canadian and European grown up con-
genital heart recommendations [13,15]. ASD clo-
sure is a well-established and safe procedure. It
can be performed on an outpatient basis with in-
tracardiac echocardiography guidance and with-
out the need for anesthesia or TEE. Such an
approach presumes that patients have undergone
a detailed TEE study to exclude associated abnor-
malities including, but not limited to, anomalous
pulmonary venous return, additional secundum
defects, fenestrations and sinus venosus defects,
and the occasional septum primum defect. MRI
and CT provide excellent detail with regard to pul-
monary venous drainage but are somewhat lacking
in specificity and sensitivity when it comes to small
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
fenestrations or other small coexisting ASDs. A
TEE also excludes left atrial thrombus and confirms
the absence of significant valvular heart disease.
The most versatile device for this application is
the Amplatzer Septal Occluder. It provides the
benefit of a self-centering design (an issue with
double umbrella devices) and is easy to implant.
The residual leak rate after placement arguably is
the lowest of any device and has the lowest rate of
visible thrombus formation of any current device
[16,17]. Many series of short- and medium-term
results are published and demonstrate excellent re-
sults. Patients younger than age 40 generally have
a complete return of their dilated right ventricular
size to normal over the after 6 months (much of
the improvement occurs immediately after the
procedure) [18]. Many patients over 40 enjoy the
same results. Most adults demonstrate objectively
improved cardiopulmonary function after closure,
including those who believe they were asymptom-
atic before the procedure [19–21]. There is a sug-
gestion that patients who do not have
arrhythmia before age 55 and who undergo clo-
sure may enjoy a lower risk of subsequently devel-
oping an arrhythmia [22].
Although a well-established procedure, the
long-term safety of ASD device closure remains
in question. Several late device erosions into the
pericardium, aorta, or other structures are re-
ported. The erosion rate is believed to involve
approximately 0.1% of cases [23,24]. Caution is
advised in cases of a deficient superior or aortic
rim, with aggressive balloon sizing and device
oversizing, and with aggressive maneuvers (push-
pull) to verify device stability. Some advocate
the abandonment of balloon sizing but have not
presented data that this is a safer approach. The
prospect of placing a large device (which results
in remodeling and shrinkage of the right heart)
into a small person remains a source of concern.
Programs that do a lot of these procedures en-
counter a severe complication at some point.
Patent foramen ovale closure
Although indications for ASD closure gener-
ally are accepted, the indications for closure of
PFO for the secondary prevention of stroke are
not. Although the evidence of benefit is limited,
the number of procedures performed and the
number of operators performing them continue
to grow. Two major trials currently are under way
examining whether or not device therapy is better
or worse than medical therapy for secondary
553
stroke prevention. These trials are slow to recruit
patients and are in jeopardy of not being com-
pleted. The data arguing for closure rest mainly
on evidence provided by single-center, retrospec-
tive, nonrandomized trials and with meta-analyses
that have taken unfortunate liberties in interpre-
tation and analysis [25,26]. The result is an unclear
future for this procedure. Which patients will ben-
efit from a percutaneous PFO closure procedure?
Will it be only those who have atrial septal aneu-
rysms or those who have multiple prior events?
Will the complications of PFO closure, including
arrhythmia, device thrombus formation, recurrent
cerebrovascular events, or access complications,
make this therapy less attractive than medical
therapy? Will patients who have other risk factors
for stroke, such as diabetes, or those who have
nonsurgical carotid disease benefit from closure
as a risk reduction strategy? The authors’ policy
regarding patients referred for PFO closure in-
volves a detailed informed consent where many
of these issues and questions are raised. The au-
thors offer every appropriate patient entry into
a randomized trial to help resolve these issues
and tell each patient that there is no conclusive
data to support closure at this time, a position
supported by the American Academy of Neurol-
ogy [27].
The treatment and follow-up of these patients
who have ASD or PFO after device closure is
another area of controversy. Most operators treat
with ASA and clopidogrel for periods varying
from 1 to 6 months or longer. The use of bubble
studies to follow these patients who have trans-
thoracic echocardiography (TTE) or TEE to
‘‘confirm closure’’ often is done without any
evidence that this type of testing is important or
helpful.
Coarctation of the aorta
Stent placement for coarctation of the aorta is
perhaps the most hazardous of all interventions
performed in the catheter laboratory. Patients
who have either native coarctation or previously
treated coarctation may be candidates for treat-
ment. In most patients, the intervention is in-
dicated for a gradient of greater than 20 mm Hg
across the coarctation site, usually in the setting of
proximal hypertension. The authors also have
intervened to treat pseudoaneurysms with covered
stents and coils. Over the past several years, the
authors have modified their approach to this
procedure to improve its safety. They routinely
554 MCLAUGHLIN
obtain access from the right radial artery and the
right femoral artery. This allows for simultaneous
pressure measurement before and after stenting
and immediate angiography after stent placement
to rule out aortic dissection or perforation. This
approach always has been one of direct stenting
(primary stenting) in adults, without progressive
balloon inflation until dissection is noted, and
then stenting as advocated by others. The authors
began with the use of the Palmaz Shatz P5014b
stent (Jonhson & Johnson Interventional, Warren,
New Jersey), which provided excellent radial
strength but relatively poor flexibility, and then
moved to the Genesis Biliary Stent (Johnson &
Johnson), which provided less radial strength in
return for flexibility. In follow-up, they have
seen the Genesis stents buckle under the recoil
of the aorta and also noted circumferential frac-
tures. Most recently, they have used the CP
PTFE-covered stent, which offers protection
from aortic perforation and dissection, the most
serious complications of this procedure. Also,
for years they have been addressing the femoral
access site as a source of complication by fully an-
ticoagulating these patients and preclosing the ac-
cess site using a suture-mediated closure device
(Perclose AT Abbott Vascular Devices, Redwood
City, California). The authors find that complete
hemostasis is possible with few complications.
The large arterial access (up to 14 Fr) required
for this procedure remains a source of concern.
Patent ductus arteriosus
Patent ductus arteriosus (PDA) closure in
adults usually is performed to reduce the risk of
endarteritis. It is said that all ducts with an
audible murmur should be closed. The interven-
tion on PDAs has been facilitated greatly using
the Amplatzer Duct Occluder (AGA Medical,
Golden Valley, Minnesota). This is a well-designed
device, which has little competition in the adult
marketplace. This is a safe and quick procedure,
which almost uniformly corrects the intended an-
atomic abnormality.
The authors, and others, have used this versa-
tile device successfully to close perivalvular leaks
around previously implanted mitral or aortic
valve prostheses. They have limited these inter-
ventions to patients who have indications of
severe mitral insufficiency associated with heart
failure or transfusion-dependent hemolysis who
are not surgical candidates or who are high-risk
surgical candidates.
et al
A look at new and emerging interventions
The next great series of innovations in cardio-
vascular intervention will be in valvular heart
disease, such as the Bonhoeffer-Melody (Med-
tronic) stented pulmonary valve. In addition, per-
cutaneous valve replacement has been achieved in
humans with two different aortic valve prostheses.
Pulmonary valve implantation initially will be of-
fered to patients who have previously implanted
conduits. At present, the largest percutaneous pul-
monary valve is 22 mm. This size will limit its use
to a select patient population for the time being.
In short order, new stent designs or other inter-
ventional techniques should be expected to allow
the treatment of patients who have larger conduits
and native outflow tracts.
There likely will be a growth of hybrid
surgical/interventional procedures that involve
sophisticated imaging to examine physiology and
anatomy immediately before and after repair.
These hybrid procedures are performed rarely in
the pediatric world and even more rarely in the
adult setting. Until stented valves are able to treat
the pulmonary insufficiency seen in aneurysmal
outflow tracts in patients who have tetralogy,
a novel hybrid approach can be envisioned.
Surgeons may, through a minimally invasive
approach, plicate the pulmonary artery or suture
a percutaneous valve in place from the epicardial
surface on a beating heart without cardiopulmo-
nary bypass. Valves in the aortic position have
been replaced via a periapical approach and
a minithoracotomy, obviating femoral arteries
large enough to handle a very large delivery
system.
New structural heart disease suites, which take
into account the need for operating room stan-
dard ventilation, anesthesia, and perfusion, will be
required. There will continue to be a proliferation
of minimally invasive procedures whereby surgery
will facilitate the introduction of a device either
percutaneously or directly on a beating heart,
obviating cardiopulmonary bypass. Eventually,
many of these procedures mostly will be percuta-
neous with surgeons and interventional colleagues
working side by side in pursuit of the safest and
most effective way to perform repair or replace-
ment of a particular valve or vessel.
Surgical strategies for repair will take into
account interventional advances, and childhood
operations will be modified so that an interven-
tional solution to a future reoperation may be
possible. An example may be the performance of
 CARDIAC CATHETERIZATION AND CONGENITAL HEART DISEASE
a modified hemi-Fontan to allow for catheter
laboratory completion of the Fontan with a cov-
ered stent from the IVC to the PA. There may be
implantation of fewer mechanical valves to permit
the percutaneous implantation of tissue valves
within failing surgically implanted stented tissue
valves, which will serve as a matrix on which to
build.
Guidance for these new procedures likely will
involve 3-D imaging modalities and device ma-
nipulation and stabilization during the proce-
dures, requiring that new expertise be developed.
We live in interesting times, but they are certain to
become more interesting, as structures as divinely
inspired as the mitral valve become the substance
of everyday catheter laboratory repair.
The future role of diagnostic catheterization in the
patients who have adult congenital heart disease
It is probable that what is known of diagnostic
catheterization will change dramatically. A di-
agnostic catheterization may be expected to be
performed in an imaging suite. Who the operator
will be remains in question. A structural map may
be created with cross-sectional imaging and then
the navigation through tortuous pulmonary ar-
teries or vessel occlusions or transeptally will be
done with a 0.014 wire with a pressure transducer
and a radiofrequency ablation assembly at its tip.
The guidance of this small device may be solely
magnetic. A nurse may place a peripheral in-
travenous line through which such a device is
introduced. Oximetry may be measured in various
chambers using similar wire tip technology with-
out the need for blood sampling. The need for
large devices, such as balloons and stents, will
continue to require access to a large or central
vein to permit their introduction.
Patients who have ACHD may no longer wait
for different appointments for different imaging
modalities. They will pass through a series of
scanners in the course of an hour. The coronaries
may be visualized via CT, the stenosed pulmonary
artery and the gradient across it assessed by MR,
followed by the crossing of the stenosis by
a magnetically directed wire over which a balloon
and stent will be introduced.
Will contrast angiography survive independent
of interventional catheterization? It may not, at
least not in its present form. Diagnostic catheter-
ization will not disappear, but it will change. It
will be less invasive, use less or no contrast, and
may take place with the operator sitting in
555
a control room using navigational equipment.
To run an aircraft carrier, one need not turn
a crank to drive the propeller.
Summary
This short introduction to diagnostic and
interventional heart catheterization in patients
who have ACHD allows readers a point of
departure for the invasive assessment and inter-
ventional treatment of the most common lesions.
Many cases occur, however, that do not fall into
a standard categorization and operators must be
prepared to remember the basic principles out-
lined in this article and use creative approaches to
define and treat the lesion optimally. Successful
outcomes require patience, perseverance, and the
learned experience of others.
References
[1] Angelini P. Abnormalities of the coronary arteries.
Normal and anomalous coronary arteries: defini-
tions and classification. Am Heart J 1989;117:
418–34.
[2] Greenberg MA, Fish BG, Spindola-Franco H. Con-
genital anomalies of the coronary arteries. Radiol
Clin North Am 1989;27:1127–46.
[3] Freedom RM, Culham JAG. Abnormalities of the
coronary arteries. In: Freedom RM, Mawson JB,
Yoo SJ, et al, editors. Congenital Heart Disease,
textbook of angiocardiography. Armonk, NY:
Futura Publishing Company, Inc.; 1997. p. 849–78.
[4] Ishikawa T, Brandt PWT. Anomalous origin of the
left main coronary artery from the right anterior aor-
tic sinus: angiographic definition of anomalous
course. Am J Cardiol 1985;55(6):770–6.
[5] Serota H, Barth CW, Seuc CA, et al. Rapid identifi-
cation of the course of anomalous coronary arteries
in adults: the ‘‘dot and eye’’ method. Am J Cardiol
1990;65:891–8.
[6] Culham JAG. Physical principles of image forma-
tion and projections in angiocardiography. In:
Freedom RM, Mawson JB, Yoo SJ, et al, editors.
congenital heart disease textbook of angiocardio-
graphy. Armonk, NY: Futura Publishing; 1997.
p. 39–93.
[7] Beekman RH 3rd, Hellenbrand WE, Lloyd TR,
et al. ACCF/AHA/AAP recommendations for train-
ing in pediatric cardiology. Task force 3: training
guidelines for pediatric cardiac catheterization and
interventional cardiology endorsed by the Society
for Cardiovascular Angiography and Interventions.
J Am Coll Cardiol 2005;46:1388–90.
[8] Qureshi SA, Redington AN, Wren C, et al. Recom-
mendations of the British Paediatric Cardiac Associ-
ation for therapeutic cardiac catheterisation in
556 MCLAUGHLIN
congenital cardiac disease. Cardiol Young 2000;10:
649–67.
[9] Ventricular septal defect. In: Freedom RM, [19]
Mawson JB, Yoo SJ, et al, editors. Congenital heart
disease textbook of angiocardiography. Armonk,
NY: Futura Publishing; 1997. p. 189–218.
[10] Brandt PW. Axially angled angiocardiography.
Cardiovasc Intervent Radiol 1984;7:166–9. [20]
[11] Griffith MJ, Carey C, Coltart DJ, et al. Inaccuracies
of using aortic valve gradients alone to grade severity
of aortic stenosis. Br Heart J 1989;62:372–8.
[12] Rahimtoola SH. Should patients with asymptomatic [21]
mild or moderate aortic stenosis undergoing coro-
nary artery bypass surgery also have valve replace-
ment for their aortic stenosis? Heart 2001;85:337–41.
[13] Therrien J, Dore A, Gersony W. CCS Consensus [22]
Conference 2001 update: recommendations for the
Management of Adults with Congenital Heart
Disease Part I. Can J Cardiol 2001;17:940–59.
[14] Ferencz C, Rubin JD, McCarter RJ, et al. Congeni- [23]
tal heart disease: prevalence at livebirth. The Balti-
more-Washington Infant Study. Am J Epidemiol
1985;121:31–6.
[15] Deanfield J, Thaulow E, Warnes C, et al. Manage-
ment of grown up congenital heart disease. The [24]
Task force on the Management of Grown up
Congenital Heart Disease of the European Society
of Cardiology. Eur Heart J 2003;24:1035–84.
[16] Anzai H, Child J, Natterson B, et al. Incidence of [25]
thrombus formation on the CardioSEAL and the
Amplatzer interatrial closure devices. Am J Cardiol
2004;93:426–31.
[17] Krumsdorf U, Ostermayer S, Billinger K, et al. Inci- [26]
dence and clinical course of thrombus formation on
atrial septal defect and patient foramen ovale closure [27]
devices in 1,000 consecutive patients. J Am Coll
Cardiol 2004;43:302–9.
[18] Veldtman GR, Razack V, Siu S, et al. Right ventric-
ular form and function after percutaneous atrial
et al
septal defect device closure. J Am Coll Cardiol
2001;37:2108–13.
Giardini A, Donti A, Formigari R, et al. Determi-
nants of cardiopulmonary functional improvement
after transcatheter atrial septal defect closure in
asymptomatic adults. J Am Coll Cardiol 2004;43:
1886–91.
Webb G, Horlick EM. Lessons from cardiopulmo-
nary testing after device closure of secundum atrial
septal defects: a tale of two ventricles. J Am Coll
Cardiol 2004;43:1892–3.
Brochu MC, Baril JF, Dore A, et al. Improvement in
exercise capacity in asymptomatic and mildly symp-
tomatic adults after atrial septal defect percutaneous
closure. Circulation 2002;106:1821.
Silversides CK, Siu SC, McLaughlin PR. Symptom-
atic atrial arrhythmias and transcatheter closure of
atrial septal defects in adult patients. Heart 2004;
90:1194–8.
Amin Z, Hijazi Z, Bass JL. Erosion of Amplatzer
septal occluder device after closure of secundum
atrial septal defects: review of registry of complica-
tions and recommendations to minimize future
risk. Catheter Cardiovasc Interv 2004;63:496–502.
Divekar A, Gaamangwe T, Shaikh N, et al. Cardiac
perforation after device closure of atrial septal
defects with the Amplatzer septal occluder. J Am
Coll Cardiol 2005;45:1213–8.
Windecker S, Wahl A, Nedeltchev K, et al. Compar-
ison of medical treatment with percutaneous closure
of patent foramen ovale in patients with cryptogenic
stroke. J Am Coll Cardiol 2004;44:750–8.
Landzberg MJ, Khairy P. Indications for the closure
of patent foramen ovale. Heart 2004;90:219–24.
Messe SR, Silverman IE, Kizer JR, et al. Practice pa-
rameter: recurrent stroke with patent foramen ovale
and atrial septal aneurysm: report of the Quality
Standards Subcommittee of the American Academy
of Neurology. Neurology 2004;62:1042–50.
 Cardiol Clin 24 (2006) 557–569

The Geneticist’s Role in Adult

Congenital Heart Disease
Francois P. Bernier, MD*, Renee Spaetgens, MD
Department of Medical Genetics, University of Calgary, Alberta Children’s Hospital,
1820 Richmond Road, SW Calgary, AB T2T5C7, Canada

The Human Genome Project promises to
revolutionize our understanding and approach
to human diseases. Specifically for congenital
heart defects (CHDs), this project promises to
identify the genes and pathways involved in
normal cardiac development and to determine
the impact of mutations in these genes in the
pathogenesis of syndromic and nonsyndromic
CHDs. For clinicians involved in the care of
adults who have congenital heart disease, some
of these advances in genetic knowledge are now
being translated into clinical care, raising the
question of what role this new era of genetic
knowledge should have in the adult congenital
heart disease (ACHD) clinic. The authors sum-
marize the clinical and molecular advances rele-
vant to the care and genetic counseling of ACHD
patients and explore the role of genetic care
providers in an ACHD clinic.
The Role of genetics
Medical geneticists and genetic counselors are
involved in the diagnosis, treatment, and counsel-
ing of patients who have or who are at risk for
genetic diseases. In a pediatric setting, the focus is
often on the short- and long-term medical and
developmental outcomes for the affected child and
on recurrence risks for the parents. In adult
patients, the role expands to provide accurate
information on the risk of having an affected child
(recurrence risk) and what options, if any, are
available for prenatal diagnosis. The accuracy of
* Corresponding author.
E-mail address: francois.bernier@calgaryhealthregion.ca
(F.P. Bernier).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.001
this genetic counseling is primarily determined by
the accuracy of the patient’s diagnosis. For many
patients in an ACHD clinic, their last medical
genetics evaluation (if one was ever completed)
may have taken place many years ago and before
the significant advances that have occurred in the
interim. For example, in many centers, newborns
who have conotruncal CHDs are now routinely
tested for 22q11 deletions, whereas few ACHD
patients, if any, will likely have been tested. The
importance of an accurate diagnosis is further
exemplified by the fact that recurrence risks vary,
in general, between zero (the risk associated, for
example, with a teratogenic etiology) and 50% (the
risk associated with an autosomal dominant con-
dition). Providing inaccurate genetic counseling
may not only falsely reassure or scare a prospective
couple but also present a significant medical legal
liability. In addition, the recurrence risk may apply
not only to the risk of having a child with a CHD
but also to a range of other birth defects or
learning disabilities in those who have syndromic
diagnoses. In an ACHD clinic, medical genetics
can play a significant role in the clinical and
genetic assessment of patients, with the goal of
arriving at an accurate etiologic diagnosis to
provide accurate counseling regarding recurrence
risk and prenatal diagnosis and screening options.
Toward an etiologic classification of congenital
heart defects
Historically, CHDs were simply classified ac-
cording to their physiologic effects. More recently,
a developmentally oriented classification has been
advocated by molecular biologists, particularly
those interested in exploiting the power of model
cardiology.theclinics.com
558 BERNIER & SPAETGENS
systems for studying cardiac development given the
apparent conservation of cardiac developmental
steps in vertebrates [1]. Ultimately, an etiologic
classification is anticipated. Historically, CHDs
were classified as isolated or associated with a syn-
drome, with the former being mainly thought to be
due to multifactorial inheritance with a minority re-
sulting from single-gene, chromosomal, or terato-
genic causes. Molecular advances are challenging
this assumption, and it is anticipated that the large
‘‘multifactorial’’ group of patients may be much
more etiologically heterogeneous [2]. Nevertheless,
the assignment of an ACHD patient into one of the
above etiologic categories remains an important
clinical exercise.
Genetics of isolated congenital heart disease
It has long been recognized that congenital
heart disease tends to cluster in families. Most
families demonstrate a multifactorial mode of
inheritance, whereby the risk of congenital heart
disease increases when one or more closely related
family members is affected. A number of rare
families who demonstrate a classic mendelian
pattern of inheritance of isolated CHDs have
recently helped in the identification of several
single-gene causes of CHDs. Although their over-
all contribution to sporadic congenital heart
disease is likely to be low, the hope is that their
identification will provide knowledge about the
underlying mechanisms of heart defects that could
ultimately result in prevention or new therapeutics
and improved clinical outcomes. To date, only
a small number of single-gene causes of isolated
heart defects have been identified, suggesting that
CHDs in general are extremely heterogeneous and
genetically complex.
Identification of single-gene causes of congen-
ital heart disease has been largely accomplished
with linkage studies in large families who have
CHDs, from the study of recurrent chromosomal
rearrangements, or based on the identification of
the molecular etiology of rare genetic syndromes
[3]. Described in the following paragraphs are
a number of single-gene causes of isolated CHDs.
GATA4 encodes a transcription factor known
to play a critical role in cardiogenesis. It is local-
ized to chromosome 8p23.1. Pathogenic muta-
tions were first identified in a large family who
had isolated autosomal dominant ostium secun-
dum atrial septal defects (ASDs) and have since
been identified in other families who have isolated
CHDs [4–7]. Although ASDs appear to be the
most common type of heart defect associated
with mutations in GATA4, additional defects
have been seen, including ventricular septal de-
fects (VSDs), atrioventricular septal defects
(AVSDs), and pulmonary valve stenosis [4].
A novel mutation in GATA4 has also been found
in two patients with isolated tetralogy of Fallot
(TOF) [8]. The molecular mechanism by which
GATA4 mutations cause septation defects (and
others types of CHDs) remains to be elucidated,
although it appears that at least some of the path-
ogenic mutations result in reduced transcriptional
activation of downstream target genes [8]. The
overall contribution of GATA4 mutations to fa-
milial CHD is probably modest. In one study,
only 2 of 16 probands with familial ASD were
found to have GATA4 mutations [5].
NKX2-5 encodes for a homeobox transcription
factor that interacts functionally with GATA4. It is
involved in the early determination of the heart
field and may play a role in formation of the car-
diac conduction system. Knockout animal models
have demonstrated that the absence of NKX2-5
leads to acardia or severe defects in cardiac mor-
phogenesis [9]. Mutations in NKX2-5 have been
identified in several families who have familial non-
syndromic ASD with conduction defects [10]. It
has also been less commonly associated with a vari-
ety of other heart defects including VSD, Ebstein’s
anomaly, TOF, fibromuscular subaortic stenosis,
double-outlet right ventricle, and hypoplastic left
heart. In addition, the NKX2-5 mutation can cause
atrioventricular conduction block with or without
associated congenital heart diseases, and all
NKX2-5 mutation carriers are thought to have
a lifelong risk for the development of heart block
and sudden death. To date, approximately 20 het-
erozygous mutations in NKX2-5 have been re-
ported to be associated with nonsyndromic
congenital heart disease [11]. Most mutations re-
sult in protein truncation or an altered DNA bind-
ing domain. Overall, the impact of NKX2-5
mutations in sporadic isolated CHDs appears to
be small [12].
Several other single-gene causes of congenital
heart disease are known. Missense mutations in
CRELD1, a cell adhesion molecule expressed dur-
ing cardiac development, have been identified in pa-
tients who have nonsyndromic sporadic AVSDs
[13]. Mutations in ZIC3, an X-linked transcription
factor, have been found in sporadic and X-linked
laterality defects, and less commonly in transposi-
tion of the great arteries (TGA) [14,15]. Mutations
in JAG1, the gene that causes Alagille syndrome,
 GENETICIST’S ROLE
have been identified in familial cases of isolated
right-sided cardiac lesions [16]. Elastin mutations
are seen in dominantly inherited isolated supraval-
var aortic and pulmonic stenosis [17].
Molecular genetic testing for mutations in the
abovementioned genes is not yet available on
a clinical basis. As we learn more about the
underlying genetic causes of congenital heart
disease and as genetic testing becomes clinically
available, important issues regarding genetic test-
ing will become important. These issues include
appropriate timing of testing, medical insurability,
confidentiality, and cost of genetic testing. Clinical
utility and relevance to patient management will
need to be demonstrated before testing becomes
routinely available.
Teratogenic causes of congenital heart disease
Environmental factors also play an important
role in the development of CHDs. A number of
teratogens are known to be associated with
CHDs, the most notable of which are maternal
diabetes and alcohol. Heart defects that are the
result of a teratogenic exposure are not associated
with an increased risk of CHDs in siblings or
offspring, assuming the exposure does not occur
in subsequent pregnancies.
Fetuses born to mothers who have diabetes
mellitus are at increased risk of heart defects as
isolated malformations or in association with
other features of diabetic embryopathy (ie, caudal
regression). The risk of heart defects in mothers
who have diabetes is 3% to 5% and is dependent
on glycemic control. In those who have good
glycemic control, the rates of CHDs are similar to
the general population risk. The most common
lesions associated with maternal diabetes are
VSDs, coarctation of the aorta, and TGA. There
is also an increased risk for laterality defects. The
association of diabetes with heart defects suggests
that this embryologic pathway is particularly
susceptible to disruption by disturbances of glu-
cose homeostasis. Other maternal medical condi-
tions are also associated with risk for CHDs in
offspring. Maternal rubella infection carries
a 35% risk of heart defects, although widespread
vaccination for rubella should prevent most con-
genital rubella infections and associated heart
defects. Maternal systemic lupus erythematosis
can cause complete heart block in a portion of the
fetuses whose mothers carry specific autoanti-
bodies (antiRo, antiLa). Maternal phenylketon-
uria is associated with approximately a 25% risk
559
of CHDs, particularly TOF. The risk of adverse
outcome (including microcephaly and mental
handicap) is related to the level of maternal serum
phenylalanine. A phenylalanine-free diet and nor-
mal serum phenylalanine levels before conception
and throughout pregnancy can reduce the risk of
malformations to that of the general population.
Prenatal alcohol exposure increases the risk
of CHD. VSD is the most common heart defect
seen with alcohol exposure, but ASD, TOF, and
a variety of other defects can occur. Major
anomalies are more likely to occur with heavy
alcohol consumption. Other common teratogens
that can increase the risk of CHD include anticon-
vulsants, lithium, and retinoic acid. Prenatal reti-
noic acid exposure, for example, is associated with
a 10% to 20% risk of conotruncal heart defects, in
addition to multiple other malformations.
Syndromic causes of congenital heart disease
Syndromic and chromosomal etiologies are an
important cause of CHDs. About 3% to 5% of
CHDs are associated with recognizable genetic
syndromes [18], and a further significant portion
are associated with other congenital malforma-
tions of unknown etiology. Although most syn-
drome diagnoses are made during infancy and
childhood in a pediatric setting, review in a setting
such as the ACHD clinic may lead to the diagno-
sis of an underlying genetic syndrome in adults
who have subtle presentations or in adults not
previously evaluated by a geneticist. CHDs are
components of many genetic syndromes. Online
Mendelian Inheritance in Man [19] lists over
300 syndromes including CHDs. Diagnosis of
syndromic causes of congenital heart disease is
important for a variety of reasons including ap-
propriate management of and screening for other
medical problems associated with a particular
syndrome and for accurate counseling for recur-
rence risk for other family members and offspring
of the affected individual. Common syndromic
and chromosomal causes of congenital heart dis-
ease that might be encountered or diagnosed in
an ACHD clinic are discussed in the following
sections and summarized in Table 1.
Aneuploidy syndromes
Turner’s syndrome
Turner’s syndrome, a chromosomal disorder
affecting women, has considerable phenotypic
560 BERNIER & SPAETGENS

Table 1
Summary of common syndromes associated with cardiac anomalies
Noncardiac
Syndrome Cardiac manifestations manifestations Etiology Inheritance pattern
Turner’s syndrome (15%–50%) Variable phenotype; Chromosomal (45, X Sporadic; affected
Predominantly short stature, karyotype and women are almost
left-sided lesions gonadal variants always infertile
(ie, BAV, coarctation); dysgenesis, mosaicism,
risk for aortic root lymphedema, structurally
dilatation renal anomalies, abnormal X
normal chromosome)
intelligence but
can have learning
disability
Down syndrome (50%) Dysmorphic facies, Chromosomal Usually sporadic;
Perimembranous VSD, mental handicap; trisomy for associated with
PDA, ASD, and hearing and vision chromosome 21 late maternal age;
AVSD are most loss, early-onset (rarely infertility in
common; 50% require dementia Robertsonian affected adults
surgical intervention translocation or
mosaicism)
22q11 microdeletion (75%) Variable phenotype; Microdeletion at Autosomal
syndrome Conotruncal palate defects, 22q11.2 detectable dominant; 50%
malformations are developmental with FISH (95%); risk to offspring
most common; TOF delay/learning w90% de novo
in 20%; anomalies of difficulties,
major neck vessels are dysmorphic facies,
also common immune
deficiency,
hypocalcemia,
psychiatric illness
in adulthood
Williams syndrome (80%–90%) Dysmorphic facies, Microdeletion at Autosomal
SVAS in 75%; any stellate iris, 7q11.12 involving dominant; 50%
artery may be developmental the elastin gene risk to offspring
narrowed; delay, short detectable with
hypertension in adults stature, connective FISH (99%); most
tissue are de novo
abnormalities,
hypercalcemia
Noonan’s syndrome (50%–90%) Variable phenotype, PTPN11 mutations Autosomal
Pulmonic stenosis short stature, in 50%; half of dominant; 50%
(20%–50%) webbed neck, cases are de novo risk to offspring
Hypertrophic dysmorphic facies,
cardiomyopathy varying
(20%–30%) developmental
delay;
cryptorchidism;
coagulation
defects
VATER association (80%) Vertebral defects, No specific recurring Sporadic; very small
CHD of any type and anal atresia, TEF, etiology risk to offspring
severity renal and limb
anomalies, normal
intelligence
(continued on next page)
 GENETICIST’S ROLE 561

Table 1 (continued )
Noncardiac
Syndrome Cardiac manifestations manifestations Etiology Inheritance pattern
Holt-Oram (75%) Upper limb TBX5 mutation Autosomal
syndrome Ostium secundum ASD; anomalies of (70%); 15% dominant; 50%
VSD; conduction varying degrees familial risk to offspring
defects (especially radial
ray) in 100%
Alagille syndrome (80%–100%) Hepatic disease, JAG1 mutation Autosomal
Pulmonary valve, artery dysmorphic facies, (70%) or deletion dominant; 50%
and branch PA most posterior (w5%); 50%– risk to offspring
commonly affected; embryotoxin, 70% of cases are
TOF in butterfly de novo
7%–15% vertebrae, renal
anomalies,
neurovascular
anomalies
CHARGE (60%–85%) Colobomas, growth CHD7 mutation Autosomal
syndrome Commonly right-sided failure, (60%) or deletion; dominant; 50%
and conotruncal developmental most cases are de risk to offspring
lesions delay, choanal novo but reduced
atresia, genital fertility
hypoplasia, ear
anomalies and
hearing loss
Abbreviations: BAV, bicuspid aortic valve; FISH, fluorescence in situ hybridization; PA, pulmonary artery; PDA,
patent ductus arteriosus; SVAS, supravalvar aortic stenosis; TEF, tracheoesophageal fistula.

variation. Some patients may be diagnosed at birth
with classic physical features of webbed neck,
excess nuchal skin, and marked lymphedema.
Other patients have only subtle features and may
not be diagnosed until adolescence or adulthood
when undergoing investigation for short stature,
amenorrhea, or infertility. Common physical fea-
tures of Turner’s syndrome include short stature,
gonadal dysgenesis, cardiovascular malforma-
tions, lymphedema (webbed neck, edema of the
hands and feet), and structural renal abnormali-
ties. Intelligence is normal but mild speech and
language delays can occur and there tends to be
a specific difficulty with spatial/perceptual skills.
Gonadal dysgenesis results in delayed or absent
secondary sexual characteristics and infertility in
most women who have Turner’s syndrome.
Cardiac anomalies are present in 15% to 50%
of individuals who have Turner’s syndrome. Left-
sided lesions predominate, with bicuspid aortic
valve being the most common defect. Coarctation
of the aorta is seen in approximately 10% of cases.
Hypoplastic left heart, VSD, anomalous pulmo-
nary venous drainage, and conduction defects have
also been seen [20]. Aortic root dilation is seen in
5% to 10% of women who have Turner’s
syndrome and can lead to aortic root dissection
in adult life. Regular echocardiographic surveil-
lance of aortic root diameter and appropriate
screening/treatment of risk factors for dissection,
such as hypertension, are important components
of management of women who have Turner’s syn-
drome during adulthood [21]. Close monitoring by
a cardiologist during pregnancy (invariably
achieved through assisted reproductive technol-
ogy) is imperative because of the risk of rupture
or dissection of the aorta in pregnancy [22].
Turner’s syndrome is classically associated
with a 45,X karyotype but can be seen with
a variety of 45,X variants including 46,XX kar-
yotypes with X chromosome structural abnormal-
ities. Mosaicism (45,X/46,XX or 45,X/46,XY) is
also frequently seen and seems to be associated
with a milder phenotype. The prevalence of
Turner’s syndrome is 1 in 2500 live births,
although prevalence in conceptuses is much higher
because of the high rate of fetal death.
Down syndrome
Down syndrome is characterized by a variety
of congenital malformations, characteristic facial
562 BERNIER & SPAETGENS
features, minor physical differences, and mental
handicap. It usually is easily diagnosed in the
newborn period based on the presence of typical
dysmorphic features that include brachycephaly,
upslanting palpebral fissures, flat facial profile,
single palmar crease, and sandal gap (wide space
between the first and second toe). Cardiac anom-
alies are the most frequent type of birth defect in
individuals who have Down syndrome, but vari-
ous other congenital malformations are also seen.
Duodenal atresia/stenosis and Hirschsprung’s dis-
ease are the most common malformations after
heart defects. Developmental delay is universal,
and all affected individuals have some degree of
mental handicap. Adults who have Down syn-
drome generally require supervised home and
work environments. Median life expectancy is
about 50 years. An important cause of this
reduced life expectancy, apart from the contribu-
tion of increased mortality secondary to CHDs in
infancy and childhood, is a very high rate of
dementia with cognitive and neuropathologic
findings identical to those seen in Alzheimer’s
disease.
Congenital heart disease is seen in about
50% of individuals who have Down syndrome.
Half of those will have a serious CHD that
requires surgical intervention. The most common
heart defects are a perimembranous VSD, fol-
lowed by patent ductus arteriosus and ASD.
AVSD is the classic CHD seen in Down syn-
drome, occurring 100 times more frequently in
Down syndrome than in the general population.
Other types of complex heart disease are also
seen. The presence, type, and severity of heart
defects are important determinants of survival in
individuals who have Down syndrome. Screening
with echocardiography is part of the routine
investigation of any infant/child who has a con-
firmed or a suspected diagnosis of Down
syndrome.
Down syndrome is caused by the presence of
an extra copy of chromosome 21, specifically
region 21q22. Most cases of Down syndrome are
the result of standard trisomy 21. Five percent of
cases are due to Robertsonian translocations
involving chromosome 21 (which may or may
not be familial) or to trisomy 21 mosaicism.
Increased maternal age is the single most impor-
tant risk factor for trisomy 21. Men who have
Down syndrome are infertile. Despite normal
sexual development in girls, there are only a small
number of cases of documented fertility in
women. Risk of recurrence in offspring is
therefore generally immaterial, although it would
be increased over that of the general population.
Microdeletion syndromes
22q11 microdeletion syndrome
The 22q11 microdeletion syndrome is the most
commonly occurring microdeletion syndrome in
humans [23]. It goes by many names (partly de-
pendent on which features are present in an indi-
vidual): velocardiofacial syndrome, Shprintzen’s
syndrome, DiGeorge syndrome or sequence, and
conotruncal anomaly face syndrome. This condi-
tion has a highly variable range of expression.
Over 180 different clinical features have been re-
ported in individuals who have 22q11 microdele-
tion. Affected individuals may have many
features of the syndrome or only one or two fea-
tures with subtle facial dysmorphism, making di-
agnosis difficult. The frequency in the general
population is thought to about 1 in 4000, but
this figure likely represents an underestimate given
the under-recognition of mild cases. The most
common findings in the 22q11 microdeletion syn-
drome are CHDs; palate abnormalities including
velopharyngeal insufficiency, submucosal cleft pal-
ate, and frank cleft palate; characteristic facial fea-
tures; and learning difficulties or developmental
delay [24]. Other manifestations are immune defi-
ciency (thymic hypoplasia), hypocalcemia (para-
thyroid hypoplasia), feeding difficulties, renal
anomalies, and hearing loss. Individuals who
have the 22q11 deletion are at increased risk for
psychiatric illness in adulthood. Typical facial fea-
tures are subtle but appear in most affected indi-
viduals. These features include a long face and
tubular shaped nose. Micrognathia, mild hyperte-
lorism, and hooded eyelids may also be seen.
CHDs occur in about 75% of individuals who
have 22q11 microdeletion syndrome. The cono-
truncal malformations are the most common
types of heart defects. TOF occurs in approxi-
mately 20% of individuals. Other conotruncal
lesions include interrupted aortic arch, VSD, and
truncus arteriosus [23]. In many centers, infants
who have TOF, interrupted aortic arch, and trun-
cus arteriosus are routinely screened for the 22q11
microdeletion. Nonconotruncal lesions can also
occur in this condition. Bicuspid aortic valve,
ASD, TGA, and anomalous origin of the coro-
nary artery have been reported. Anomalies of
the internal carotid arteries and other major
neck vessels are also a common finding.
 GENETICIST’S ROLE
The 22q11 microdeletion results from a deletion
of about 3 megabases (Mb) on the distal end of
the long arm of chromosome 22. A small propor-
tion of affected individuals have smaller 1.5-Mb
deletions. A deletion is visible on a routine high-
resolution karyotype in about 15% of cases. The
rest of the cases are diagnosable only with fluores-
cence in situ hybridization (FISH) using a probe
specific for this region. FISH studies detect dele-
tions in over 95% of affected individuals. The de-
letion results in haploinsuffiency for over 30 genes
in a typical 3-Mb deletion. Of these genes, TBX1
is believed to be responsible for most heart and
vascular anomalies [25,26]. Ninety-three percent
of individuals who have the 22q11 microdeletion
have de novo deletions. The risk of 22q11 dele-
tions in offspring of affected individuals is 50%.
The severity of the condition in offspring or sib-
lings is variable and cannot be predicted based
on the phenotype of the proband. It is not uncom-
mon to diagnose a mildly affected parent after the
birth of a more severely affected infant.
Williams syndrome
Williams syndrome is characterized by dys-
morphic facial features, developmental delay,
short stature, and CHDs. A range of connective
tissue abnormalities including hoarse voice, in-
guinal/umbilical hernia, bowel/bladder diverticu-
lae, rectal prolapse, joint limitation or laxity, and
soft, lax skin are frequently observed. Hypercal-
cemia or hypercalciuria are common. The range
of intellectual impairment is wide, ranging from
low average intelligence to severe mental retarda-
tion. The cognitive profile is unique, with partic-
ular strengths in verbal language and memory
skills in contrast to very poor visuospatial abili-
ties. The behavioral phenotype is characterized by
anxiety, attention problems, and overfriendliness
to strangers. The term ‘‘cocktail party’’ personal-
ity is frequently used to describe these patients.
Distinctive facial characteristics include bitempo-
ral narrowing, periorbital fullness, short nose, full
nasal tip, midface hypoplasia, long philtrum, full
lips, wide mouth, and prominent earlobes. Adults
typically have a long face and neck resulting in
a more gaunt appearance. Also present is a stel-
late/lacy iris pattern.
The cardiovascular features of this condition
are secondary to elastin arteriopathy. Cardiac
abnormalities are seen in 80% to 90% of individ-
uals. Supravalvar aortic stenosis (SVAS) is the
most clinically significant and most common
563
cardiovascular finding, occurring in about 75% of
affected individuals. SVAS can take the form of
a discrete narrowing above the valve (hourglass
stenosis) or a more diffuse long-segment aortic
hypoplasia. These lesions tend to be progressive.
Thirty percent to 40% of individuals who have
SVAS eventually require surgical correction. Al-
though the aorta is most frequently involved, any
artery may be narrowed. Supravalvar pulmonic
stenosis, peripheral pulmonic stenosis, renal artery
stenosis, and coronary artery involvement can be
seen. Unlike the aortic lesions, however, the
pulmonary lesions tend to regress with time.
Hypertension is common in adolescents and adults.
Mitral valve prolapse and aortic insufficiency have
also been reported in adults. In general, cardiovas-
cular disease tends to be more severe in men.
Williams syndrome is caused by a submicro-
scopic deletion of 7q11.23. The commonly deleted
region spans 1.5 Mb, and 21 genes have so far
been mapped within this region. Many of the
manifestations (arteriopathy, hoarse voice, her-
nias, and connective tissue abnormalities) of
Williams syndrome are due to deletion (and
therefore haploinsufficiency) of the elastin gene.
FISH studies demonstrate deletion of the elastin
gene in 99% of individuals who have Williams
syndrome. Families who have isolated familial
SVAS typically have point mutations within the
elastin gene and do not demonstrate deletions
with FISH. Most cases of Williams syndrome are
de novo. An affected individual has a 50% chance
of passing the condition to his/her offspring. The
prevalence of Williams syndrome is about 1 in
10,000 live births.
Single-gene disorders
Noonan syndrome
Noonan syndrome is a relatively common
syndrome characterized by short stature, CHDs,
broad or webbed neck, pectus deformities, vary-
ing degrees of developmental delay, cryptorchi-
dism in male patients, and dysmorphic facial
features. Facial appearance changes over time
and is most subtle in the adult. Key features
irrespective of age include low-set posteriorly
rotated ears with fleshy helices, vivid blue or
blue-green irides, widely spaced eyes with epican-
thal folds, and thick or droopy eyelids. Other
features of this syndrome include a variety of
ocular abnormalities, renal anomalies (generally
mild), and scoliosis. Various coagulation defects
564 BERNIER & SPAETGENS
may occur in Noonan syndrome and affect
approximately one third of individuals. The coa-
gulopathy may range from severe surgical hemor-
rhage to clinically mild bruising or to laboratory
anomalies with no clinical sequelae. Routine
screening for coagulopathy at the time of di-
agnosis should include prothrombin time, partial
thromboplastin time, and platelet count. Acetylsa-
licylic acid should be avoided in those who have
laboratory or clinical evidence of coagulopathy,
and physicians should be aware of the potential
for prolonged bleeding in those undergoing surgi-
cal procedures. Men who have Noonan syndrome
are frequently infertile secondary to cryptorchi-
dism. The prevalence of Noonan syndrome is
about 1 in 2500, and mild cases are often
unrecognized.
The frequency of congenital heart disease in
individuals who have Noonan syndrome is 50%
to 90%. The frequency of Noonan syndrome
among children who have congenital heart disease
is approximately 1.4%. A stenotic, often dysplas-
tic, pulmonary valve is the most common heart
defect in Noonan syndrome and is seen in 20% to
50% of individuals. Seven percent of those who
have valvular pulmonary stenosis have an under-
lying diagnosis of Noonan syndrome. Dysplastic
valves, with reduced mobility rather than fusion
of valve commissures, are rare in non-Noonan in-
dividuals. Hypertrophic cardiomyopathy is also
a relatively frequent cardiovascular finding in
Noonan syndrome, occurring in about 20% to
30% of affected individuals [27]. It can range
from mild localized ventricular septal hypertrophy
to severe hypertrophic cardiomyopathy associated
with significant mortality. The cardiomyopathy
can present in utero, during infancy, or during
childhood. Clinical and echocardiographic fea-
tures are identical to nonsyndromic hypertrophic
cardiomyopathy, although it appears that ar-
rhythmia and sudden death are less common in
the hypertrophic cardiomyopathy associated
with Noonan syndrome. Other cardiac defects fre-
quently seen in this condition include ASD, VSD,
supravalvar pulmonic stenosis, branch pulmonary
artery stenosis, TOF, coarctation of the aorta, and
even partial AVSDs.
Noonan syndrome is an autosomal dominant
condition demonstrating marked variability of
expression. About half of the cases arise de
novo. Half of the cases are familial, and it is not
uncommon to diagnose Noonan syndrome in
a mildly affected parent following the diagnosis
of an affected infant or child. Missense mutations
in the PTPN11 gene have been identified in about
50% of individuals who have Noonan syndrome
[28]. Sequence analysis of PTPN11 is available
on a clinical basis.
Holt-Oram syndrome
Holt-Oram syndrome is the best recognized of
the heart-hand syndromes. It is characterized by
cardiac defects and upper-limb anomalies of vari-
able severity. Skeletal defects involve the upper
limbs exclusively and are usually bilateral and
asymmetric. They typically involve the radial ray.
Defects range from mild involvement with clino-
dactyly (curved fingers), limited supination, or
narrow sloping shoulders to more severe defects
with absent, hypoplastic, or triphalangeal thumb.
At the most severe end of the spectrum are
reduction deformities that may include any part
of the upper limb. The left side is typically more
severely affected than the right side. At a mini-
mum, all affected individuals have an abnormal
carpal bone that may be the only evidence of
disease. Posterior hand radiographs can identify
such defects. The broad range of severity of these
findings is such that some individuals who have
the mildest upper-limb malformations and no or
only mild CHDs may escape diagnosis. In general,
the manifestations of Holt-Oram syndrome tend
to be more severe in women.
The CHDs most commonly observed in Holt-
Oram syndrome are ostium secundum ASD and
VSD, especially those occurring in the muscular
trabeculated septum. More severe cardiac defects
such as hypoplastic left heart and total anomalous
pulmonary veins have been reported. CHDs are
seen in about 75% of affected individuals, and
a further proportion of affected individuals may
have nonspecific ECG abnormalities. Individuals
who have Holt-Oram syndrome, with or without
a congenital heart malformation, are at risk for
cardiac conduction defects including complete
heart block and atrial fibrillation. There appears
to be some correlation between the severity of the
upper-limb defect and the severity of the cardiac
lesion.
Holt-Oram syndrome is an autosomal domi-
nant disorder caused by heterozygous mutations
in the transcription factor TBX5 [29,30]. Over
70% of individuals who meet strict diagnostic cri-
teria (ie, upper-limb defect plus personal or family
history of structural or conductive heart disease)
have a TBX5 mutation [31]. Clinical testing is
available for TBX5 mutations. The estimated
 GENETICIST’S ROLE
frequency of Holt-Oram syndrome is 1 in 100,000
births. About 85% of cases are the result of new
mutations. There is significant variable expressiv-
ity observed in individuals who have Holt-Oram
syndrome, within and among families who have
the same mutation. The phenotype of affected off-
spring, therefore, cannot be accurately predicted.
In general, of those who have inherited a TBX5
mutation, one third have a severe reduction defect
of the upper limb (ie, absent thumb), and 5% have
phocomelia.
Alagille syndrome
Alagille syndrome is a highly variable, multi-
system condition involving the liver, heart, eyes,
face, and skeleton. Bile duct paucity has tradi-
tionally been required for diagnosis, and hepatic
presentation with conjugated hyperbilirubinemia
or cholestasis, often in the neonatal period, is
typical. Progression to cirrhosis or liver failure
necessitating transplant occurs in about 20% of
affected individuals. The classic eye finding is
posterior embryotoxin, an eye abnormality that
results in a white ring at the periphery of the
cornea and is a benign anterior chamber finding.
Butterfly vertebrae (clefting of the vertebral body)
are seen in about 50% of individuals but are
rarely symptomatic. Less frequently encountered
problems include pancreatic insufficiency and
a variety of structural renal abnormalities. Neuro-
vascular anomalies predisposing to intracranial
bleeds are seen in about 15% of individuals. The
overall mortality rate is about 10% and attribut-
able to cardiac and liver disease and vascular
accidents. Characteristic facial features are seen in
most individuals, although they tend to be more
subtle in adulthood. Features include prominent
forehead, pointed chin, deep-set eyes, hypertelor-
ism, and bulbous nose.
Cardiac involvement occurs in 80% to 100%
of individuals who have Alagille syndrome. Most
heart defects are hemodynamically insignificant.
The pulmonary vasculature (including valve, ar-
tery, and branches) is most frequently involved.
The most common defect is peripheral pulmonary
stenosis, which tends not to be progressive. Other
heart defects include TOF, VSD, ASD, aortic
stenosis, and aortic coarctation. The clinical out-
comes of those who have TOF and Alagille
syndrome may be poor compared with those
who have typical isolated TOF [32]. TOF occurs
in 7% to 16% of individuals who have Alagille
syndrome.
565
Alagille syndrome is an autosomal dominant
condition. It is caused by mutations in the JAG1
gene [30,33]. JAG1 is a cell surface protein that
functions in the notch signaling pathway. Muta-
tions in JAG1 are identified in about 70% of clin-
ically affected individuals [34]. Deletions of the
JAG1 gene, identifiable with FISH studies, are
found in a further 5% to 7% of individuals who
have Alagille syndrome. Larger deletions are asso-
ciated with a more severe phenotype and addi-
tional clinical features such as developmental
delay or hearing loss. Clinical testing by sequence
analysis for point mutations and FISH analysis
for deletions is available. Fifty percent to 70%
of JAG1 mutations are de novo. The molecular
characterization of Alagille syndrome has shown
that affected individuals may have only very sub-
tle or isolated findings. Such patients are often di-
agnosed only following the diagnosis of a more
classically affected offspring. Risk to offspring of
an affected parent is 50%. No genotype/pheno-
type correlations have been found, so the severity
of clinical features cannot be predicted based on
genotype or the degree to which a parent is af-
fected. The prevalence of Alagille syndrome is 1
in 70,000 to 1 in 100,000, although this range
may be an underestimate.
CHARGE syndrome
CHARGE syndrome is a multiple congenital
anomaly syndrome characterized by colobomas of
the iris, retina, or optic disc; CHD; postnatal
growth failure; developmental delay; choanal
atresia; genital hypoplasia (particularly in men);
and ear abnormalities that can include external
and internal ear anomalies and hearing loss. Other
relatively common features are facial palsy, renal
anomalies, orofacial clefts, and tracheoesophageal
fistula.
CHDs are seen in about 60% to 85% of
individuals who have CHARGE syndrome. The
severity and the spectrum of heart defects vary,
although there appears to be a preponderance of
right-sided lesions and conotruncal defects. About
75% of heart defects in patients who have
CHARGE syndrome require surgical intervention.
CHARGE syndrome was recently discovered
to be caused by mutations in the CHD7 gene,
which codes for a chromodomain helicase DNA
binding protein [35]. Mutations in CHD7 are
found in about 60% of patients meeting diagnos-
tic criteria for CHARGE syndrome. Whole gene
deletions of CHD7 are also reported. Clinical
566 BERNIER & SPAETGENS
testing by sequence analysis for point mutations
and FISH analysis for deletions are available.
Most mutations and deletions are de novo.
Many patients who have CHARGE syndrome,
particularly men, have delayed or absent puberty
and infertility. Of those who reproduce, the risk
to offspring is 50%. The prevalence of CHARGE
syndrome is estimated to be about 1 in 9000 to 1
in 12,000.
Associations
VATER association
VATER association is a sporadic constellation
of congenital malformations that occur together
more often than expected by chance alone. It is
a diagnosis of exclusion. VATER is characterized
by a combination of vertebral defects, anal
atresia, CHDs, tracheoesophageal fistula, renal
anomalies, and limb defects. Rarely do individuals
have all of the above anomalies. Development and
intelligence are normal. Dysmorphic facies are not
present. The prognosis of the condition depends
on which features are present, their severity, and
the degree to which they can be treated.
CHDs are the most common finding in the
VATER association. Approximately 80% of
affected individuals have cardiac involvement.
Heart defects can be of any type and severity.
No specific recurring etiology has been estab-
lished for VATER. The prevalence is 1.6 in
10,000. There does not appear to be a large risk
for recurrence in siblings or offspring of individ-
uals who have VATER association.
Genetics and adult patients who have congenital
heart disease
In addition to providing cardiac care, ACHD
clinics can play an important role in identifying
patients who might benefit from additional genetic
assessment and counseling. This counseling may be
particularly relevant for patients of reproductive
age. Determining whether patients are interested in
obtaining information on recurrence risks should
be explored with all ACHD patients [36].
Additional patients who might benefit from
further assessment include those who have non–
CHD-related medical problems, congenital anom-
alies, or learning/cognitive deficits. It may be more
difficult for some clinicians to consider whether
the patient has a more subtle pattern of physical
features that might suggest the presence of an
underlying syndrome. For example, the hyper-
telorism (wide spacing of the eyes) seen in Noonan
syndrome can vary from subtle to striking.
Screening for patients who have a family history
of CHDs should also be undertaken by asking
whether any first-degree (parents, siblings, and
children) or second-degree (grandparents, aunts/
uncles, and nieces/nephews) relatives have a his-
tory of CHDs [2]. The presence of a single affected
third-degree relative (eg, first cousin) is unlikely to
be significant. A family history of major malfor-
mations, unexplained infant deaths, or significant
learning/cognitive disabilities could be an indica-
tion that an underlying syndrome is segregating
in the family (Box 1).
The role of 22q11 testing merits specific
attention. Beauchesne and colleagues [37] identi-
fied six patients who had 22q11 deletions in a co-
hort of 103 consecutive adults who had high-risk
CHD lesions (prevalence, 5.8%). In particular,
they noted that patients often had few if any fea-
tures that might have otherwise identified them as
being at higher risk for 22q11 deletions. They ad-
vocated 22q11 testing in all patients who have
high-risk lesions, which is a view supported by
other investigators [2,37]. Patients who have
22q11 deletions should be screened for associated
Box 1. Checklist for identifying adult
congenital heart disease patients who
might benefit from genetic assessment
and counseling
, Potential parents
, Unexplained medical conditions (eg,
short stature, cholestasis, immune
deficiency, psychiatric illness,
hearing loss)
, Noncardiac malformations
, Facial dysmorphism
, Learning or cognitive disabilities
, 22q11 deletion status unknown in
a patient who has a high-risk CHD
(eg, outflow tract anomaly,
pulmonary atresia/stenosis, tricuspic
atresia, transposition)
, Family history of
, CHDs
, Major malformations
, Unexplained infant deaths or
, Significant learning/cognitive
disabilities
 GENETICIST’S ROLE
medical problems in a genetics clinic or a multidis-
ciplinary 22q11deletion clinic (if one is available
locally) and offered further genetic counseling.
Recurrence risks
Patients identified as being of reproductive age
but for whom the remaining screening questions
are normal likely have isolated CHDs and should
be counseled regarding their specific recurrence
risk, recognizing that limited and at-times con-
flicting data are available on recurrence risks
(Table 2) [38,39]. Recurrence risks for offspring
are higher than sibling recurrence risks, and for
many CHDs, women appear to have a higher re-
currence risk than men. The reason for the latter
finding remains unknown. For most patients
who have an isolated nonfamilial CHD, quoting
a recurrence risk of 5% to 6% for women and
2% to 3% for men is appropriate until further
data on CHD-specific recurrence risks become
available. The recurrence risk for syndromic pa-
tients who have autosomal dominant conditions
is 50%; for recessive conditions, the risk will de-
pend on the carrier frequency in the population
but will usually be quite low (see Table 1).
Prenatal testing and screening
Women at increased risk of having a child who
has a CHD are offered fetal echocardiography at
18 to 20 weeks’ gestation [40]. Fetal nuchal trans-
lucency measurement between 11 and 14 weeks’
gestation is a good screening test for CHD
because the prevalence of CHD increases
Table 2
Risk of recurrence in offspring of parents who have congenital heart defects
Risk to offspring (%)
1994 Meta-analysis [38]
CHD Of women
Atrioventicular septal defect 11.6
Aortic stenosis 8.0
Coarctation of the aorta 6.3
Pulmonary stenosis 5.3
Atrial septal defect 6.1
Ventricular septal defect 6.0
Persistent ductus arteriosus 4.1
Tetralogy of Fallot 2.0
Transposition great arteries
Anomalous pulmonary venous drainage
Abnormal situs and other abnormal connections
All CHDs 5.8
567
exponentially with fetal nuchal translucency thick-
ness [41]. For patients at particularly high risk,
early fetal echocardiography (11–13 weeks) can
be considered if the expertise is available. Invasive
prenatal, molecular, or cytogenetic testing follow-
ing amniocentesis or chorionic villus sampling is
available to patients who have identified single-
gene or cytogenetic etiologies for their CHD.
The offering of these later investigations should
be accompanied by genetic counseling.
Models of genetic service in adult congenital heart
disease clinics
No literature could be identified by the authors
on the subject of models of genetic service pro-
vision to patients who have ACHD. Most ACHD
clinics refer to local genetic clinics, but some
clinics have developed multidisciplinary care
models. This latter model would likely promote
more integrated care. In a recent survey of
cardiologists and clinical geneticists regarding
the provision of genetic services to hypertrophic
cardiomyopathy patients, the majority of both
groups indicated they preferred a model in which
each group was responsible for specific steps of
a cardiogenetic assessment [42]. With respect to
ACHD patients, cardiologists may well be willing
to undertake screening of appropriate patients to
refer to genetics while leaving the further assess-
ment and counseling of identified patients to
genetics. ACHD clinics should nevertheless
consider the inclusion of genetic care providers
in the development of local care pathways and
work toward multidisciplinary models of care.
1998 British collaborative study [39]
Of men Of women Of men
4.3 7.9 7.7
3.8
3.0
3.5
3.5
3.6
2.0
1.4 4.5 1.6
5.9
7.1 15.4
3.1 5.7 2.2
568 BERNIER & SPAETGENS
Summary
With improved treatment of CHDs, patients
now frequently survive to adulthood. CHDs are
etiologically heterogeneous; however, there is
significant evidence that genes play an important
role in the etiology of most CHDs. Most ACHD
patients have a recurrence risk of between 2% and
50%. Although the realization that they face
a recurrence risk may not alter the reproductive
plans of many patients, some may appreciate
frank discussions regarding their risks and their
reproductive options. Medical geneticists can play
in important role in the identification, assessment,
and counseling of at-risk ACHD patients and
should be included as part of the multidisciplinary
team caring for this patient population.
References
[1] Pelech AN, Broeckel U. Toward the etiologies of
congenital heart diseases. Clin Perinatol 2005;
32(4):825–44 [vii].
[2] Goldmuntz E. The genetic contribution to congeni-
tal heart disease. Pediatr Clin North Am 2004;51(6):
1721–37 [x].
[3] Grossfeld PD. The genetics of congenital heart dis-
ease. J Nucl Cardiol 2003;10(1):71–6.
[4] Garg V, Kathiriya IS, Barnes R, et al. GATA4 mu-
tations cause human congenital heart defects and
reveal an interaction with TBX5. Nature 2003;
424(6947):443–7.
[5] Hirayama-Yamada K, Kamisago M, Akimoto K,
et al. Phenotypes with GATA4 or NKX2.5 muta-
tions in familial atrial septal defect. Am J Med Genet
A 2005;135(1):47–52.
[6] Okubo A, Miyoshi O, Baba K, et al. A novel
GATA4 mutation completely segregated with atrial
septal defect in a large Japanese family. J Med Genet
2004;41(7):e97.
[7] Sarkozy A, Conti E, Neri C, et al. Spectrum of atrial
septal defects associated with mutations of NKX2.5
and GATA4 transcription factors. J Med Genet
2005;42(2):e16.
[8] Nemer G, Fadlalah F, Usta J, et al. A novel muta-
tion in the GATA4 gene in patients with tetralogy
of Fallot. Hum Mutat 2006;27(3):293–4.
[9] Ueyama T, Kasahara H, Ishiwata T, et al. Myocar-
din expression is regulated by Nkx2.5, and its func-
tion is required for cardiomyogenesis. Mol Cell
Biol 2003;23(24):9222–32.
[10] Schott JJ, Benson DW, Basson CT, et al. Congenital
heart disease caused by mutations in the transcription
factor NKX2–5. Science 1998;281(5373):108–11.
[11] McElhinney DB, Geiger E, Blinder J, et al. NKX2.5
mutations in patients with congenital heart disease.
J Am Coll Cardiol 2003;42(9):1650–5.
[12] Hobbs CA, Cleves MA, Keith C, et al. NKX2.5 and
congenital heart defects: a population-based study.
Am J Med Genet A 2005;134(2):223–5.
[13] Robinson SW, Morris CD, Goldmuntz E, et al. Mis-
sense mutations in CRELD1 are associated with car-
diac atrioventricular septal defects. Am J Hum
Genet 2003;72(4):1047–52.
[14] Gebbia M, Ferrero GB, Pilia G, et al. X-linked situs
abnormalities result from mutations in ZIC3. Nat
Genet 1997;17(3):305–8.
[15] Ware SM, Peng J, Zhu L, et al. Identification and
functional analysis of ZIC3 mutations in heterotaxy
and related congenital heart defects. Am J Hum
Genet 2004;74(1):93–105.
[16] Krantz ID, Smith R, Colliton RP, et al. Jagged1 mu-
tations in patients ascertained with isolated congen-
ital heart defects. Am J Med Genet 1999;84(1):
56–60.
[17] Tassabehji M, Metcalfe K, Donnai D, et al. Elastin:
genomic structure and point mutations in patients
with supravalvular aortic stenosis. Hum Mol Genet
1997;6(7):1029–36.
[18] Meberg A, Otterstad JE, Froland G, et al. Outcome
of congenital heart defects–a population-based
study. Acta Paediatr 2000;89(11):1344–51.
[19] Online Mendelian Inheritance in Man. OMIM.
McKusick-Nathans Institute for Genetic Medicine,
Johns Hopkins University (Baltimore, MD) and Na-
tional Center for Biotechnology Information, Na-
tional Library of Medicine (Bethesda, MD), 2000.
Available at: http://www.ncbi.nlm.nih.gov/omim/.
Accessed March 15, 2006.
[20] Mazzanti L, Cacciari E. Congenital heart disease in
patients with Turner’s syndrome. Italian Study
Group for Turner Syndrome (ISGTS). J Pediatr
1998;133(5):688–92.
[21] Elsheikh M, Casadei B, Conway GS, et al. Hyper-
tension is a major risk factor for aortic root dilata-
tion in women with Turner’s syndrome. Clin
Endocrinol (Oxf) 2001;54(1):69–73.
[22] Karnis MF, Zimon AE, Lalwani SI, et al. Risk of
death in pregnancy achieved through oocyte dona-
tion in patients with Turner syndrome: a national
survey. Fertil Steril 2003;80(3):498–501.
[23] Perez E, Sullivan KE. Chromosome 22q11.2 dele-
tion syndrome (DiGeorge and velocardiofacial
syndromes). Curr Opin Pediatr 2002;14(6):
678–83.
[24] Antshel KM, Kates WR, Roizen N, et al. 22q11.2
deletion syndrome: genetics, neuroanatomy and
cognitive/behavioral features keywords. Child Neu-
ropsychol 2005;11(1):5–19.
[25] Merscher S, Funke B, Epstein JA, et al. TBX1 is re-
sponsible for cardiovascular defects in velo-cardio-
facial/DiGeorge syndrome. Cell 2001;104(4):
619–29.
[26] Yagi H, Furutani Y, Hamada H, et al. Role of TBX1
in human del22q11.2 syndrome. Lancet 2003;
362(9393):1366–73.
 GENETICIST’S ROLE
[27] Nishikawa T, Ishiyama S, Shimojo T, et al. Hyper-
trophic cardiomyopathy in Noonan syndrome.
Acta Paediatr Jpn 1996;38(1):91–8.
[28] Tartaglia M, Mehler EL, Goldberg R, et al. Muta-
tions in PTPN11, encoding the protein tyrosine
phosphatase SHP-2, cause Noonan syndrome. Nat
Genet 2001;29(4):465–8.
[29] Basson CT, Bachinsky DR, Lin RC, et al. Mutations
in human TBX5 [corrected] cause limb and cardiac
malformation in Holt-Oram syndrome. Nat Genet
1997;15(1):30–5.
[30] Li L, Krantz ID, Deng Y, et al. Alagille syndrome is
caused by mutations in human Jagged1, which en-
codes a ligand for Notch1. Nat Genet 1997;16(3):
243–51.
[31] McDermott DA, Bressan MC, He J, et al. TBX5
genetic testing validates strict clinical criteria for
Holt-Oram syndrome. Pediatr Res 2005;58(5):
981–6.
[32] McElhinney DB, Krantz ID, Bason L, et al. Analysis
of cardiovascular phenotype and genotype-pheno-
type correlation in individuals with a JAG1 muta-
tion and/or Alagille syndrome. Circulation 2002;
106(20):2567–74.
[33] Oda T, Elkahloun AG, Pike BL, et al. Mutations in
the human Jagged1 gene are responsible for Alagille
syndrome. Nat Genet 1997;16(3):235–42.
[34] Krantz ID, Colliton RP, Genin A, et al. Spectrum
and frequency of jagged1 (JAG1) mutations in Ala-
gille syndrome patients and their families. Am J
Hum Genet 1998;62(6):1361–9.
569
[35] Vissers LE, van Ravenswaaij CM, Admiraal R, et al.
Mutations in a new member of the chromodomain
gene family cause CHARGE syndrome. Nat Genet
2004;36(9):955–7.
[36] Uebing A, Steer PJ, Yentis SM, et al. Pregnancy and
congenital heart disease. BMJ 2006;332(7538):
401–6.
[37] Beauchesne LM, Warnes CA, Connolly HM, et al.
Prevalence and clinical manifestations of 22q11.2
microdeletion in adults with selected conotruncal
anomalies. J Am Coll Cardiol 2005;45(4):595–8.
[38] Nora JJ. From generational studies to a multilevel
genetic-environmental interaction. J Am Coll
Cardiol 1994;23(6):1468–71.
[39] Burn J, Brennan P, Little J, et al. Recurrence risks in
offspring of adults with major heart defects: results
from first cohort of British collaborative study.
Lancet 1998;351(9099):311–6.
[40] Cohen MS, Frommelt MA. Does fetal diagnosis
make a difference? Clin Perinatol 2005;32(4):
877–90 [viii].
[41] Atzei A, Gajewska K, Huggon IC, et al. Relation-
ship between nuchal translucency thickness and
prevalence of major cardiac defects in fetuses with
normal karyotype. Ultrasound Obstet Gynecol
2005;26(2):154–7.
[42] van Langen IM, Birnie E, Schuurman E, et al. Pref-
erences of cardiologists and clinical geneticists for
the future organization of genetic care in hypertro-
phic cardiomyopathy: a survey. Clin Genet 2005;
68(4):360–8.
 Cardiol Clin 24 (2006) 571–585
The Anesthesiologist’s Role in Adults
with Congenital Heart Disease
Jane Heggie, MD, FRCP, Jacek Karski, MD, FRCP*
Department of Anesthesia, Toronto General Hospital, 3 Eaton North, 200 Elizabeth Street,
Toronto, Ontario M5G 2C4, Canada
Adults with congenital heart disease (CHD)
are a considerable challenge to anesthesiologists
because of the management they require not only
during cardiac surgery but also during noncardiac
surgery, acute (postoperative) pain, labor and
delivery, and while undergoing conscious sedation
for diagnostic and interventional procedures. The
adult patient population with CHD includes those
with uncorrected, corrected, and palliated con-
genital heart defects. Patients presenting for repair
or revision of a complex congenital heart defect
will likely be under the care of an existing CHD
team in a tertiary care center [1,2]. The increasing
number of adults with CHD mandates the crea-
tion of more specialized centers to manage the
long term and emergency care of these patients
[3–5]. The anesthesiologist is a crucial member
of the CHD team. This article addresses the issues
that must be considered in establishing anesthetic
and perioperative services in a center that pro-
vides specialized care for adults with CHD.
Adult patients with CHD have unique adap-
tive physiologic considerations. They often have
a co-existing disease or anatomy with a surgical
intervention planned that is in conflict with these
considerations, creating management hurdles for
the perioperative team. The challenge for the
anesthesiologist is to understand the underlying
pathophysiology of each CHD patient and to
anticipate areas of conflict by tailoring the intra-
operative management to minimize the adverse
effects the co-existing disease or the operative
procedure on the patient’s hemodynamics. Several
recent reviews are directed toward anesthesia in
E-mail address: jacek.karski@uhn.on.ca (J. Karski).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.006
this challenging group of patients for cardiac and
noncardiac surgery [6–10], as well as reviews in the
internal medicine and cardiology literature pro-
viding overviews of CHD in adults [1,11–14].
Although much effort has been made to
transition the teenager successfully to an adult
CHD center, a substantial number of patients are
‘‘lost to follow-up’’ [15]. In addition to managing
the patient’s perioperative care, the anesthesiolo-
gist may have to reunite the patient with a CHD
cardiologist. Patients will often minimize the ex-
tent of their cardiac disease and focus on the sim-
plest explanation of their problem [16]. In a survey
of 1941 adult patients who had CHD and who
had been lost to follow-up (O5 years with no car-
diology contact), Wacker [17] found that 96%
regarded themselves as healthy and fit, and 68%
had no regular medical follow-up.
Preoperative assessment
The first encounter with the CHD patient is
either at the preadmission outpatient consultation
or as an inpatient before a surgical procedure or
therapeutic intervention. The urgency of the pro-
posed procedure will dictate how much detailed
information can be obtained. In emergency situ-
ations, such as an acute abdomen or cardiover-
sion, the history will be obtained while preparing
to act, and it is imperative that consultant anes-
thesiologists have knowledge of and experience
with CHD or access to colleagues who do either
within their own department or by telephone.
An anesthesia consultation includes a discus-
sion of previous surgery and intensive care unit
(ICU) admissions. A source of tremendous anxi-
ety for some adult CHD patients is the memory of
cardiology.theclinics.com
572 HEGGIE & KARSKI
previous episodes of awareness in the operating
room or the ICU [10,18,19]. A review of systems
includes questions about asthma, smoking, recre-
ational drugs, alcohol use, renal function, the gas-
trointestinal tract, endocrine issues, and the
presence of and type of a pacemaker, as well as
when it was last checked. In all women of child-
bearing age, the date of the last normal menstrual
cycle and the possibility of pregnancy must be
considered. In elective cases, extremes of body
weight have warranted a delay of surgery for nu-
tritional supplementation or weight loss. Previous
anesthetic records and ICU admissions often con-
tain information vital to the anesthetic plan but
may not be included in cardiology summaries or
dictated surgical notes and may take a few days to
procure. The patient’s parents may provide much
of the history and may need as much reassurance
as the patient [16]. It is important to establish
good rapport at this time because the anesthesiolo-
gist and his or her ICU colleagues will have an
intense exposure to the family in the days to come.
It is helpful to have a clear picture of what the
patient’s innate anatomy was. With this informa-
tion, the anesthesiologist can follow how surgical
intervention has altered the patient’s hemodynam-
ics. The blood pressure in both arms should be
assessed and the arterial and venous access sites
inspected in anticipation of monitoring and re-
suscitation. Airway assessment and management
are critical to the plan for induction and for
extubation of the patient. The presence of
systemic to pulmonary artery shunts (Blalock-
Taussig [BT], Potts, Waterston-Cooley) or a cavo-
pulmonary anastomosis (Glenn or Fontan) is
important to note in planning the ventilation
and positioning. In a classic BT shunt, the subcla-
vian artery is divided and anastomosed end-
to-side to the ipsilateral pulmonary artery. The
ipsilateral radial pulse is lost or diminished.
Some patients may have had bilateral BT shunts
or a BT shunt on one side and arterial cutdowns
on the other side. In these patients, monitoring
of the central arterial pressure will need to be
done from a femoral or high brachial artery.
More recent BT shunts may have used a Gore-
Tex graft from the intact subclavian to the ipsilat-
eral pulmonary artery. The ipsilateral radial pulse
is intact. A Waterston-Cooley shunt is an anasto-
mosis between the posterior aspect of the aorta
and the anterior aspect of the right pulmonary ar-
tery. A Potts shunt is between the side of the tho-
racic aorta at the level near the ductus origin and
the left pulmonary artery. The superior vena cava
may be connected directly to both pulmonary ar-
teries (bidirectional cavopulmonary) or to the
right pulmonary artery (classic Glenn). Long-
standing Glenn shunts may be associated with
hypoxemia related to pulmonary arteriovenous
shunts, as well as systemic venous collaterals [20].
Pathophysiology of congenital heart disease
Respiratory issues
Most adults who have CHD have abnormal
static lung function when compared with age-
matched controls [21]. Although this may have
relevance in the recovery room or ICU, it is less
likely to influence management of ventilation in
the operating room. The ability to increase car-
diac output in response to stress is reduced in
a wide selection of congenital lesions [22]. Patients
may have hypoplastic pulmonary arteries and a di-
minished number of alveoli related to a previous
systemic to pulmonary artery shunt [23]. This dif-
ference is a major consideration in positioning the
patient in the operating room if the patient’s
hypoplastic lung is the dependant lung or, worse,
the dependent lung in one-lung anesthesia.
Hypoxemia and shunting
Low oxygen delivery to tissues may be caused
by inadequate central oxygenation of blood as
occurs in right-to-left shunts. Patients with left-
to-right or bidirectional shunts will have decreased
delivery of oxygen to the systemic circulation to the
extent that pulmonary blood flow exceeds systemic
blood flow. The presence of high oxygen saturation
does not ensure adequate oxygen delivery. If
a patient has a large atrial septal defect (ASD) or
common atrium without any outflow tract obstruc-
tion, a dramatic fall in pulmonary vascular re-
sistance (PVR) will cause a greater mismatch in
blood flow by increasing pulmonary and decreas-
ing systemic circulations. The patient’s increased
arterial oxygen saturation is at the expense of
systemic output; therefore, a pink patient may
paradoxically develop a metabolic acidosis. In the
management of left-to-right shunts, one should
always keep in mind the balance of PVR and
systemic vascular resistance (SVR), regardless of
the level of the shunt (atrial, ventricular, or great
vessels), and choose anesthetic agents and ventila-
tion strategies that will maintain a balance of the
PVR and SVR. Residual left-to-right shunts also
predispose patients to ventilation perfusion mis-
matching. There are theoretical delays with
 THE ANESTHESIOLOGIST’S ROLE
inhalational induction in the presence of a right-to-
left shunt and in intravenous induction in the
presence of a left-to-right shunt, but they have
little clinical impact.
Complex cyanotic patients may never have had
a palliative surgical procedure but may develop
substantial aortopulmonary collaterals. As a re-
sult, in addition to an intracardiac right-to-left
shunt, these patients have a significant left-to-right
shunt through their collaterals. Aortopulmonary
collaterals may connect with the central pulmo-
nary arteries or supply a segment of lung in-
dependently of the pulmonary circulation. The
coronary arteries compete for diastolic blood flow
with these aortopulmonary collaterals, and ma-
neuvers that increase the left-to-right shunt
compromise myocardial perfusion, putting an in-
creased strain on the systemic ventricle such that it
may start to fail and contribute to pulmonary
venous hypertension and pulmonary edema [24].
Hematologic considerations in congenital heart
disease
The hematologic considerations in complex
cyanotic heart disease are problematic in any
surgery. Oxygen delivery to the tissues depends
on many variables, which include the inspired
oxygen concentration, minute ventilation, diffu-
sion capacity, cardiac output, tissue extraction,
and hemoglobin concentration. The increase in
red blood cell mass is an adaptive response to
hypoxemia resulting in many cyanotic patients
having hematocrits in excess of 70%. There is no
agreed upon level of expected hemoglobin for
a given degree of hypoxemia, but a patient with
a resting saturation under 85% on room air
should have a hemoglobin of at least 18. Decom-
pensated erythrocytosis is defined as a high
hematocrit with moderate to severe symptoms of
hyperviscosity [25]. Iron deficiency is common in
cyanotic heart disease and causes red cells to
change from flexible biconcave disks to inflexible
microspheres. Microcytic iron-deficient erythro-
cytes have reduced oxygen-carrying capacity and
reduced deformability, exacerbating symptoms
of hyperviscosity and decreasing oxygen delivery
[26]. Preoperative fasting or a bowel preparation
may create or exacerbate symptoms of hypervis-
cosity, and preoperative hydration should be
maintained orally or intravenously.
Production of red cells is increased in response
to tissue hypoxia but not that of other cell lines or
coagulation factors. Tempe and Virmani have
573
published a comprehensive review of coagulation
abnormalities in cyanotic heart disease describing
thrombocytopenia, platelet function abnormali-
ties, decreased production of coagulation factors,
vitamin K deficiency, and primary fibrinolysis
[26]. The problem of perioperative coagulopathy
in cyanotic heart disease is well known, and ade-
quate transfusion services must exist when con-
templating surgery, including labor and delivery.
Preparing the patient
To avoid strokes, precautions against air
emboli must be taken in all patients with right-
to-left shunts, but ‘‘air filters’’ or ‘‘bubble traps’’
are not a substitute for meticulous de-airing of all
intravenous lines, stopcocks, and injection ports.
Air filters frequently obstruct at the most in-
opportune times, particularly with propofol, mak-
ing it difficult to infuse blood through the line.
Before connecting the intravenous line to the
patient, one should recheck the line because
bubbles can mysteriously reappear. When inject-
ing into a line, it is helpful to use a stopcock with
the syringe held vertically with the injection port
inferior to the syringe. One should aspirate before
injection and tap the connector, allowing any air
bubbles to float to the superior aspect of the
syringe. Nitrous oxide (N2O) may increase PVR
and exacerbate an air embolus, and its use is
discouraged.
Premedication should be given with the pa-
tient’s physiology in mind. Knowledge of room air
saturation before sedation is also helpful in mon-
itoring the patient in the presurgical waiting area.
Patients who are dependent on maintaining a bal-
ance of PVR and SVR should wait in a monitored
setting. Too much sedation may result in hyper-
carbia and falling oxygen saturation, whereas an
antibiotic such as vancomycin may drastically
lower the SVR. In complex cases, the choice of
anesthetic agents is not as important as the
knowledge and care with which the anesthesiolo-
gist manages the patient’s hemodynamics.
Pharmacology of anesthetic agents
The components of general anesthesia include
hypnosis, narcosis, and surgical exposure. The
delivery can be intravenously or via volatile
anesthetic agents. Conscious sedation involves
the use of hypnotics and narcotics in a spontane-
ously breathing patient and in the past has been
referred to as a ‘‘neuroleptic technique.’’
574 HEGGIE & KARSKI
For various reasons, volatile (inhalational)
anesthetic agents cause dose-dependent decreases
in the systemic blood pressure. Halothane and
enflurane cause a dose-dependent decrease in
myocardial contractility, whereas isoflurane, sevo-
flurane, and desflurane cause a decrease in SVR
with little effect on cardiac output. All volatile
anesthetic agents lower the arrhythmogenic
threshold for endogenous and exogenous epineph-
rine. Halothane has little direct effect on the
coronary vasculature, whereas isoflurane, sevo-
flurane, and desflurane all cause mild coronary
vasodilatation. These agents all attenuate the
effects of myocardial ischemia, and the subject
of ischemic preconditioning is currently a major
focus of investigation. The volatile anesthetic
agents display variable effects on the pulmonary
vasculature including hypoxic pulmonary
vasoconstriction. Halothane causes a flow-
independent pulmonary vasoconstriction, whereas
sevoflurane and desflurane do not alter hypoxic
pulmonary vasoconstriction. All volatile anes-
thetic agents inhibit the baroreceptor reflex [27].
The volatile anesthetic agents are used as
a component of general anesthesia, and the agent
selected should complement the physiologic goals
in the adult patient with CHD. For example,
a patient with a cavopulmonary anastomosis will
do better with isoflurane than halothane because
the cardiac output is maintained, and there is
a drop in the SVR while PVR is maintained.
Patients with dynamic outflow tract obstruction
do well with halothane because it decreases
myocardial contractility without dropping the
vascular resistance.
Nitrous oxide is an insoluble gas used as an
adjunct to general anesthesia and is delivered in
conjunction with oxygen in concentrations of
50% to 70%. It is used occasionally in obstetric
anesthesia in a concentration of 50% mixed with
50% oxygen when a patient arrives in active or
precipitous labor and there is insufficient time for
an epidural. Nitrous oxide causes a decrease in
venous capacitance, moderate increases in pulmo-
nary arterial pressures, and a mild decrease in
contractility, limiting its use in patients who have
CHD. Additionally, nitrous oxide diffuses rapidly
into closed gas spaces within the body such as the
pneumothorax, pneumoperitoneum, and venous
air emboli. Alveolar concentrations of 50% to
75% correlate with twofold to fourfold increases
in the volume of the gas space. The potential for
venous air emboli constitutes an absolute contra-
indication of nitrous oxide [28].
The intravenous hypnotic agents include pro-
pofol, barbiturates (thiopental), benzodiazepines
(midazolam), etomidate, and ketamine. Propofol
is an alkylphenol that is widely used in anesthesia
practice for induction, maintenance of anesthesia,
and sedation in the ICU. In healthy patients it
causes a significant decrease in the systolic arterial
pressure, stroke volume, cardiac index, and SVR.
Midazolam is a water soluble benzodiazepine with
rapid onset and short duration of action and is
cleared significantly faster than diazepam or
lorazepam. When used in induction doses of
0.05 to 0.2 mg/kg, it is associated with mild de-
creases in the mean arterial pressure and heart
rate. Midazolam has no analgesic properties and
is used routinely in the induction of cardiac pa-
tients in conjunction with fentanyl. Thiopental is
a barbiturate with a short onset of action, rapid
redistribution half-life, but a much longer elimina-
tion half-life than midazolam or propofol. It is
known to increase venous capacitance, decrease
venous return, decrease cardiac index, and in-
crease heart rate. It will significantly decrease car-
diovascular parameters in hypovolemic patients
and patients with impaired ventricular function.
Etomidate is a carboxylated imidazole derivative
that is reported to have minimal cardiovascular
effects but has been shown to suppress directly
adrenal-cortical function. It is not commercially
available in the jurisdiction of the author. It ap-
pears to have advantages in emergency situations
to induce a patient who is hypovolemic or has
tamponade. The controversy of the suppression
of the stress response has limited the popularity
of this drug in anesthesia and intensive care med-
icine [29]. Ketamine is another agent that has ad-
vantages in the unstable and hypovolemic patient
but with unpleasant side effects of hallucinations
and vivid dreams (nightmares). It is a phencycli-
dine derivative with profound analgesic proper-
ties. In contrast to the other hypnotic agents
mentioned, it stimulates the cardiovascular system
by the central nervous system, causing increases in
heart rate and pulmonary arterial pressure. These
effects can be minimized by premedication with
benzodiazepines. Ketamine is a direct myocardial
depressant [27].
Opioids can be naturally occurring, semisyn-
thetic, or synthetic and act at the opioid receptor
to produce a pharmacologic effect. The term nar-
cotic is used interchangeably but has a legal impli-
cation meaning a controlled substance that can
produce dependency and addiction. Examples of
naturally occurring opioids are morphine and
 THE ANESTHESIOLOGIST’S ROLE
codeine. Semisynthetics include hydromorphone
and meperidine. Synthetics include fentanyl, su-
fentanil, and remifentanil. Morphine is associated
with significant hypotension owing to histamine
release. Its use in cardiac anesthesia diminished
with the advent of synthetic narcotics that have
minimal cardiovascular impact [27]. Although
there are minimal cardiovascular considerations,
opiates are well known to be central respiratory
depressants and in the spontaneously breathing
patient will raise the apneic threshold (the alveolar
carbon dioxide level at which respiration is
stimulated) [28].
Neuromuscular blockers are either depolariz-
ing, such as succinylcholine, or nondepolarizing,
such as pancuronium and rocuronium. The
patient is paralyzed, and the surgeon is provided
with sufficient exposure to operate. They have no
direct effect on the myocardium and few cardio-
vascular consequences. Pancuronium may cause
a mild tachycardia. The duration of neuromuscu-
lar blockade depends on the agent used as well as
the hepatic and renal reserve of the patient. These
factors must be taken into consideration in
planning the postoperative emergence and extu-
bation of the patient.
Regional anesthesia
Regional anesthesia can include neural axial
blocks (epidural and spinal anesthesia) as well as
region-specific techniques such as axillary, ingui-
nal, and others. Regional anesthesia is often used
in conjunction with general anesthesia and con-
scious sedation to provide intraoperative and
postoperative pain relief. All patients must have
a normal coagulation profile. Both spinal and
epidural anesthesia cause a significant drop in
preload and afterload because the sympathetic
nervous system is affected in addition to the
sensory and motor blockade. These hemodynamic
effects are more pronounced with spinal anesthe-
sia. Absolute contraindications to epidural and
spinal anesthesia are local infection at the site of
injection, sepsis, and patient refusal.
Epidural anesthesia involves insertion of
a catheter into the potential space exterior to the
dura. The technique is most commonly used in the
lumbar region for relief of pain in labor and for
cesarean section. The advantage of an epidural is
that local anesthetic and narcotic can be slowly
injected to achieve the desired block. Hemody-
namic consequences such as a drop in preload
575
from venodilation and a drop in SVR can be
anticipated, particularly with arterial monitoring.
Thoracic epidurals are used in conjunction with
general anesthesia for management of postopera-
tive pain during upper abdominal and thoracic
surgery.
Spinal anesthesia involves a smaller needle
advanced across the dura. Local anesthetics and
opioids are injected directly into the cerebrospinal
fluid. This technique is limited to the lumbar
region as the spinal cord ends at L1. Only one
dose is administered, and the block has a finite
duration. Should the block wear off before the end
of the procedure, the patient will likely need
conversion to a general anesthetic. The onset of
the block is rapid, as are the hemodynamic
consequences. The height and distribution of the
block can be influenced by altering the density of
the local anesthetic solution in relation to the
cerebrospinal fluid and positioning of the patient.
The duration of the block is dependent on the
agent used and whether adrenaline is added to the
local anesthetic solution. Spinal anesthesia is
widely used in cesarean sections; however, in the
adult parturient with CHD, an epidural with
a slower onset of action would be desirable.
Situations in which spinal anesthesia would be
contraindicated would include unrestricted left-
to-right shunts (because of the possibility of
converting them to right-to-left shunts) and pre-
load-dependent lesions, such as patients with
a Fontan circulation, Ebstein’s anomaly, tetralogy
of Fallot with right ventricular dysfunction, and
patients with left-sided outflow tract obstruction.
Noncardiac surgery
The general considerations for operating on
the adult with CHD can be demanding. Non-
cardiac surgical procedures also pose particular
challenges, including one-lung ventilation, jet
ventilation for thoracic and tracheal procedures,
and procedures that involve insufflation of CO2
into a body cavity. There are also issues related
to placing the patient in the lateral, prone, and
Trendelenburg positions, and head up for pro-
longed periods of time. Warner and coworkers
[30] retrospectively reviewed the records of 276
adult CHD patients less than 50 years old over
a 6-year period and found that patients with cya-
nosis who received treatment for heart failure and
who were in poor general health were at higher
risk of complications. This study also found that
576 HEGGIE & KARSKI
operations on the respiratory and neurologic sys-
tems were associated with high frequencies of
complications. Ammash and coworkers [31] retro-
spectively evaluated 58 adult CHD patients with
Eisenmenger syndrome and found that among
the 24 patients who underwent noncardiac surgery
there were two deaths. One of the deaths followed
an appendectomy in a community hospital; the
other occurred in a high-risk patient having spinal
fusion at a tertiary care center who was in the prone
position for an anesthetic time of 525 minutes.
Whether CHD patients should be referred to
regional tertiary care centers for care by a subspe-
cialist cardiac anesthesiologist or cared for at their
local center where their primary cardiologist can
follow them is controversial. Many regional car-
diac surgical centers do not provide cardiac surgery
for adults with CHD but do have cardiac and
thoracic trained anesthesiologists and intensive
care and blood transfusion facilities that would
be able to attend to the situations outlined in the
following sections.
Positioning
Lateral
In the lateral position, the anesthetized para-
lyzed patient has a mismatch of ventilation and
perfusion. Nephrectomy, thoracotomy, and hip
procedures are examples of surgery performed in
the lateral position. There is relatively good
ventilation and decreased perfusion in the ‘‘up’’
nondependent lung and relatively decreased
ventilation but good perfusion in the ‘‘down’’
dependent lung. There is a degree of right-to-left
shunt in the ‘‘down’’ dependent lung during two-
lung ventilation in the lateral position and a re-
sultant increase in the alveolar arterial oxygen
gradient [28]. These factors must be taken into
consideration in positioning a patient with pulmo-
nary arteriovenous fistulas and systemic arterial-
to-pulmonary arterial shunts, particularly if the
shunt has caused hypoplasia of one of the lungs.
Head up (reverse Trendelenburg)
Head and neck surgery, as well as procedures
in the upper abdomen such as cholecystectomy,
require the head-up position. The head-up posi-
tion combined with raised intra-abdominal pres-
sure, laparotomy, and laparoscopic insufflation of
CO2 will dramatically reduce preload. In upper
abdominal surgery, the abdominal contents or
the insufflation of CO2 will push upward against
the diaphragm and necessitate higher inspiratory
pressures. This factor in combination with the
decrease in venous return will compromise
preload-dependent right ventricles in patients
who have Ebstein’s anomaly, tetralogy of Fallot
with right ventricular dysfunction, or passive pul-
monary perfusion (eg, cavopulmonary anastomo-
sis and Fontan circulations) in whom adequate
venous filling pressures and low intrathoracic
pressures are needed. In head and neck surgery,
central venous monitoring with brachial central
lines enables the anesthesiologist to maintain ade-
quate filling pressures. Consideration of conver-
sion of a laparoscopic to an open procedure
may be needed in some situations. Cases that in-
volve caval occlusion or compression will necessi-
tate open and frequent communication between
the anesthesiologist and the surgeon in addition
to central venous monitoring and ventilation strat-
egies to minimize intrathoracic pressure.
Head down (Trendelenburg)
Gynecologic and urologic procedures involve
steep head-down positioning, that is, Trendelen-
burg and lithotomy, and may be problematic for
patients with Glenn shunts, Fontan procedures,
and other cavopulmonary anastomoses in which
systemic venous hypertension could compromise
cerebral perfusion. The abdominal contents are
pushed up against the diaphragm, and the patient
must be mechanically ventilated with higher
inspiratory pressures even though he or she may
have difficulty tolerating positive pressure venti-
lation. If the central venous pressures are high,
increasing contractility with an inotrope will
improve the cerebral perfusion pressure.
Prone
The prone or face-down position has physio-
logic consequences for CHD patients. Once the
patient is draped and the case has started, the
anesthesiologist will have no access to the precor-
dium or neck and limited access to the airway.
Spinal instrumentation is the most common pro-
cedure in which the prone position is used, but
rectal procedures may also be done prone. Central
venous access must be obtained before position-
ing. There is impaired venous return owing to
direct venous compression in the pelvis and sub-
sequent venous pooling in the legs. Compression
of the abdominal and thoracic compartments can
occur despite the use of a bolster or special frame
designed to allow the abdominal compartment to
hang free [28]. Cardiopulmonary resuscitation is
not possible in the prone position, and a bed
(on which to roll the patient) must be kept
 THE ANESTHESIOLOGIST’S ROLE
immediately available should an event occur. Pads
for cardioversion can be applied before draping.
The series by Ammash and coworkers reported
a death in an Eisenmenger patient following
spinal surgery in the prone position [31]. A case
report of a 14-year-old boy with a Fontan circula-
tion who successfully underwent spinal instrumen-
tation in the prone position highlights these
challenges [32].
One-lung ventilation
Situations in which one-lung ventilation may
be needed would be the placement or extraction of
epicardial atrial pacemaker wires, thoracotomy
for pulmonary hemorrhage, chylothorax, or lung
biopsy, and procedures involving the descending
thoracic aorta. Anesthesiologists can ventilate the
two lungs independently by means of double-
lumen endotracheal tubes, or one lung with the
use of bronchial blockers. Once the ‘‘up’’ non-
dependent lung is not ventilated, there is addi-
tional right-to-left shunting because that lung is
perfused and not ventilated. Minute alveolar
ventilation is increased in the lung that is
‘‘down’’ dependent and able to compensate for
carbon dioxide elimination, but there will likely be
a further increase in the alveolar arterial oxygen
gradient and a drop in the oxygen saturation of
the arterial blood [28]. It does not require much
imagination to see that this scenario could convert
a previous left-to-right shunt to a right-to-left
shunt. This situation may be especially difficult
in patients who have Eisenmenger syndrome and
a high PVR, as well as in Glenn and Fontan circu-
lations where the PVR must be kept as low as
possible to fill the left atrium [6,33]. Thoracic sur-
geons also use a technique of carbon dioxide in-
sufflation into the thoracic cavity referred to as
‘‘video-assisted thoracoscopic surgery’’ where the
lung is partially collapsed and, in addition to the
mechanical increase in PVR and right-to-left
shunt in the nonventilated lung, the arterial
carbon dioxide level is elevated from the insuf-
flated carbon dioxide.
Laparoscopic procedures
Laparoscopic procedures involve insufflation
of carbon dioxide into the peritoneal cavity
creating a pneumoperitoneum and are widely
used in general, gynecologic, and urologic surgery.
The insufflation of CO2 causes a rise in the arterial
carbon dioxide concentration owing to vascular
577
absorption across the peritoneum and intravascu-
lar embolization. The extent of the cardiovascular
changes is primarily dependent on the intra-
abdominal pressure attained and the patient’s
position. This rise in intra-abdominal pressure is
transmitted to the chest, causing decreased lung
volumes and decreased compliance necessitating
increased inflation pressures in mechanically
ventilated patients [34]. The head-up position
and laparoscopy will reduce venous return,
cardiac output, and mean arterial pressure and
will increase PVR [35]. Although the patient will
have increased intrathoracic pressures, increasing
the respiratory rate with smaller tidal volumes
and pharmacologic interventions to dilate the pul-
monary circulation and improve contractility can
overcome the limitations of laparoscopic surgery.
If the anesthesiologist is unable to compensate,
conversion to an open procedure must be contem-
plated and the threshold for this decision
discussed with the surgeon and attending cardiol-
ogist ahead of time.
Shared airway
Situations in which the surgeon and the
anesthesiologist share the airway involve laryn-
goscopy and bronchoscopy using the technique of
jet ventilation. Tracheal complications of a pro-
longed ICU stay would be one of the indications
for these procedures, and a patient may need
evaluation if a complicated cardiac revision is
under consideration. In a diagnostic scope, the
ventilation can be on 100% oxygen, but if the
operation involves laser of the airway, the portion
of the procedure using laser must be done on
room air to prevent an airway fire. The anesthe-
siologist delivers intermittent short bursts of
pressurized air/oxygen to the airway, and the
conventional method of monitoring is observation
of arterial blood gases and movement of the chest
wall. In between the brief periods of ventilation,
while the patient is left apneic so the surgeon can
operate on the trachea or larynx, the oxygen
saturation will drop and the arterial CO2 will
rise. The anesthetic is delivered completely intra-
venously, and these cases require paralysis and
vigilance in monitoring, usually with an arterial
line. The nature of the operation compromises
the anesthesiologist’s ability to ventilate the pa-
tient effectively, and there is no way of regulating
or monitoring the pressure delivered to the lungs.
This situation would be problematic for any pa-
tient dependent on a low PVR, such as the patient
578 HEGGIE & KARSKI
with tetralogy with right ventricular dysfunction
and pulmonary insufficiency, as well as patients
with Fontan circulations. Complications include
pneumothorax, pneumomediastinum, and com-
promise of airway patency.
Diagnostic and interventional services
Anesthesiologists are increasingly involved in
cardiologic interventional and diagnostic services
either as a provider of anesthetic services or as
a collaborator in perioperative transesophageal
echocardiography (TEE) monitoring in cardiac
surgery [36,37]. Anesthesia for diagnostic cathe-
terization or TEE must permit the patient’s phys-
iology under anesthesia to reflect his or her awake
and ambulatory state as much as possible. Most
anesthetic agents lower SVR and lessen the sever-
ity of regurgitant lesions. Ketamine is well known
to stimulate the central sympathetic nervous
system and has been found to counteract the
SVR-lowering effects of the hypnotic propofol
when the two are used concomitantly. Ketamine
has the added advantage of allowing spontaneous
breathing on an inspired oxygen concentration
close to that of room air. It has the disadvantage
of occasionally causing unpleasant or dysphoric
dreams when used alone [38,39].
The same considerations would apply to anes-
thesia for TEE, because the goal is not to alter the
patient’s ambulatory physiology. These proce-
dures are performed most easily with the patient
in the lateral position and breathing spontane-
ously after the oropharynx has been sprayed with
topical lidocaine. Appropriate recovery facilities
must be available, and it is safer to conduct these
studies in a monitored facility, operating room,
post anesthetic recovery room, or coronary care
unit, particularly if the indication for TEE is to
rule out thrombus before cardioversion. The
cardioversion can be performed immediately fol-
lowing the procedure, before emergence from
anesthesia, with the patient in the lateral position
as the airway is unprotected owing to the topical
local anesthetic. In an operating room or ICU
setting, any complications related to the cardio-
version can be addressed quickly because the
appropriate staff and equipment are immediately
available.
In interventional cases, the diagnosis has been
established, the cardiologist must be able to
conduct the procedure safely without distraction,
and the patient must be free of pain [36]. Exam-
ples would include stenting of coarctation,
percutaneous valve insertion, insertion of biven-
tricular pacing leads, and ablation therapy. Inter-
ventional cases necessitate a deeper level of
conscious sedation or general anesthesia and pos-
sible neuromuscular blockade [37]. When diagnos-
tic angiography precedes the intervention, it is
reasonable to use conscious sedation as outlined
previously and convert to a general anesthetic
for the intervention. Recovery of the patient is
best served in the post anesthetic care unit or
a monitored ICU setting not only for hemody-
namic monitoring but also because the degree of
somnolence and the risk of incomplete reversal
of a neuromuscular blockade may exceed the ca-
pabilities of the recovery area of the interventional
suite [37].
Arrhythmias and cardioversions
Arrhythmias are common in adult patients
with CHD. In a large survey of adults with
CHD, Kaemmerer and colleagues demonstrated
that one fourth of all admissions to a specialized
adult CHD facility were urgent and unscheduled.
The most common reason for readmission was
cardiovascular, of which arrhythmia was the most
common [4]. A cross-sectional study of 2609
adults with CHD found that sudden death was
the most common cause of death (26%) [40]. Ven-
tricular arrhythmias occur more commonly in
patients with left-sided obstruction, tetralogy of
Fallot, and severe ventricular dysfunction [41].
The Euro Heart Survey program published a ret-
rospective cohort study performed from 1998 to
April 2004 of more than 4000 adult patients
with one of eight different congenital heart diag-
noses. The incidence of supraventricular arrhyth-
mias was 18% overall and highest in the Fontan
group (45%), followed by secundum ASD
(28%), transposition (26%), and tetralogy of Fal-
lot (20%). The incidence of ventricular arrhyth-
mias was 5% overall and highest in the tetralogy
of Fallot group (14%) [3].
Anesthesia for cardioversion is a simple and
uncomplicated procedure that requires a short
hypnotic agent, support of the airway, and venti-
lation for a brief period of time. Nevertheless,
cardioversions are fraught with complications, and
the anesthesiologist must anticipate these compli-
cations and have a clear back up plan in mind
before proceeding. A telephone consultation with
a cardiovascular anesthesiologist experienced with
CHD may be warranted. The procedure should
take place in a monitored environment that is
 THE ANESTHESIOLOGIST’S ROLE
accessible to ancillary personnel. Should the pro-
cedure result in a bradyarrhythmia, the means to
pace should be available. This treatment may be via
a temporary wire via central venous access or may
be performed transcutaneously in the patient who
has a cavopulmonary anastomosis (as there will be
no intravenous access from above to the ventricle).
The procedure may be preceded by TEE to rule out
atrial thrombus. Of 52 adults undergoing the
Fontan operation, 17 (33%) had asymptomatic
thrombi in the right atrium and one in the left
atrium, surprisingly with no correlation to the
presence of arrhythmias [42]. The considerations
for pacemaker insertion as well as defibrillator
implant are much like those outlined in preparing
for noncardiac surgery but with the additional
concerns outlined previously.
Labor and delivery
Women with CHD comprise the majority of
cardiac patients presenting for labor and delivery.
These patients are best managed in a high-risk
pregnancy unit by a multidisciplinary team from
obstetrics, cardiology, anesthesia, and neonatol-
ogy [43]. Comprehensive reviews of pregnancy
and heart disease are in the recent literature,
including didactic accounts of the normal cardio-
vascular changes of pregnancy and the implica-
tions for patients with specific congenital heart
defects [44–47]. A large prospective multi-center
study of 562 pregnant women with cardiac disease
(74% with CHD) with 599 pregnancies defined
four predictors of primary cardiac events: (1)
a prior cardiac event or arrhythmia, (2) a baseline
New York Heart Association classification of
heart disease greater than II or cyanosis, (3) left-
sided heart obstruction, and (4) reduced systemic
ventricular function. There was no association be-
tween the type of delivery, vaginal versus cesarean
section, and peripartum cardiac events [48]. An-
other study reviewed the records of 90 pregnancies
in 53 women with CHD using the same definitions
of obstetric, neonatal, and cardiac events as in the
previous study and found a cardiac event rate of
19%, of which congestive heart failure was the
most common followed by arrhythmias. Indepen-
dent risk factors for pulmonary edema included
a history of congestive heart failure, smoking,
and a decreased subpulmonary ejection fraction.
All but one of the patients in this study received
an epidural [49].
A team conference with anesthesia in atten-
dance should be held early in the patient’s
579
pregnancy to plan and disseminate the treatment
goals. Epidurals are desirable in most patients,
although anticoagulants will preclude this if not
stopped in sufficient time. Although there are
congenital defects in which any drop in afterload
must be avoided, low concentrations of local
anesthetics and the use of intrathecal and epidural
narcotics will diminish this risk, particularly if
initiated slowly with appropriate monitoring and
vasopressors available. The extent of invasive
monitoring will be dictated by the patient’s
functional class and type of defect. Central venous
and especially pulmonary arterial monitoring
should only be used when the information is
essential and cannot be obtained any other way.
In a patient with a significant aortic coarctation,
arterial monitoring below the obstruction is
necessary to ascertain the placental perfusion
pressure; the proximal aortic pressure may not
drop significantly with a low thoracic level epidu-
ral, but the distal aortic pressure may drop
significantly [50]. The use of nitrates in managing
pulmonary hypertension will adversely affect the
uterine tone. There are limited reasons to avoid
vaginal delivery. The only cardiac indications for
cesarean section are Marfan syndrome with a di-
lated aortic root and the failure to stop warfarin
in sufficient time before labor [43].
Perioperative transesophageal echocardiography
Institutions vary as to whether anesthesiolo-
gists or cardiologists provide perioperative TEE
monitoring. Who is best suited to do this is subject
to debate. Regardless of the primary specialty, the
TEE echocardiographer must have additional
training in the reading of complex congenital
echocardiography [51,52]. The echocardiographer
should function independently from the role of
the primary anesthesiologist and should be avail-
able to return to monitor the patient. Ideally,
a collaborative relationship between the cardiol-
ogy service and the anesthesiology service will
exist. An exhaustive TEE review in the adult
CHD patient has been published recently by
Russell and coworkers [53].
Cardiac surgery
Patients presenting for repair of a complex
congenital anomaly or revision of a previous
repair will do so to an established congenital
cardiac surgical center, whether pediatric or adult
580 HEGGIE & KARSKI
in focus. The acuity of these procedures is much
greater than the situations outlined previously,
and co-existing disease, particularly respiratory,
renal, and hepatic, may disqualify some patients
for cardiac surgery. Adults of small stature will
have considerations independent of their cardiac
anatomy and may require smaller priming vol-
umes for the heart lung machine, a prudent choice
of central venous catheters, airway considerations,
and vigilance in the administration of parenteral
fluids. Large bore intravenous lines with the
capability to transfuse fluids at high flow rates
are routine, and care must be taken that a ‘‘run-
away liter’’ is not inadvertently administered. The
adult CHD patient of small stature and compro-
mised ventricular function may not tolerate such
a volume load. Precautions for repeat sternotomy
include blood readily available for transfusion,
heparinization before sternotomy, the use of
antifibrinolytics, and access to femoral vessels
for cannulation (true for all patients requiring
repeat sternotomy). Anesthetic considerations for
cardiac surgery in each individual congenital heart
defect, beyond the issues outlined herein, are
beyond the scope of this article. The adult with
CHD is a core topic of a didactic cardiac
anesthesia fellowship, but not all graduates will
necessarily be involved in the care of these
patients. The following discussion presents a brief
overview of issues that the cardiac anesthesiolo-
gist will encounter and provides an overview for
cardiologists participating in the perioperative
care of these patients. Several recent reviews detail
the management of cardiac anesthesia for this
population as well as anesthetic management of
specific defects [6,7,10,54,55].
The ICU should be in close proximity to the
operating room, and transport should occur in
a seamless and monitored fashion. On arrival to
the ICU, there must be a surgical and anesthesia
report to the nursing, respiratory, and medical
staff. Hemodynamic goals and respiratory param-
eters should be clearly outlined, and a sketch or
drawing of the patient’s preoperative and post-
operative anatomy is helpful. Much of what
occurs in the ICU is a natural extension of the
anesthetic management in the operating room,
and in difficult cases the ICU staff should be
available to the operating room to participate in
the post bypass care and the determination of
goals of care. Once the patient is admitted to the
ICU, the attending cardiologist should be in-
volved on a daily basis even though the patient
may be under the care of the attending surgeons
and attending ICU physicians. CHD cardiologists
can contribute to the postoperative management
of arrhythmias, echocardiography interpretation,
and interventional consultations. Knowledge of
the ICU events will also be helpful in managing
the patient on transfer to the ward setting.
Atrial septal defect, ventricular septal defects,
atrioventricular septal defects, and patent
ductus arteriosus
Most cases of secundum ASD and patent
ductus arteriosus (PDA) can be closed in the
catheterization laboratory with a device [56].
Some ventricular septal defects (VSD) can be
managed in the interventional suite [57]. Rarely,
patients will need to be transported to the operat-
ing room for retrieval of a malpositioned or
embolized device. On the rare occasions when
a device has embolized, the immediate anesthesia
goal is to prevent further travel.
Regardless of the location of the septal defect
or the presence of a PDA and the type of closure,
there are some basic considerations for the cardiac
anesthesiologist as outlined in the anesthesia of
patients with left-to-right shunts.
Patients who have their defects closed surgi-
cally are likely to have additional cardiac anom-
alies or have been referred for a Maze procedure
to treat atrial arrhythmias. These patients are
likely to be older and may also have significant
coronary artery disease [58]. Patients who have
mitral regurgitation in association with a primum
ASD may have pulmonary hypertension that may
not be entirely due to increased pulmonary blood
flow [59]. Precautions for maintaining an optimal
PVR should be observed as outlined previously.
Pulmonary vasodilators may be needed post
bypass in addition to maintaining an alkalotic
pH and ventilating the patient at tidal volumes
close to functional residual capacity with minimal
airway pressures. Most patients successfully wean
from bypass without intravenous nitrates or milri-
none, but these agents may be necessary. Occa-
sionally, one will have to add inhalational nitric
oxide [60]. A method of portable nitric oxide
delivery is essential for leaving the operating
room and delivering the patient to the ICU.
Tetralogy of Fallot
It is unusual to see an adult with uncorrected
tetralogy of Fallot, although this does occur in
immigrants whose anatomy was favorable for
 THE ANESTHESIOLOGIST’S ROLE
long-term survival or who originated from a coun-
try where surgical intervention was not available
[61–63]. Anesthesia for an uncorrected tetralogy
of Fallot is well described in Lake’s text of con-
genital heart anesthesia [64]. The most common
reasons for reoperation in adults with tetralogy
of Fallot corrected in childhood are to repair the
right ventricular outflow tract or replace the re-
gurgitant pulmonary valve, to repair a VSD patch
leak, or to repair severe tricuspid regurgitation
[65]. Anesthetic and ICU management must antic-
ipate right ventricular dysfunction, and measures
to minimize PVR post bypass and to improve
contractility may be indicated, although these
patients usually do very well [66]. The use of a pul-
monary arterial catheter is unnecessary. Postoper-
ative management should take into account
keeping the central venous pressure high as well
as maintaining a low PVR with a mild respiratory
alkalosis and avoiding extremes of lung volumes.
Left ventricular dysfunction will impact negatively
on the right ventricular output and should be
managed with inotropes and afterload reduction.
Arrhythmias should be anticipated and treated
preoperatively in the electrophysiology suite or
intraoperatively with cryoablation. Patients in
whom severe tricuspid regurgitation was an issue
preoperatively may need management of a coagul-
opathy secondary to hepatic congestion.
Ebstein’s anomaly
Ebstein’s anomaly is characterized by adher-
ence of one or more tricuspid leaflets to the
underlying myocardium with apical displacement
of the leaflets, ‘‘atrialization’’ of the right ventricle
and a dilated right atrium, and tricuspid regurgi-
tation. Many patients have an ASD that is often
associated with a right-to-left shunt. Surgery may
include closure of the atrial defect, performance of
an antiarrhythmic procedure, reduction of the
right atrium, and repair or replacement of the
tricuspid valve [67]. Pre-bypass considerations
must account for the right-to-left shunt as well
as the severe tricuspid regurgitation. Post bypass
one should anticipate right ventricular dysfunc-
tion and take measures to optimize contractility
and reduce the afterload of the right ventricle
[68]. Pulmonary vasodilators such as milrinone
or inhalational nitric oxide used with inotropes
such as dobutamine may be needed in the postby-
pass period. Chronic hepatic congestion from
severe tricuspid regurgitation may affect the
severity of postoperative coagulation.
581
Transposition of the great arteries
with a Mustard or Senning procedure
Adults with a transposition of the great arteries
usually have had an atrial type of repair (Mustard
or Senning operation) as a young child. This
operation leaves the patient with the morphologic
right ventricle as the systemic ventricle and the
tricuspid valve as the systemic atrioventricular
valve, both of which may fail in adult life.
Complications also include baffle leaks, superior
vena cava obstruction, inferior vena cava obstruc-
tion, pulmonary venous obstruction, and arrhyth-
mias (sinus node, atrial, or ventricular). These
patients seldom present for cardiac reoperation
but may present for pacemaker insertion or other
surgical procedures. Anesthetic management of
systemic ventricular (right ventricular) failure in
a Mustard patient is much like managing a patient
with a dilated cardiomyopathy. Central venous
monitoring in the Mustard patient can be prob-
lematic. A pulmonary artery catheter or pacing
wire must traverse the intra-atrial baffle to get to
the morphologic left ventricle in the subpulmo-
nary position. There is the potential for inducing
arrhythmia. The pulmonary arterial catheter will
appear to make a sharp ‘‘hairpin’’ turn or loop as
it enters the pulmonary artery from the morpho-
logic left ventricle. The morphologic right ventri-
cle (systemic ventricle) often becomes dilated
and hypokinetic. Anesthetic agents with minimal
effects on contractility that lower SVR are
desirable.
Transposition of the great arteries
with pulmonary stenosis and a Rastelli
procedure
Transposition of the great arteries may be
associated with a VSD and pulmonary stenosis.
The Rastelli procedure redirects blood flow from
the left ventricle to the aorta at the ventricular level,
and a valved extracardiac conduit is placed from
the right ventricle to the pulmonary artery. These
patients frequently present for reoperation because
of obstruction of the extracardiac conduit. The
right ventricle to pulmonary artery conduit lies just
beneath the sternum and is vulnerable to external
compression, as well as trauma or laceration on
re-entering the chest. Occasionally, the conduit is
critically narrowed, and all monitoring lines as well
as femoral cannulation and heparinization are
performed before induction and endotracheal in-
tubation performed with ensuing positive pressure
582 HEGGIE & KARSKI
ventilation. Spontaneously breathing intravenous
conscious sedation with propofol and ketamine
mixtures are effective. Adequate venous return and
filling pressures must be maintained, and anesthetic
agents that lower preload by causing venodilation
must be avoided.
Congenitally corrected (left) transposition
of the great arteries
Congenitally corrected transposition of the
great arteries requires discordant atrioventricular
and ventriculoarterial connections. The circula-
tion is physiologically viable, but the morphologic
right ventricle is, from birth, the systemic ventri-
cle. Such patients may present for systemic tri-
cuspid valve replacement, for replacement of the
previously placed valved conduit, or for correc-
tion of a VSD and pulmonary stenosis in a fashion
similar to that of the Rastelli procedure. The
anesthetic management is similar to that for
a dilated cardiomyopathy, as described in the
section on Mustard patients. Patients may present
with systemic ventricular dysfunction, systemic
atrioventricular valve regurgitation, or atrioven-
tricular block [69]. The associated lesions or com-
plications usually dictate the patient’s operative
and anesthetic care.
Glenn and Fontan operations
Single ventricle physiology is treated by by-
passing the right side of the heart by direct
connection of the inferior vena cava or superior
vena cava to the pulmonary arteries using a bi-
directional cavopulmonary shunt or the Fontan
operation (complete nonpulsatile systemic venous
return to the pulmonary artery). The superior
vena cava may be connected directly to both
pulmonary arteries (bidirectional cavopulmonary)
or to the right pulmonary artery only (classic
Glenn operation). A cavopulmonary shunt may
be a stage on the path to a Fontan operation.
Fontan described a total right-sided heart bypass
that has been subject to numerous modifications
over the years [70–74]. Although the differences in
anatomy may not change one’s approach to main-
taining the physiology, the anatomic differences in
these modifications may influence central venous
monitoring. In a patient with a classic Glenn op-
eration, a jugular or subclavian venous catheter
will not reach the heart, nor will it in a bidirec-
tional Glenn operation if the superior vena cava
has been divided from the right atrium. The
pressure in these central venous catheters reflects
the pulmonary arterial pressure in the involved
lung.
Most adult patients will present for a Maze
procedure and revision of a previous Fontan
operation [74,75]. The anesthetic care of this
group of patients is among the most challenging
in an adult cardiac anesthesia practice. It is impor-
tant to know if the Fontan anastomosis is fenes-
trated, with a right-to-left shunt. Precautions for
venous air emboli must be taken. Fontan patients
are highly dependent upon adequate preload. Pul-
monary blood flow is dependent upon the gradi-
ent between the vena cava and the left atrium;
therefore, increased PVR (induced by hypoxia,
hypercapnia, light anesthesia, increased/decreased
extremes in lung volumes) will have deleterious ef-
fects. The pre-bypass goals are no different than
post bypass. For patients to thrive, they must
have a controlled regular rhythm (preferably
sinus), sufficient right-sided filling pressures, min-
imal PVR, a competent systemic atrioventricular
valve, maintained left ventricular function, and
afterload reduction for the left ventricle. The
only means to monitor left atrial filling pressure
in these patients is by using a left atrial line. The
gradient between the right-sided filling pressures
and the left atrial line (transpulmonary pressure)
should be above 10 mm Hg, and colloids includ-
ing blood products are administered if left atrial
pressures are less than 8 mm Hg or transpulmo-
nary pressures less than 8 mm Hg. Gradients of
more than 2 to 3 mm Hg can be clinically signifi-
cant, and confirmation by echo or cardiac cathe-
terization is warranted. Methods to lower or
optimize the PVR include hyperventilation and
administration of bicarbonate for a pH greater
than 7.50, an increase in the inspired concentra-
tion of oxygen, optimization of ventilation with
appropriate sedation and paralysis if needed,
and pulmonary vasodilators both intravenous
and inhalational (milrinone, nitrates, prostacyclin,
nitric oxide). Atrial pacing or sequential atrioven-
tricular pacing at a higher rate, inotropes to
improve contractility, and afterload reduction
will all help to improve the cardiac output of the
single ventricle.
Coagulopathy is a major issue. Once hemosta-
sis is obtained, ventilation is weaned, preferably
with pressure support, and early extubation is the
goal. It is prudent to supervise extubation of the
patient, because prolonged Valsalva breaths,
retching, or coughing against the ventilator may
precipitate hemodynamic instability. Problems
 THE ANESTHESIOLOGIST’S ROLE
with pleural effusions and ascites may arise and
interfere with ventilation and should be treated
promptly with chest drains and peritoneal cathe-
ters. In patients with acute renal insufficiency
requiring renal replacement therapy, the insertion
of the vascular access line must be done with
caution because these lines are of relatively wide
diameter and may enter or partially occlude the
Glenn shunt. Arrhythmias are poorly tolerated,
and aggressive attempts to convert to sinus
rhythm should be contemplated with the advice
of the patient’s CHD cardiologist. Hypercoagula-
bility is another issue, and anticoagulation and
mechanical factors are indicated to prevent deep
venous thrombosis. The optimal type or duration
of anticoagulation is not clear [42,76,77].
Summary
Anesthesia for adults with CHD has many
challenging physiologic considerations. Collabo-
rative relationships of a multidisciplinary team
including cardiology, cardiac surgery, anesthesi-
ology, and intensive care are essential to ensure
positive outcomes in this population for non-
cardiac and cardiac surgery.
Acknowledgments
The author thanks Dr. Gary Webb and
Dr. William Williams for their encouragement
and leadership.
References
[1] Niwa K, Perloff JK, Webb GD, et al. Survey of
specialized tertiary care facilities for adults with
congenital heart disease. Int J Cardiol 2004;96(2):
211–6.
[2] Warnes CA, Liberthson R, Danielson GK, et al.
Task force 1: the changing profile of congenital heart
disease in adult life. J Am Coll Cardiol 2001;37(5):
1170–5.
[3] Engelfriet P, Boersma E, Oechslin E, et al. The spec-
trum of adult congenital heart disease in Europe:
morbidity and mortality in a 5 year follow-up
period. The Euro Heart Survey on adult congenital
heart disease. Eur Heart J 2005;26(21):2325–33.
[4] Kaemmerer H, Fratz S, Bauer U, et al. Emergency
hospital admissions and three-year survival of adults
with and without cardiovascular surgery for congen-
ital cardiac disease. J Thorac Cardiovasc Surg 2003;
126(4):1048–52.
[5] Therrien J, Gatzoulis M, Graham T, et al. Canadian
Cardiovascular Society Consensus Conference 2001
583
update: Recommendations for the Management of
Adults with Congenital Heart Disease–Part II. Can
J Cardiol 2001;17(10):1029–50.
[6] Lovell AT. Anaesthetic implications of grown-up
congenital heart disease. Br J Anaesth 2004;93(1):
129–39.
[7] Stayer SA, Andropoulos DB, Russell IA. Anesthetic
management of the adult patient with congenital
heart disease. Anesthesiol Clin North Am 2003;
21(3):653–73.
[8] Findlow D, Doyle E. Congenital heart disease in
adults. Br J Anaesth 1997;78(4):416–30.
[9] Galli KK, Myers LB, Nicolson SC. Anesthesia for
adult patients with congenital heart disease undergo-
ing noncardiac surgery. Int Anesthesiol Clin 2001;
39(4):43–71.
[10] Heggie J, Poirier N, Williams W, et al. Anesthetic
considerations for adult cardiac surgery patients
with congenital heart disease. Semin Cardiothorac
Vasc Anesth 2003;7(2):141–52.
[11] Webb G. Improving the care of Canadian adults
with congenital heart disease. Can J Cardiol 2005;
21(10):833–8.
[12] Gatzoulis MA. Adult congenital heart disease:
education, education, education. Nat Clin Pract
Cardiovasc Med 2006;3(1):2–3.
[13] Brickner ME, Hillis LD, Lange RA. Congenital
heart disease in adults: first of two parts. N Engl
J Med 2000;342(4):256–63.
[14] Brickner ME, Hillis LD, Lange RA. Congenital
heart disease in adults: second of two parts.
N Engl J Med 2000;342(5):334–42.
[15] Reid GJ, Irvine MJ, McCrindle BW, et al. Preva-
lence and correlates of successful transfer from pedi-
atric to adult health care among a cohort of young
adults with complex congenital heart defects. Pediat-
rics 2004;113(3):197–205.
[16] Walker F. Precious adults: a lesson in grown-up
congenital heart disease. Lancet 2003;362(9379):
241.
[17] Wacker AE. Outcome of operated and unoperated
adults with congenital cardiac disease lost to
follow-up for more than five years. Am J Cardiol
2005;95:776–9.
[18] Curtis S, Stuart G. Outcome in congenital heart
disease. Curr Paediatr 2005;15:549–56.
[19] Tempe DK, Siddiquie RA. Awareness during
cardiac surgery. J Cardiothorac Vasc Anesth 1999;
13(2):214–9.
[20] Magee AG, McCrindle BW, Mawson J, et al.
Systemic venous collateral development after the
bidirectional cavopulmonary anastomosis: preva-
lence and predictors. J Am Coll Cardiol 1998;
32(2):502–8.
[21] Rigolin VH, Li JS, Hanson MW, et al. Role of
right ventricular and pulmonary functional abnor-
malities in limiting exercise capacity in adults with
congenital heart disease. Am J Cardiol 1997;80(3):
315–22.
584 HEGGIE & KARSKI
[22] Fredriksen PM, Veldtman G, Hechter S, et al.
Aerobic capacity in adults with various congenital
heart diseases. Am J Cardiol 2001;87(3):310–4.
[23] Godart F, Qureshi SA, Simha A, et al. Effects of
modified and classic Blalock-Taussig shunts on the
pulmonary arterial tree. Ann Thorac Surg 1998;
66(2):512–7.
[24] Fabre OH, Porte HL, Godart FR, et al. Long-term
cardiovascular consequences of undiagnosed intra-
lobar pulmonary sequestration. Ann Thorac Surg
1998;65(4):1144–6.
[25] Oechslin E. Hematological management of the
cyanotic adult with congenital heart disease. Int
J Cardiol 2004;97(Suppl):1109–15.
[26] Tempe DK, Virmani S. Coagulation abnormali-
ties in patients with cyanotic congenital heart
disease. J Cardiothorac Vasc Anesth 2002;16(6):
752–65.
[27] Grogan K, Nyhan D, Berkowitz D. Pharmacology
of anesthetic drugs. In: Kaplan RLK, editor.
Kaplan’s cardiac anesthesia. 5th edition. Philadel-
phia: WB Saunders; 2006. p. 165–212.
[28] Miller R. Anesthesia. New York: Churchill Living-
stone; 2005.
[29] Jackson WL Jr. Should we use etomidate as an
induction agent for endotracheal intubation in
patients with septic shock? A critical appraisal.
Chest 2005;127(3):1031–8.
[30] Warner MA, Lunn RJ, O’Leary PW, et al. Out-
comes of noncardiac surgical procedures in children
and adults with congenital heart disease: Mayo Peri-
operative Outcomes Group. Mayo Clin Proc 1998;
73(8):728–34.
[31] Ammash NM, Connolly HM, Abel MD, et al. Non-
cardiac surgery in Eisenmenger syndrome. J Am
Coll Cardiol 1999;33(1):222–7.
[32] Vischoff D, Fortier LP, Villeneuve E, et al. Anaes-
thetic management of an adolescent for scoliosis sur-
gery with a Fontan circulation. Paediatr Anaesth
2001;11(5):607–10.
[33] Heller AR, Litz RJ, Koch T. A fine balanced
one-lung ventilation in a patient with Eisenmenger
syndrome. Br J Anaesth 2004;92(4):587–90.
[34] Gerges JG, Ghassan EK, Jabbour-Khoury SI. Anes-
thesia for laparoscopy: a review. J Clin Anesth 2006;
18:67–78.
[35] Gutt C, Mehrabi A, Schemmer P, et al. Circulatory
and respiratory complications of carbon dioxide in-
sufflation. Dig Surg 2004;21(2):95–105.
[36] Andropoulos DB, Stayer SA. An anesthesiologist
for all pediatric cardiac catheterizations: luxury or
necessity? J Cardiothorac Vasc Anesth 2003;17(6):
683–5.
[37] Chessa M, Carrozza C, Butera G, et al. The impact
of interventional cardiology for the management of
adults with congenital heart disease. Catheter Cardi-
ovasc Interv 2006;67:258–64.
[38] Oklu E, Bulutcu FS, Yalcin Y, et al. Which anes-
thetic agent alters the hemodynamic status during
pediatric catheterization? Comparison of propofol
versus ketamine. J Cardiothorac Vasc Anesth
2003;17(6):686–90.
[39] Kogan A, Efrat R, Katz J, et al. Propofol-ketamine
mixture for anesthesia in pediatric patients undergo-
ing cardiac catheterization. J Cardiothorac Vasc
Anesth 2003;17(6):691–3.
[40] Oechslin EN, Harrison DA, Connelly MS, et al.
Mode of death in adults with congenital heart dis-
ease. Am J Cardiol 2000;86(10):1111–6.
[41] Khairy P, Dore A, Talajic M, et al. Arrhythmias in
adult congenital heart disease. Future drugsdexpert
review of cardiovascular therapy 2006;4(1):1–32.
[42] Balling G, Vogt M, Kaemmerer H, et al. Intracar-
diac thrombus formation after the Fontan opera-
tion. J Thorac Cardiovasc Surg 2000;119(4):745–52.
[43] Siu S, Coleman J. Heart disease and pregnancy.
Heart 2001;85:710–5.
[44] van Mook WN, Peeters L. Severe cardiac disease
in pregnancy. Part I. Hemodynamic changes and
complaints during pregnancy, and general manage-
ment of cardiac disease in pregnancy. Curr Opin
Crit Care 2005;11(5):430–4.
[45] van Mook WN, Peeters L. Severe cardiac disease in
pregnancy. Part II. Impact of congenital and ac-
quired cardiac diseases during pregnancy. Curr
Opin Crit Care 2005;11(5):435–48.
[46] Colman J, Siu S. Pregnancy in adult patients with
congenital heart disease. Prog Pediatr Cardiol
2003;17:53–60.
[47] Koos BJ. Management of uncorrected, palliated,
and repaired cyanotic congenital heart disease in
pregnancy. Prog Pediatr Cardiol 2004;19:25–45.
[48] Siu SC, Sermer M, Colman JM, et al. Prospective
multicenter study of pregnancy outcomes in women
with heart disease. Circulation 2001;104(5):515–21.
[49] Khairy P, Ouyang DW, Fernandes SM, et al. Preg-
nancy outcomes in women with congenital heart dis-
ease. Circulation 2006;113(4):517–24.
[50] Walker E, Malins AF. Anaesthetic management of
aortic coarctation in pregnancy. Int J Obstet Anesth
2004;13(4):266–70.
[51] Moran AM, Geva T. Con: pediatric anesthesiolo-
gists should not be the primary echocardiographers
for pediatric patients undergoing cardiac surgical
procedures. J Cardiothorac Vasc Anesth 2001;
15(3):391–3.
[52] Sangwan S, Au C, Mahajan A. Pro: pediatric anes-
thesiologists should be the primary echocardiog-
raphers for pediatric patients undergoing cardiac
surgical procedures. J Cardiothorac Vasc Anesth
2001;15(3):388–90.
[53] Russell IA, Rouine-Rapp K, Stratmann G, et al.
Congenital heart disease in the adult: a review
with Internet-accessible transesophageal echocar-
diographic images. Anesth Analg 2006;102(3):
694–723.
[54] Andropoulos DB, Stayer SA, Skjonsby BS, et al.
Anesthetic and perioperative outcome of teenagers
 THE ANESTHESIOLOGIST’S ROLE
and adults with congenital heart disease. J Cardio-
thorac Vasc Anesth 2002;16(6):731–6.
[55] Akpek EA, Miller-Hance WC, Stayer SA, et al.
Anesthetic management and outcome of complex
late arterial-switch operations for patients with
transposition of the great arteries and a systemic
right ventricle. J Cardiothorac Vasc Anesth 2005;
19(3):322–8.
[56] Chessa M, Carminati M, Butera G, et al. Early and
late complications associated with transcatheter
occlusion of secundum atrial septal defect. J Am
Coll Cardiol 2002;39(6):1061–5.
[57] Carminati M, Butera G, Chessa M, et al. Transcath-
eter closure of congenital ventricular septal defect
with Amplatzer septal occluders. Am J Cardiol
2005;96(12A):52L–8L.
[58] Gatzoulis MA, Freeman MA, Siu SC, et al. Atrial
arrhythmia after surgical closure of atrial septal
defects in adults. N Engl J Med 1999;340(11):
839–46.
[59] Gatzoulis MA, Hechter S, Webb GD, et al. Surgery
for partial atrioventricular septal defect in the adult.
Ann Thorac Surg 1999;67(2):504–10.
[60] Budts W, Van PN, Gillyns H, et al. Residual pulmo-
nary vasoreactivity to inhaled nitric oxide in patients
with severe obstructive pulmonary hypertension and
Eisenmenger syndrome. Heart 2001;86(5):553–8.
[61] Morton BC, Morch JE, Williams WG. Primary
repair of Fallot’s tetralogy in an older adult:
a 20-year follow-up. Can J Cardiol 1996;12(12):
1268–70.
[62] Presbitero P, Prever SB, Contrafatto I, et al. As orig-
inally published in 1988: results of total correction of
tetralogy of Fallot performed in adults. Updated in
1996. Ann Thorac Surg 1996;61(6):1870–3.
[63] Hughes CF, Lim YC, Cartmill TB, et al. Total intra-
cardiac repair for tetralogy of Fallot in adults. Ann
Thorac Surg 1987;43(6):634–8.
[64] Lake C, Booker P. Pediatric cardiac anesthesia. Phil-
adelphia: Lippincott Williams & Wilkins; 2004.
[65] Oechslin EN, Harrison DA, Harris L, et al. Reoper-
ation in adults with repair of tetralogy of Fallot: in-
dications and outcomes. J Thorac Cardiovasc Surg
1999;118(2):245–51.
585
[66] Gatzoulis MA, Elliott JT, Guru V, et al. Right and
left ventricular systolic function late after repair of
tetralogy of Fallot. Am J Cardiol 2000;86(12):
1352–7.
[67] Dearani JA, Danielson GK. Surgical management
of Ebstein’s anomaly in the adult. Semin Thorac
Cardiovasc Surg 2005;17(2):148–54.
[68] Lerner A, Dinardo JA, Comunale ME. Anesthetic
management for repair of Ebstein’s anomaly. J Car-
diothorac Vasc Anesth 2003;17(2):232–5.
[69] Connelly MS, Liu PP, Williams WG, et al. Congen-
itally corrected transposition of the great arteries in
the adult: functional status and complications.
J Am Coll Cardiol 1996;27(5):1238–43.
[70] Fontan F, Baudet E. Surgical repair of tricuspid
atresia. Thorax 1971;26(3):240–8.
[71] Mavroudis C, Zales VR, Backer CL, et al. Fenes-
trated Fontan with delayed catheter closure: effects
of volume loading and baffle fenestration on cardiac
index and oxygen delivery. Circulation 1992;86(5
Suppl):II85–92.
[72] Kreutzer J, Keane JF, Lock JE, et al. Conversion of
modified Fontan procedure to lateral atrial tunnel
cavopulmonary anastomosis. J Thorac Cardiovasc
Surg 1996;111(6):1169–76.
[73] Mavroudis C, Backer CL, Deal BJ, et al. Fontan
conversion to cavopulmonary connection and ar-
rhythmia circuit cryoablation. J Thorac Cardiovasc
Surg 1998;115(3):547–56.
[74] Mavroudis C, Backer CL, Deal BJ, et al. Total
cavopulmonary conversion and maze procedure
for patients with failure of the Fontan operation.
J Thorac Cardiovasc Surg 2001;122(5):863–71.
[75] van den Bosch AE, Roos-Hesselink JW, Van DR,
et al. Long-term outcome and quality of life in adult
patients after the Fontan operation. Am J Cardiol
2004;93(9):1141–5.
[76] Walker HA, Gatzoulis MA. Prophylactic anticoagu-
lation following the Fontan operation. Heart 2005;
91(7):854–6.
[77] Monagle P, Karl TR. Thromboembolic problems
after the Fontan operation. Semin Thorac Cardio-
vasc Surg Pediatr Cardiol Surg Annu 2002;5:
36–47.
 Cardiol Clin 24 (2006) 587–605

The Team Approach to Pregnancy

and Congenital Heart Disease
Henryk Kafka, MD, FRCPC, FACCa,b,*, Mark R. Johnson, PhD,
MRCP, MRCOGc, Michael A. Gatzoulis, MD, PhD, FACCa
a
Adult Congenital Heart Disease Centre, Royal Brompton Hospital, Sydney Street, London SW3 6NP, UK
b
Division of Cardiology, Queen’s University Cardiovascular Laboratory, Kingston General Hospital,
76 Stewart Street, Kingston, ON, Canada K7L2V7
c
Academic Obstetrics and Gynaecology, Chelsea and Westminster, 369 Fulham Road, London SW10 9NH, UK

The adult cardiologist is being called upon to
see an increasing number of patients who have
congenital heart disease. The incidence of this
most common inborn defect is generally estimated
at 0.8% of all births [1,2]. The advances in pediat-
ric cardiology and surgery have resulted in
a marked improvement in survival. Fifty years
ago, only 25% of infants who had congenital
heart disease (CHD) would survive their first
year and now we expect about 85% to survive
into adulthood [3]. There are no firm figures as
to the real size of this adult population but it is es-
timated at 800,000 adult patients who have CHD
in the United States. With an annual birth rate of
4 million [4], the United States will add another
28,000 infants who have CHD each year, with
about 24,000 of these surviving into adulthood
and joining the ever-growing list of adults who
have congenital heart disease. Survival into adult-
hood certainly does not equate with cure of their
congenital lesions [3,5] and they continue to face
medical, surgical, and psychosocial obstacles [5–8].
Among these, some of the most challenging revolve
Dr. Kafka has been supported through the
Detweiler Fellowship of the Royal College of
Physicians and Surgeons of Canada and through
a grant from the Department of Medicine at Queen’s
University, Canada.
* Corresponding author. Division of Cardiology,
Queen’s University Cardiovascular Laboratory, King-
ston General Hospital, 76 Stewart Street, Kingston,
ON, Canada K7L2V7.
E-mail address: kafkamd@usa.net (H. Kafka).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.009
around sexual function, reproductive issues, and
pregnancy [9–11].
In the face of the improved survival among
those who have congenital cardiac lesions and
with the decreasing incidence of rheumatic heart
disease, CHD is now the most common structural
cardiac problem in women of childbearing age in
North America and industrialized nations [12–14].
Most cardiac patients being seen in relation to
pregnancy, therefore, have CHD and we must un-
derstand the effects of pregnancy on these patients
and the risks faced by the mother and her offspring
to choose the best plan for ongoing management.
Effects of pregnancy on the circulatory system
Important maternal cardiovascular adaptations
start during the first trimester, peak in the late
second and early third trimesters, and largely
resolve by 6 weeks postpartum [12,15–17]. These
adaptations are summarized in Table 1. Several re-
cent reviews are available for more detailed descrip-
tions of these hemodynamic changes [12,14–16].
The increases in stroke volume and heart rate
[17] are preceded by a decrease in systemic vascu-
lar resistance evident early in the first trimester.
This decrease in systemic vascular resistance is
caused by the effects of pregnancy hormones
and other vasoactive substances, including circu-
lating prostaglandins [12,18–20]. The peripheral
vasodilatation is accompanied by an increased
glomerular filtration rate and may be the trigger
for sodium and water retention [18,19], resulting
in the increased circulating volume. There is
cardiology.theclinics.com
588 KAFKA et al

Table 1 estrogen receptors have been found in human

Cardiovascular adaptations during pregnancy aortic tissue. A gradual dilatation of the aorta,
Average change renal, and placental vessels deals with the in-
Parameter during pregnancy creased volume and cardiac output [24–26].
Blood volume \ 35% There is a significant increase in the procoagu-
Cardiac output \ 40%–43% lant activity linked to elevations in factors V, VII,
Stroke volume \ 30% VIII, IX, X, XII, von Willebrand factor, and
Heart rate \ 15%–17% fibrinogen that peaks around term. These changes
Systemic vascular resistance Z 15%–21% help to prevent hemorrhage during endovascular
Mean arterial pressure No significant change trophoblastic invasion and with hemostasis during
Systolic blood pressure Z 3–5 mm Hg delivery but they are associated with an increased
Diastolic blood pressure Z 5–10 mm Hg incidence of thromboembolic complications [27]. It
Central venous pressure No significant change
has been estimated that the risk for a thrombotic
Serum colloid Z 14%
osmotic pressure
event is 6 times higher in pregnancy and 11 times
Hemoglobin Z 2.1 g/dL higher in the puerperium [28].
O2 consumption \ 30% Although the described changes are of benefit in
Right Ventricle \ 18% the setting of a normal maternal circulation, these
(end diastole)a same changes can be challenging for the mother
Right Atriuma \ 19% who has CHD and these challenges rise with the
Left Ventricle \ 6% increasing complexity of the lesions. Patients who
(end diastole)a have shunts (ventricular septal defect [VSD], atrial
Left Atriuma \ 12% septal defect [ASD]) and valvular regurgitation
a
Echocardiographic dimensions. (pulmonary regurgitation, aortic regurgitation)
Adapted and modified from Abbas AE, Lester SJ, generally tolerate this added volume load and in-
Connolly H. Pregnancy and the cardiovascular system. creased cardiac output well. The presence of heart
Int J Card 2005;98:179–89. failure or symptoms before the pregnancy, how-
ever, can increase the risks even in this group. It is
pregnancy-associated enlargement of the cardiac obvious that patients who have severe obstructive
chambers (Table 1), which is more apparent in lesions (aortic stenosis, pulmonary stenosis) and
the right-sided chambers [21]. The peripheral va- those who have severe pulmonary hypertension
sodilatation can be associated with postural symp- cope poorly with the hemodynamic changes of
toms, and caval pressure from the gravid uterus pregnancy because of their limited ability to aug-
may exacerbate hypotension further [22]. ment cardiac output. Those who have poor ventric-
Labor and delivery bring their own cardiovas- ular function may be expected to have similar
cular changes and challenges. The pain and difficulties [13,29]. One result of the plasma volume
anxiety during labor can increase heart rate and expansion is reduced serum colloid osmotic pres-
stroke volume. A further 300 to 500 mL of blood sure and, in consequence, a lowered threshold for
is transfused into the circulation with each uterine pulmonary edema [14].
contraction and these factors combined can result The situation is further complicated for the
in cardiac output increases of as much as 80% patient who has cyanotic heart disease. These
above prepregnancy levels [16]. Use of an epidural patients have chronic erythrocytosis and their
anesthetic can modulate these increases in cardiac own underlying hemostatic disorder which, cou-
output. About 500 mL of blood is lost during a nor- pled with the thrombophilic tendencies of preg-
mal vaginal delivery and a normal cesarean delivery nancy, could be expected to further increase the
is expected to result in the loss of 600 to 1000 mL. thromboembolic risks. Moreover, the drop in
After delivery, there is relief of the caval compres- systemic vascular resistance may predispose to
sion and a further augmentation of maternal circu- increased right-to-left shunting and, consequently,
lating volume by the return of blood from the worsening maternal and fetal hypoxemia [9,12,15].
placental sinusoids [12]. There has been some Marfan syndrome, bicuspid aortic valve [24],
suggestion that this sudden augmentation of vol- and coarctation [3] have all been associated with
ume may lead to elevated levels of atrial natri- an increased risk for aortic dissection linked to
uretic peptide and postpartum diuresis [23]. their underlying aortopathy. This risk is further
Changes in the structure of arteries and veins enhanced during pregnancy as a result of preg-
have been documented during pregnancy, and nancy-related tissue changes and the increased
 PREGNANCY AND CONGENITAL HEART DISEASE
hemodynamic stresses on the aorta, being espe-
cially high in Marfan syndrome when the aortic
root exceeds 4 cm [24]. Aortic root enlargement
has been described in repaired tetralogy of Fallot
[14,30] but the dissection risk in pregnant patients
has not been reported.
In addition to the abnormal response of the
patient who has CHD to the normal cardiovascular
adaptations of pregnancy, there are the complica-
tions that can occur even in normal patients during
pregnancy. Preeclampsia can seriously threaten
a patient who has CHD (especially any form of
pulmonary hypertension) because of the hyperten-
sion, renal dysfunction, and disseminated intravas-
cular coagulation [31]. Urinary tract infections are
common in pregnancy and the fever-induced
tachycardia and vasodilation may be poorly toler-
ated. Arrhythmias during pregnancy [32–35] can
occur in seemingly well individuals and may be re-
lated to the direct electrophysiologic effects of hor-
mones, the changes in autonomic tone, and the
physiologic cardiac dilatation [17,21,36]. Not sur-
prisingly, they are more common during labor
and delivery. Tachyarrhythmias may decompen-
sate the patient who has an obstructive lesion or
poor systemic ventricular function.
Risks
Any discussion about pregnancy needs to start
with an understanding and an assessment of the
associated risks. These risks fall into four cate-
gories, each of which requires careful discussion:
maternal risk during pregnancy, fetal risk, risk for
recurrence of CHD in the offspring, and risk for
early parental loss or disability. It must be empha-
sized, however, that accurate risk assessment may
not be available for every type of adult CHD
situation and that even the most quoted reports are
based on relatively small numbers of patients. An
in-depth review has recently tabulated the wide
range of maternal and fetal risks [37].
Maternal risk during pregnancy
Over the past 80 years there has been a signifi-
cant decrease in the maternal mortality rate from 1
in 250 to 1 in 10,000 [31]. The reported maternal
mortality rates in CHD vary from 1 in 1000 [31]
to as high as 1 in 2 for patients who have Eisen-
menger syndrome. Although congenital disease is
the most common cardiac lesion during pregnancy
in industrialized nations, fewer than one third of
cardiac deaths during pregnancy in the United
589
Kingdom from 1987 to 2002 were attributable to
CHD [38]. Myocardial infarction, aortic dissection,
and cardiomyopathy each accounted for about
20% of the cardiac deaths. During 2000 to 2002,
suicide was the leading cause of maternal death in
the United Kingdom, followed by cardiac disease
and thromboembolism [38].
In general, increasing maternal risk is associ-
ated with anatomic complexity, severity of ob-
structive lesions, poor function of the systemic
ventricle, poor functional class, and the presence
and degree of cyanosis [28]. The CARPREG
group validated their prediction rule by following
599 pregnancies, of which 445 involved congenital
heart defects, and identifying those who had pri-
mary cardiac events (pulmonary edema, stroke,
tachyarrhythmia or bradyarrhythmia requiring
treatment, cardiac arrest, or cardiac death). Thir-
teen percent of these patients suffered a primary
cardiac event. This finding established four reli-
able predictors of these primary cardiac events:
 Prior cardiac event (CHF, transient ischemic
attack, stroke) or symptomatic arrhythmias
requiring treatment
 Baseline New York Heart Association
(NYHA) greater than class II, or cyanosis
 Left heart obstruction (mitral valve area !2 cm2,
aortic valve area !1.5 cm2, or peak left ventri-
cular outflow gradient O30 mm Hg by Doppler
echocardiography)
 Reduced systemic ventricular function (ejec-
tion fraction !40%).
In the absence of any of these predictors, the risk
for a primary cardiac event was 5%, rising to 27%
in the presence of one predictor and to 75% when
there were two or more predictors (there were no
patients who had Eisenmenger syndrome in this
study) [13]. A recent study of 90 pregnancies in
women who had exclusively CHD has validated
the CARPREG index score and has demonstrated
that cardiac risk assessment could be improved
further by adding decreased subpulmonary ventric-
ular systolic function or severe pulmonary regurgi-
tation to the CARPREG risk index [29]. In fact,
mothers who had severe pulmonary regurgitation
or reduced subpulmonary ventricular ejection
fraction were 10 times more likely to have a cardiac
complication. In addition to these predictors,
a thorough knowledge of the patient’s congenital
heart lesion is necessary to deal with the specific
risks of different types of lesions. These can be orga-
nized into low, moderate, and high risk and exam-
ples are presented in Tables 2, 3, and 4 [28].
590 KAFKA
Table 2
Lesions associated with low maternal risk
Lesion Exclude before pregnancy
VSD Pulmonary arterial
hypertension
ASD (unoperated) Pulmonary arterial
hypertension
Ventricular dysfunction
Coarctation (repaired) Recoarctation
Aneurysm formation
at the site
of repair (MRI)
Associated lesion,
such as bicuspid
aortic valve
(þ/ aortic stenosis or
aortic regurgitation),
ascending aortopathy
Systemic hypertension
Ventricular dysfunction
Tetralogy of Fallot Severe right
(repaired) ventricular outflow
tract obstruction
Severe pulmonary
regurgitation
Right ventricular
dysfunction
DiGeorge syndrome
(present in up to 15%
of patients)
From Uebing A, Steer PJ, Yentis SM, et al. Pregnancy and congenital heart disease. BMJ 2006;332:401–6.
To apply these prognostic indicators, one
needs to have completed a thorough assessment
of the patient, including history, physical exami-
nation, echocardiography, and relevant blood
work, if indicated. In some patients it may be
necessary to proceed to exercise testing, MRI, or
cardiac catheterization to make as reliable an
assessment as possible regarding prognosis and
risk for the pregnancy.
Fetal risk
In a pregnancy complicated by preexisting
heart disease, the fetus may be at greater risk for
the usual pregnancy complications of prematurity
and growth restriction but also may be exposed to
the risks of lower maternal oxygen levels, drugs
given to support the maternal cardiovascular
et al
Recommended treatment
during pregnancy and
Potential hazards peripartum
Arrhythmias Antibiotic prophylaxis
Endocarditis (unoperated for unoperated or
or residual VSD) residual VSD
Arrhythmias Thromboprophylaxis if bed
Thromboembolic events rest is required
Consider low-dose
aspirin during pregnancy
Preeclampsia Beta blockers if necessary to
Aortic dissection control systemic blood
Congestive heart failure pressure
Endarteritis Consider elective
cesarean section before
term in case of aortic
aneurysm formation or
uncontrollable systemic
hypertension
Antibiotic prophylaxis
Arrhythmias Consider preterm
Right ventricular failure delivery in the rare
Endocarditis case of right
ventricular failure
Antibiotic prophylaxis
state, and inheritance of a congenital cardiovas-
cular condition. Prematurity carries with it the
potential complications of respiratory distress
syndrome, intracranial hemorrhage, and necrotiz-
ing enterocolitis. When prematurity is combined
with growth restriction these problems are mag-
nified several-fold, and inevitably these pregnan-
cies are complicated by increased rates of stillbirth
and neonatal death [39]. The same risk factors
that predict maternal complications also can pre-
dict fetal and neonatal adverse effects, namely
NYHA class greater than II, cyanosis, or impor-
tant left heart obstruction [13,39]. The risk to
the fetus is heightened further by smoking, mater-
nal age (!20 or O35), and therapy with anticoag-
ulants [13,29]. In general, fetal mortality reports
have ranged from 6% to 24% [37] but can be as
high as 50% in Eisenmenger syndrome [40]. For
 PREGNANCY AND CONGENITAL HEART DISEASE
Table 3
Lesions associated with moderate maternal risk
Exclude before
Lesion pregnancy
Mitral stenosis Severe stenosis
Pulmonary venous
hypertension
Aortic stenosis Severe stenosis
(peak pressure gradient
on Doppler O 80 mm Hg,
ST segment depression,
symptoms)
Left ventricular dysfunction
Systemic right Ventricular dysfunction
ventricle: Severe systemic AV valve
TGA after regurgitation
atrial switch Arrhythmias (brady- and
procedure; tachyarrhythmias)
ccTGA Heart failure (NHYA O II)
Obstruction of venous
pathways after atrial
switch as venous
blood flow significantly
increases during
pregnancy
Cyanotic lesions Ventricular dysfunction
without pulmonary
hypertension
Fontan-type Ventricular dysfunction
circulation Arrhythmias
Heart failure (NYHA O II)
Abbreviations: ACE, angiotensin converting enzyme; AV, atrioventricular, ccTGA, congenitally corrected TGA;
TGA, transposition of the great arteries.
From Uebing A, Steer PJ, Yentis SM, et al. Pregnancy and congenital heart disease. BMJ 2006;332:401–6.
591
Recommended treatment
during pregnancy and
Potential hazards peripartum
Atrial fibrillation Beta blockers
Thromboembolic events Low-dose aspirin
Pulmonary edema Consider bed rest
during the third
trimester with additional
thromboprophylaxis
Antibiotic prophylaxis
Arrhythmias Bed rest during third trimester
Angina with thromboprophylaxis
Endocarditis Consider balloon
Left ventricular aortic valvotomy
failure (for severe symptomatic
valvar stenosis)
or preterm cesarean section
if cardiac decompensation
ensues (cardiopulmonary
bypass carries a 20% risk
for fetal death)
Antibiotic prophylaxis
Right ventricular Regular monitoring
dysfunction of heart rhythm
(potentially persisting Restore sinus
after pregnancy) rhythm in case
Heart failure of atrial flutter
Arrhythmias (DC cardioversion usually
Thromboembolic events effective and safe)
Endocarditis Alter afterload
reduction therapy
(stop ACE inhibitors; con-
sider beta blockers)
Low-dose aspirin (81 mg)
Antibiotic prophylaxis
Hemorrhage (bleeding Consider bed rest
diathesis) and oxygen supplementation
Thromboembolic events to maintain oxygen
Increased cyanosis saturation and promote
Heart failure oxygen tissue delivery
Endocarditis (thromboprophylaxis with
low molecular weight
heparin)
Antibiotic prophylaxis
Heart failure Consider anticoagulation
Arrhythmias with low molecular
Thromboembolic weight heparin and aspirin
complications throughout pregnancy
Endocarditis Maintain sufficient
filling pressures and avoid
dehydration during delivery
Antibiotic prophylaxis
592 KAFKA et al

Table 4
Lesions associated with high maternal risk
Recommended treatment
during pregnancy and
Lesion Exclude before pregnancy Potential hazards peripartum
Marfan syndrome Aortic root Type A dissection Beta blockers
dilatation O4 cm of the aorta in all patients
Elective cesarean
section when aortic
root O45 mm
(w35 weeks of gestation)
Eisenmenger syndrome; Ventricular dysfunction 30%–50% risk Therapeutic termination
other pulmonary arterial Arrhythmias for death related should be offered
hypertension to pregnancy If pregnancy continues close
Arrhythmia cardiovascular
Heart failure monitoring, early bed
Endocarditis for rest, pulmonary
Eisenmenger vasodilator therapy with
syndrome supplemental oxygen
should be considered
Close monitoring
necessary for 10 days
postpartum
From Uebing A, Steer PJ, Yentis SM, et al. Pregnancy and congenital heart disease. BMJ 2006;332:401–6.

the cyanotic patient, hemoglobin level and oxygen
saturation are powerful predictors of fetal out-
come. In women who have cyanotic congenital le-
sions, a hemoglobin of 16 or less was associated
with a live birth rate of 71%, whereas the live
birth rate fell to 8% when hemoglobin was 20 or
higher. In a similar way, there was a 92% live
birth rate for those who had oxygen saturation
of 90 or higher but only 12% for those who had
oxygen saturation of 85 or less [41].
Prematurity is associated not only with an
increased risk for neonatal death but also with
significant long-term morbidity [42,43], including
cerebral palsy and bronchopulmonary dysplasia
with its associated oxygen dependency for the
surviving neonate. Maternal cyanosis, obstructive
lesions, or poor ventricular function can result
in the failure to meet the oxygen and nutrient
demands of the fetus adequately, slowing fetal
growth and sometimes requiring early delivery.
Intrauterine growth restriction has been shown
to have important long-term consequences rang-
ing from neurodevelopment to vascular abnor-
mality, diabetes, and dyslipidemia [44]. By
detecting growth restriction as early as possible,
we can initiate therapy to improve both the mater-
nal situation and the fetus. If that is not possible,
or is unsuccessful, steps need to be taken to plan
the optimal delivery time, including the use of
steroids to mature fetal lungs if the delivery is nec-
essary before 34 weeks. Ultrasound assessments to
monitor fetal growth should be done every 2 to 4
weeks depending on the clinical situation [31].
Risk for recurrence of congenital heart disease
in the offspring
The overall risk for any child being born with
CHD is estimated at 0.8%, whereas the risk for
a structural cardiac lesion in the offspring of
a parent who has CHD ranges from 2% to 50%
(Table 5) with a wide range of these recurrence
risks reported in several studies [28,37,45–49].
The recurrence risk seems to depend on the par-
ent’s specific cardiac defect and on which parent
is affected. There is a general impression that the
risk for recurrence may be higher if the mother
is the affected parent [37,46] but this does not
hold true for all defects or all reports. Of course,
the recurrence rate in autosomal dominant lesions
such as Holt-Oram, or Marfan syndrome is 50%.
Special attention needs to be paid to the clinical
syndromes that are part of the wide spectrum of
22q11 deletions. Once believed to be a rare cause
of DiGeorge syndrome, it is now apparent that
22q11 deletions are far more common [48,50].
They may be detected in up to 50% of children
born with interrupted arch or persistent truncus
 PREGNANCY AND CONGENITAL HEART DISEASE 593

Table 5 loving, and providing financially for the child.

Recurrence risk for congenital heart disease These matters should be considered by a potential
Risk for Recurrence (%) parent who has congenital heart disease. Adults
Mother Father
who have CHD may have difficulty obtaining
CHD defect affected affected health insurance [6,8]. Some patients have an in-
creased risk for premature death and disability
Atrioventricular septal 7.9–13.9 4.3–7.7
or may have to face the possibility of repeated ma-
defect
Aortic stenosis 8.0–13.9 3.8
jor surgery [3,5]. These issues must be discussed
Coarctation 4.1–6.3 3.0–6.0 despite their difficulty. The matter of mortality
Atrial septal defect 4.6–11 1.5–3.6 and the fact that the patient has not been ‘‘cured’’
Ventricular septal defect 6.0–15.6 3.0–3.6 may come up for the first time in these discus-
Pulmonary stenosis 5.3–6.5 2.0–3.5 sions. Although the physician wants to be as frank
Persistent ductus arteriosus 4.1 2.0–2.5 as possible, the data about patient prognosis is
Tetralogy of Fallota 2.0–4.5 1.6 constantly changing with improvements and inno-
Marfan syndrome 50 50 vations in cardiac care and all involved must un-
22q11 deletion syndromes 50 50 derstand that discussions about prognosis can
a
Assuming that the parent does not have 22q11 only involve generalizations. The patient may
deletion. benefit from the skill and counseling of a psychol-
Data from Refs. [28,37,45–49]. ogist or psychiatrist or advanced practice nurse
with experience in such situations.

arteriosus and 10% to 15% of children who have
tetralogy of Fallot or pulmonary atresia. Associ-
ated defects, such as facial dysmorphism, nasal
speech, learning difficulties, cleft palate, hearing
impairment, psychiatric illness, or talipes equino-
varus, are clues to possible 22q11 deletions, but
there are individuals with an isolated cardiac or
vascular defect who have a 22q11 deletion and
run a 50% risk for passing that deletion to an off-
spring [48,50]. The contribution of a clinical ge-
neticist in these complex cases is invaluable.
Because of the greater risk for CHD in the
offspring of a parent who has congenital heart
disease, consideration should be given to offering
ultrasound scans to screen for congenital lesions
[28,31]. A normal fetal nuchal translucency mea-
surement at 12 to 13 weeks has a negative predic-
tive value of 99.9% for fetal heart disease and can
be used to reassure a parent who has congenital
heart disease. For those who have a strong family
history of congenital heart disease, a specialist fe-
tal echocardiogram should be performed at 14 to
16 weeks. This scan can detect moderate to severe
lesions but cannot rule out all forms of CHD be-
cause the fetal heart is so small at this stage. This
test needs to be followed up with a further echo-
cardiographic scan at 18 to 22 weeks [31,51].
Risk for early parental loss or disability
Becoming a parent carries a serious responsi-
bility to care for the child, including nurturing,
Care of the patient (Appendix 1)
Prepregnancy care and counseling
Individuals who have CHD need to be coun-
seled about reproductive issues as they reach
adolescence. This need cannot be overemphasized
because up to 80% of pregnancies in teenagers are
unplanned [52]. We know that an unplanned preg-
nancy can be a life-altering event for any teenager,
but it can be life-threatening for those who have
CHD. Counseling and teaching thus should take
place long before the patient reaches the adult
congenital heart clinic [6,53] and the pediatric car-
diologist needs to ensure that such counseling has
occurred. It has been reported that some parents
of adolescent patients tend to regard their matur-
ing offspring as asexual [6] and avoid questions
about sexuality or reproduction. As a conse-
quence, these adolescents harbor misconceptions
and fears about their sexual function and ability
to conceive, and the dangers associated with preg-
nancy. They have reported fears about death dur-
ing sexual activity and fears about transmitting
their cardiac defect to offspring [6]. Because
some CHD may be accompanied by intellectual
impairment, these counseling sessions may be
challenging and require judgment and sensitivity
in addressing all the relevant issues. Frank discus-
sions about sexual activity and reproductive issues
may be hampered by the presence of a parent or
guardian. This same difficulty also can be present
in the adult clinic because these young women and
594 KAFKA
men have been patients since birth and often con-
tinue to be accompanied by their parents during
clinic visits. Patients, both adults and adolescents,
should be given the opportunity to meet with their
caregivers for at least part of their visits without
their parents so that they can ask questions, an-
swer questions, and receive information not only
about sexual functioning but also about their
heart condition and prognosis.
When the patient arrives in the adult clinic, one
cannot assume that she (or he) has a full un-
derstanding of the reproductive issues. The pa-
tient may be reluctant to receive counseling
because there are no plans to start a family in
the near future. Counseling and education should
be offered to everyone so that patients, male and
female, understand the general risks inherent for
themselves and their offspring and the more
specific risks concerning their particular congeni-
tal defect and repair. There needs to be careful
and sensitive discussion about the residual defects
or expected deterioration that would increase
pregnancy risks or significantly limit life expec-
tancy. This discussion is done not to impose the
physician’s opinions onto the patient but to pro-
vide the information and the teaching necessary to
allow the patient to make the most appropriate
decision about future parenthood and contracep-
tion. These discussions should not be limited to
female patients because the issues of risk for
recurrence and risk for premature parental dis-
ability or loss apply to fathers also.
To provide the best information about risks to
the patient, it is necessary to have a thorough
understanding of the congenital defect and the
outcomes of the previous surgical and catheter
interventions in that patient. This testing may be
more comprehensive than is routine to give a clear
and informed picture. Furthermore, to decrease
risk during pregnancy, one may even consider
surgical or catheter interventions that might
otherwise have been postponed to a later date.
Such procedures can be justified clearly in the
preparations for a planned pregnancy because
catheter and surgical interventions during preg-
nancy carry their own special risks [28,54] and cer-
tain residual defects markedly increase maternal
and fetal risks.
Contraception
Even in patients who do not have CHD any
discussion about contraception must deal with the
issues of compliance, reliability, and safety. These
et al
discussions need to involve an obstetrician who is
familiar with the special challenges of contracep-
tion in patients who have CHD because there is
no ideal contraceptive method. The consequences
of an unplanned pregnancy in this group of
patients can be grave, and despite recommenda-
tions, contraceptive teaching is not always suc-
cessful, with one study reporting that 6 of 35
women who had CHD had an unplanned preg-
nancy [55]. A complete recent review of contra-
ception for patients who have CHD is available
[31].
Because of their unacceptable failure rates,
natural methods of birth control cannot be
recommended for patients in the high and mod-
erate maternal risk groups (see Tables 3 and 4)
Barrier methods (condom and diaphragm) have
failure rates of 2% to 50% depending on the com-
pliance with their use. Spermicidal jelly or cream
is recommended to be used in conjunction with
the condom or diaphragm. These barrier methods
may end up being the best choice for women who
face unacceptably high risks with the other tech-
niques. In the case of failure of contraception in
the high maternal risk group, early termination
of pregnancy should be given due consideration.
The oral contraceptive pill remains a popular
technique among young women. There are two
categories of oral contraceptive: the combined pill
with estrogen and progestogen, and the low-dose
pill (or mini-pill) that contains a very low dose of
progestogen. The combined pill is significantly
more effective, with a failure rate of less than 1 per
1000 women per year. Because it is associated with
a fourfold increased risk for thrombosis [31], how-
ever, it cannot be recommended for the woman
who already has a defect that predisposes to
thrombosis (cyanosis, atrial arrhythmias, valvular
prostheses, Fontan-type circulations). Conversely,
it may be a good choice for the lower-risk patient
because of its effectiveness. In comparison, the
low-dose progestogen pill does not predispose to
thrombosis but is not as effective, with a failure
rate of 1 in 200 per year. Furthermore, its effec-
tiveness depends on close compliance to a rigid
daily regimen and therefore it is not a good choice
for most adolescents [52]. In those cases, an alter-
native hormonal approach may involve the use of
an injectable progestogen.
Intrauterine devices generally had not been
recommended for these patients because of their
previous association with menorrhagia, increased
risk for pelvic infections, and bacteremias [6,16].
A newer generation of a progestogen-impregnated
 PREGNANCY AND CONGENITAL HEART DISEASE
device actually reduces menstrual bleeding (amen-
orrhea is common) and has a low risk for pelvic
infection or ectopic pregnancy [28]. The progesto-
gen coating carries no increased risk for thrombo-
embolism and these coils can be left in place for
five years, thus eliminating any compliance issues
[31]. The actual process of coil insertion, regard-
less of type, may carry a risk for endocarditis
and requires antibiotic prophylaxis against genito-
urinary organisms. Occasionally the insertion pro-
cedure may be complicated by a vasovagal
reaction that could be poorly tolerated by an indi-
vidual who has unstable hemodynamics [16]. It
has been recommended that coil insertions for
such women be done in the operating room with
the anesthesia staff present [31].
Sterilization can be considered for the patient
who has decided to avoid pregnancy permanently
or who has decided not to have any more children.
The procedure is done by laparoscopy and carries
an anesthetic and surgical risk that can be sub-
stantial for the patient who has complex anatomy
or Eisenmenger syndrome. The risk for pregnancy
after sterilization has been considered to be
between 0.2% and 1% per year [31]. Occasionally,
a young patient who has a fear of pregnancy may
request sterilization, often with urging from her
family. Such requests should be met with careful
explanation and exploration of options before
proceeding [6].
Vasectomy is a low-risk option for the male
patient who has decided not to have offspring and
it can be considered for the male partner of
a woman who has congenital heart disease. The
couple may decide against vasectomy if the
woman’s life expectancy is likely to be shorter
and the man may wish to have children with
a future partner.
Termination
When contraception fails, a decision needs to be
made quickly about whether to carry on with the
pregnancy or to proceed to termination because the
risks of termination, although small, do increase
with advancing gestational age. In general, one can
state that termination of pregnancy is about half as
dangerous as continuation with the pregnancy [31].
Dilation and suction curettage under local anes-
thetic carries a low complication rate as a first-
trimester procedure, but for those at moderate
and high risk it should be performed in a hospital
operating room rather than in an outpatient setting
[52]. Medical abortion with oral antiprogesterones
595
and vaginally administered prostaglandins is
effective in the first 7 weeks but again must be
performed in hospital because there are several
unpredictable effects, such as systemic vasodilata-
tion and heavy bleeding, that may contribute to
hypotension and destabilize the patient. Certainly
medical abortion is not to be considered for pa-
tients who have Eisenmenger syndrome (or other
types of pulmonary hypertension) or for those
who have cyanosis.
The decision to terminate may be made for
personal or medical reasons. An unplanned preg-
nancy in a moderate- (see Table 3) or high-risk pa-
tient (see Table 4) could be considered a medical
indication for termination. If the patient wishes
to continue with the pregnancy, she should be re-
ferred to a high-risk multidisciplinary tertiary care
center clinic.
Prenatal care
With thoughtful investigation and assessment,
a plan for prenatal care should be established
before the patient becomes pregnant. If that has
not occurred, it is important to generate such
a plan as soon as possible in the pregnancy. In this
way, the criteria previously described can be used
to rate the risk of the patient. Low-risk patients
(see Table 2) can receive their prenatal care from
their usual specialists, with ongoing advice avail-
able from the high-risk pregnancy team at the ter-
tiary center. Patients whose risk profile is
moderate and high (see Tables 3 and 4) should
be referred to the team with the plan that labor
and delivery will occur at that tertiary center.
Like other prenatal clinic visits, these focus on
monitoring fetal growth and development but also
include a CHD screening ultrasound and a close
assessment of the mother’s cardiovascular status
at each visit, looking for symptoms and signs of
heart failure, arrhythmia, or endocarditis. Such
complications may be precipitated by problems,
such as anemia or urinary tract infection, which,
although minor in a healthy woman, are of greater
significance in a woman who has preexisting heart
disease. Preeclampsia can be particularly disas-
trous in the setting of complex congenital defects
and especially so for patients who have Eisen-
menger syndrome and it must be specifically
sought out. A planned program for prenatal visits
has been described by Steer [31] and its intensity
depends on the assessed maternal risk. For mod-
erate- to high-risk patients, frequent visits are rec-
ommendeddevery 2 weeks until 24 weeks and
596 KAFKA et al

then weekly thereafterdalong with special ante-
natal records designed for the care of pregnant
women who have heart disease (see Fig. 1).
Planned echocardiography is necessary for pa-
tients at risk for aortic root dilatation and should
be used liberally for any patient who has a new
murmur or suspicion of clinical deterioration. Ra-
diographs should be avoided if at all possible, and
all the usual precautions need to be taken if radio-
logic examination is necessary [56]. Cardiovascu-
lar MRI generally is accepted to be safe in the
setting of pregnancy [57]. There is no human evi-
dence to suggest risk to the fetus during the first
trimester but until the absolute safety of MRI
during the susceptible period can be established,
it is best to avoid scans in the first trimester except
situations in which the benefits outweigh the po-
tential risks [57].
Some patients require hospitalization during
the third trimester for treatment of complications,
for bed rest or closer monitoring, and for oxygen
therapy in cyanotic patients [28]. Because of the
thrombophilia of pregnancy, it is important that
these pregnant patients on bed rest receive low
molecular weight heparin (LMWH) and anti-
thromboembolism stockings.
There is no clear consensus with regard to
endocarditis prophylaxis at the time of delivery.

Fig. 1. Example of antenatal record for women who has cardiac disease. This record facilitates ongoing documentation
of findings and progress during the antenatal visits. The observations at each visit are cardiac and obstetric, designed to
detect any maternal or fetal problems at an early stage. The cardiac diagnosis (lesion) needs to be as specific as possible.
The form provides space for notes about fetal cardiac and anomaly ultrasound examinations. The plan for delivery (elec-
tive cesarean section or trial of labor) should be noted. 5ths palp, amount of fetal head palpable; appt, appointment; BP,
blood pressure; EDD, estimated date of delivery; FH, fetal heart; S/B, seen by; SFH, symphysis/fundal height; SOB,
short of breath. (Courtesy of Philip Steer and the High Risk Obstetric Team, Chelsea and Westminster Hospital, London
UK.)
 PREGNANCY AND CONGENITAL HEART DISEASE
The current American Heart Association guide-
lines do not call for routine prophylaxis for uncom-
plicated delivery. Nevertheless, because of the
devastating consequences of endocarditis in this
setting, some centers routinely give IV antibiotics
to patients who have high-risk lesions [16], whereas
others use prophylaxis selectively for the patient
who has any form of instrumented vaginal delivery
or cesarean section [31].
Thromboembolism is six times more common
during pregnancy [28], even for women who do not
have congenital heart disease. There is an even
greater risk for the cyanotic patient and the patient
who has a Fontan circulation. The patient who has
a left-to-right shunt faces a risk for paradoxic em-
bolism even in the nonpregnant state and that risk
is probably enhanced during pregnancy. Low-
dose aspirin [22,58] can be used safely for these pa-
tients until the 36th week of pregnancy. A major
challenge is the management of chronic anticoagu-
lation throughout pregnancy, especially in the
patient who has a prosthetic valve [9,59,60]. Warfa-
rin crosses the placenta, has teratogenic effects early
in pregnancy, and poses increased risks of fetal
hemorrhage during pregnancy, whereas heparin
does not cross the placenta and is safer for the fetus
but is associated with more maternal thromboem-
bolism [59,60]. The American College of Chest Phy-
sicians (ACCP) Conference on Antithrombotic
Therapy has recommended one of three regimens,
namely, using unfractionated heparin (UFH),
LMWH, or a combination of warfarin and heparin
(Box 1) for pregnant patients. There is a fourth op-
tion of continuing warfarin throughout pregnancy
until the 36th week in the high-risk valve patient
and then using heparin until delivery [28], but this
approach is rarely used in North America because
of concerns about warfarin embryopathy and fetal
hemorrhage [60]. For patients at high risk because
of prosthetic heart valves, one may consider the ad-
dition of low-dose aspirin, despite the fact that
there are no data concerning this combined use in
pregnancy [60].
Atrial arrhythmias occur commonly during
pregnancy and may not be well tolerated if
associated with decreases in cardiac output [37].
Antiarrhythmic drugs should be avoided if at all
possible; however, adenosine, digoxin, procaina-
mide, flecainide, quinidine, and calcium channel
blockers generally are accepted as safe [28,58].
There had been concerns about the use of beta
blockers because of a past association with intra-
uterine growth restriction. Although this risk is
now considered small, close monitoring for
597
Box 1. Recommended regimens for the
management of anticoagulation through
pregnancy from the Seventh ACCP
Conference on Antithrombotic Therapy
(2004)
Each of these is a Grade 1C
recommendation; that is, the risk-to-
benefit ratio is clearly favorable but it
is based largely on observational
studies and should be considered to
be an intermediate-strength
recommendation that may change
when stronger evidence is available.
OR
Adjusted-dose, twice-daily LMWH
throughout pregnancy in doses
adjusted either to keep a 4-hour
postinjection anti-Xa heparin level at
approximately 1.0 to 1.2 U/mL
(preferable) or according to weight
OR
Aggressive adjusted-dose UFH
throughout pregnancy (ie,
administered SC every 12 hours in
doses adjusted to keep the
mid-interval activated partial
thromboplastin time at least twice
control or to attain an anti-Xa heparin
level of 0.35 to 0.70 U/mL)
OR
UFH or LMWH (as above) until the 13th
week, change to warfarin until the
middle of the third trimester, and then
restart UFH or LMWH.
Abbreviations: LMWH, low molecular
weight heparin; SC, subcutaneously; UFH,
unfractionated heparin. From Bates SM, Greer
IA, Hirsh J, et al. Use of antithrombotic agents
during pregnancy. Chest 2004;126:627S–44S.
growth restriction is recommended. Lidocaine or
sotalol can be used for ventricular tachycardia
but amiodarone should be avoided if at all possi-
ble and used only under special circumstances
[12,37,58,61]. Direct current (DC) cardioversion
is considered safe in the pregnant patient [9,15].
598 KAFKA
It is common for pregnant women who do not
have heart disease to complain of fatigue, dysp-
nea, or pedal edema and it may be challenging to
distinguish the signs and symptoms of a normal
pregnancy from those of early heart failure [62].
Furthermore, even patients who had few symp-
toms before pregnancy can develop heart failure
during the pregnancy. Pulmonary edema occurred
in 42 of 599 pregnancies and 34 of these women
had been in class I or II before the pregnancy
[13]. Heart failure is associated with restricted fe-
tal growth and prematurity and may require ad-
mission for optimization of medical treatment.
Diuretics are relatively safe in pregnancy but an-
giotensin-converting enzyme inhibitors and angio-
tensin-receptor blockers are contraindicated
[28,37,58]. Nitrates are safe and can be used
with hydralazine for combined preload and after-
load reduction in the pregnant patient [58,63,64].
Spironolactone is not recommended because of
a potential to feminize the male fetus. Amiloride
has been considered generally safe in pregnancy
[28,58] but there have been reports of teratogene-
sis in rats [65]. Amiloride should be used only
if potassium supplementation has not been
effective.
There are patients who continue to deteriorate
clinically despite optimal medical therapy. De-
pending on the stage of pregnancy, termination or
early delivery may be considered. If termination
or delivery are not options, then careful assess-
ment needs to be performed to decide whether the
patient may benefit from a surgical or catheter
intervention. If surgery is considered, it is best to
avoid cardiopulmonary bypass because of the
significant associated fetal mortality. If cardiopul-
monary bypass is necessary, the use of high flow/
high pressure cardiopulmonary bypass and the
avoidance of hypothermia are recommended.
Catheter interventions may carry increased long-
term radiation-associated risks to the child but
have a low fetal mortality rate [54]. The most
common interventions during pregnancy in pa-
tients who have CHD involve pulmonary or aortic
valve stenosis. Fortunately such interventions
rarely are required.
Labor and delivery
Because pregnant women can present at any
time, day or night, with labor or urgent compli-
cations, there is a need for constant availability of
staff and facilities. This availability is even more
necessary for the medium- to high-risk CHD
et al
patient. Any tertiary care center that is providing
a high-risk joint cardiac obstetric service must be
capable of providing full access to the proper
facilities and the specialized labor and delivery
team at all hours. This team would include the
obstetrician, the anesthetist, the neonatologist, the
cardiologist, and the intensivist, all of whom
should be experienced in the management of the
pregnant patient who has CHD.
Labor and delivery plans are made during the
course of the clinic visits in consultation with
the patient. There must be clear documentation of
the plan with copies available to the team at all
times, and to her local hospital if she lives some
distance away. It is advisable to provide copies of
the notes and delivery plans (Fig. 2) to the patient
so that they are immediately available on her
arrival. The process of planning and the issues
surrounding labor and delivery have been de-
scribed thoroughly in a recent review [31].
Previously, elective cesarean section under gen-
eral anesthesia had been the chosen mode of
delivery for many women who had congenital heart
disease, in the belief that this would provide the
most controlled environment for the delivery with
the planned presence of the experienced team
members and would avoid the stresses associated
with pushing during the second stage of labor.
Elective cesarean section, however, increases the
risk for hemorrhage, thrombosis, and infection
[31], and carries about 1.2 to 1.4 times the risk of
a vaginal delivery. The stresses of labor can now
be managed with the use of low-dose slow incre-
mental epidural anesthesia, and with the availabil-
ity of the high-risk team at all times there are no
real advantages to elective cesarean section in these
patients, except for obstetric indications.
Spontaneous labor is preferred to induced
labor and it may be advisable to admit some
patients who live a distance away to the tertiary
center to await spontaneous delivery. Occasion-
ally, it may be necessary to induce labor for
obstetric reasons or in the face of cardiovascular
decompensation [28]. With effective epidural anes-
thesia, uterine contractions usually have no evi-
dent adverse hemodynamic effects. The second
stage of labor requires pushing and can have det-
rimental effects if prolonged. Part of the delivery
plan, therefore, should establish how long the pa-
tient can push without significant hemodynamic
compromise (see Fig. 2). Once that limit has
been reached, delivery can be assisted by forceps
or ventouse (vacuum) extraction [31]. After deliv-
ery, relief of caval obstruction increases venous
 PREGNANCY AND CONGENITAL HEART DISEASE 599

Fig. 2. Example of a delivery management plan for patients who have cardiac disease. The cardiac diagnosis needs to be
as specific as possible. The form allows the physician to choose whether the cardiac team needs to be contacted imme-
diately on admission and whether the patient will be admitted electively at a certain gestational age (number of weeks).
A decision will have been made whether to proceed with elective cesarean section or to attempt a trial of vaginal delivery.
For each situation there are several choices regarding anesthetic and monitoring. For vaginal delivery, a decision is made
whether to allow any pushing or to provide a maximum time for pushing (minutes), after which assisted delivery is un-
dertaken. Decisions about prophylactic antibiotics and which medications are to be continued are indicated on the form.
Any postdelivery medication plans should be documented also. Plans for stay in the high-dependency unit (step-down
unit) or intensive care unit will also be noted. BP, blood pressure; CCT, controlled cord traction; CVP, central venous
pressure; EKG, electrocardiogram; HDU, high-dependency unit; LMW, low molecular weight heparin; SaO2, oxygen
saturations; TEDS, thromboembolic deterrent stockings. (Courtesy of Philip Steer and the High Risk Obstetric
Team, Chelsea and Westminster Hospital, London UK.)
600 KAFKA
return and this is augmented by the transfusion of
blood previously in the maternal placental bed.
This return may overwhelm the patient’s ability
to cope with the extra volume load. Conversely,
failure of uterine retraction may lead to uterine
hemorrhage and a loss of volume. Oxytocic drugs
can have marked hemodynamic effects and should
be given cautiously at the lowest effective dosage,
with close attention to the uterine response to
avoid excessive blood loss.
A key to successful labor and delivery is the
ongoing assessment of the fetal status and the
mother’s hemodynamic situation. Continuous
fetal monitoring is necessary in these cases.
Maternal monitoring includes continuous electro-
cardiogram and pulse oximetry and measurement
of blood pressure. Intra-arterial monitoring has
the advantage of communicating rapid blood
pressure changes [28,31] and is recommended in
patients who have aortopathy (eg, Marfan, coarc-
tation). A central venous line can be of value in
the high-risk or decompensating patient but
must be placed by an experienced operator with
knowledge of the patient’s complex anatomy.
An in-line filter (air filter or bubble trap) for the
intravenous line is recommended for any patient
who has right-to-left shunting or a significant
potential for right-to-left shunting.
Thrombotic risk is enhanced during the puer-
perium with most thromboembolic deaths occur-
ring postpartum [38]. Prophylactic LMWH has
been recommended for all patients in the moderate-
and high-risk groups, selected patients in the lower-
risk group, and all those having cesarean section
[28,38]. Anticoagulation with heparin can be re-
started safely within 6 to 12 hours after delivery
and the mother can safely breastfeed. Warfarin
usually is resumed in 48 hours but may need to be
delayed in individuals at increased risk for postpar-
tum hemorrhage. Although digoxin and flecainide
do not enter breast milk in any significant amounts,
beta blockers have been reported occasionally to
cause bradycardia in the breastfed infant and
should be used with careful monitoring of the infant
[31,58,61].
There may be patient decompensation during
any of these stages, including the puerperium, and
careful maternal monitoring must continue
throughout. In the event of clinical deterioration,
there must be rapid access to the cardiologist and
to an intensivist who understands the patient’s
complex underlying cardiac physiology. Ideally
the intensivist would have been informed of the
patient’s admission to the labor room and would
et al
have had an opportunity to review the patient, the
notes, and the plan of therapy.
Summary
There can be significant risks for mother and
offspring in the setting of parental congenital
heart disease. The key to a happy outcome is
early assessment and counseling and advanced
planning through specialized centers and teams.
A well-structured team should be able not only to
direct the management of the pregnancy but also
to provide guidance to the patient about issues
relating to prognosis, sexual activity, contracep-
tion, genetic risks, and parental responsibility.
Acknowledgments
The authors thank Dr. Alan Magee for his
expert review of the manuscript and advice on the
pediatric challenges in these patients and Pro-
fessor Philip Steer and the High Risk Obstetric
Team at the Chelsea and Westminster Hospital
for permission to reproduce their forms.
Appendix 1
Organization of a multidisciplinary team for
pregnancy and reproductive issues in the patient
who has congenital heart disease
The team approach
The preceding sections have outlined the issues
and management challenges in providing the best
care for the CHD patient during her or his
reproductive years. Neither the cardiologist nor
the obstetrician alone has the training, experience,
or expertise to deal with all the necessary aspects
involved in the reproductive issues for these
patients. A collaborative team approach is re-
quired and the members of the team change as the
needs of the patient change. Such teams can be
referred to as high-risk pregnancy teams but that
term can be misleading and it may result in some
patients not being referred because they are not
pregnant, are not actively planning pregnancy, or
are men. Any suggested alternative name (multi-
disciplinary pregnancy or reproductive services
clinic) tends to be just as awkward, however, and
no less intimidating to the patient. The name of
the clinic matters less than its composition and the
training and expertise of its members. It may seem
inappropriate to include a pediatric clinic in a
discussion of the management of adult patients
 PREGNANCY AND CONGENITAL HEART DISEASE
who have congenital heart disease, but the issues
of reproductive health and pregnancy can be con-
sidered adult issues, regardless of whether they
are addressed at the age of 14 or 18.
Team 1: Pediatric reproductive issues team
The patient who has CHD enters the repro-
ductive period of her (or his) life long before
transfer to the adult clinic. This means that the
first assessment of, and discussion about, repro-
ductive issues should take place in a pediatric
setting [6]. The pediatric cardiologist has had
years of ongoing cardiac assessment of this patient
and usually has already discussed adolescent is-
sues, such as body piercing and alcohol and to-
bacco use. The pediatric cardiologist is therefore
in the best situation to discuss the issues of diag-
nosis and prognosis with the patient, and to en-
sure that the patient has had the opportunity to
ask questions. The patient may feel more comfort-
able speaking with the advanced practice nurse as-
signed to the clinic or may benefit from meeting
with a psychologist or psychiatrist to help deal
with some of the difficult issues that may arise
around prognosis. Female patients should be
seen by an obstetrician (or by the family physi-
cian, pediatrician, or counselor) for an assessment
and counseling about sexual activity and contra-
ception. This topic may be awkward to broach
with the patient and the family. In this type of sit-
uation, it may be best to inform the family and the
patient that in the early teen years it is important
to deal with issues of reproductive and gyneco-
logic health for all patients, not just those who
have increased cardiac risk. In this way, the refer-
ral is merely part of the usual care plan for all pa-
tients. The referral could be to the obstetrician
who is already involved with the adult pregnancy
clinic because he or she can discuss confidently the
risks of future pregnancies based on the assess-
ment and on the information from the pediatric
cardiologist. Male patients, especially those who
have heritable disorders, also would benefit from
counseling about sexual activity and contracep-
tion. Depending on the nature of the congenital
heart lesion and the patient’s maturity and inter-
est, he or she could be referred for an assessment
and counseling by a clinical geneticist.
Team 2: Adult reproductive issues team
Once the patient has been referred to an adult
cardiologist, no assumption should be made that
any of the assessments or counseling recommen-
ded above has taken place [53]. It is necessary to
601
determine what the patient knows, and, if the pa-
tient is agreeable, to arrange a referral to the adult
reproductive issues team. This team should consist
of a cardiologist with training and expertise in
adult CHD and in the management of the preg-
nant patient, and an obstetrician with special inter-
est and experience in the care of pregnant patients
who have congenital heart disease. Ideally, the pa-
tient would have already seen the obstetrician for
assessment but it has been reported that many
women, even sexually active women, are not re-
ferred until their late 20s, 30s, or even after they
become pregnant [6]. These two specialists make
up the core of the team and can be supplemented
when necessary with a clinical geneticist having ex-
perience in cardiovascular genetics and counseling,
and a psychologist or psychiatrist to help the pa-
tient deal with some of the difficult aspects of his
or her current clinical situation and prognosis. In
addition, the advanced practice nurse assigned to
the clinic can be a source of comfort and counsel-
ing for the patient. The cardiologist’s assessment is
based on information from the pediatric cardiolo-
gist and the referring cardiologist and is supple-
mented by a current examination and further
investigation, if necessary. In this way, a risk pro-
file can be established and decisions made about
the risks of future pregnancy and the most appro-
priate form of contraception. Infertility issues also
can be discussed and options presented. After the
initial assessment and counseling are complete,
the patient may return to her own clinic, to be re-
ferred back whenever necessary.
Team 3: High-risk pregnancy team
Once the patient is pregnant, she needs a re-
ferral to the high-risk pregnancy team. Ideally, she
would have been assessed by Team 2 and would
already have a well-documented risk profile. If
not, the first step is to determine her risk status by
whatever investigations can be carried out safely
in her current stage of pregnancy. The second step
is to determine if there are any medical reasons
not to proceed with the pregnancy and to counsel
the patient regarding that (keeping in mind that it is
her decision whether to continue with the preg-
nancy), and the third step is to draw up the program
for prenatal care. In addition to the cardiologist
and the obstetrician, the anesthetist is an im-
portant member of this team and will want to
assess the risks for epidural and for general
anesthesia and explain all possible procedures
and risks to the patient. Documentation of the
prenatal plan and the plan for labor and
602 KAFKA
delivery is crucial and it is essential that there is
open access to this plan at all times so that it is
available whenever the patient presents, or to
advise another center should she present there.
Team 4: High-risk labor and delivery team
In addition to the cardiologist, obstetrician,
and anesthetist, the team must include a neonatol-
ogist and an intensivist. Access to a pediatric
cardiologist is advisable, especially if a screening
ultrasound has revealed a cardiac defect. In the
case of an autosomal dominant lesion in the
parent, a referral for a pediatric cardiology
assessment is prudent, even in the absence of
a fetal cardiac defect. The intensivist should be
alerted early in the course of labor and should
have an understanding of the patient’s cardiac
defect and how her altered physiology and hemo-
dynamics may complicate any intensive care unit
stay. Because the patient may present at any time
with labor or complications, team members must
be available at all times with complete access to the
notes and plans that have been so carefully laid out
in the clinic. The goal is to have an uneventful
delivery with most contingencies planned for, in the
presence of experienced specialists whose expertise
allows them to react quickly in the case of
complications [9,31].
Frequently asked questions
It is difficult to arrange to have all these special-
ists in one clinic at any time. Do they all have to see
the patient at the same time?
Ideally, the patient should see all the team
members at her first assessment so that she gets
a real understanding of their roles and interac-
tions and the team members can freely discuss the
situation among themselves and with the patient.
It is acceptable to have the team members see the
patient in separate locations and at separate times
but the need to have open communication and full
access to notes and plans cannot be overempha-
sized. Subsequent visits (eg, prenatal monitoring)
do not need to involve all team members unless
there have been concerns.
Which patient needs to be referred to these
clinics?
If the patient’s cardiologist believes that she is in
the low-risk group (see Table 2), she does not neces-
sarily need to be referred to the clinic. It may be pru-
dent, however, to refer all patients for an initial
assessment and risk profile by the experienced spe-
cialists in the clinic. They then can put together
a plan for the patient to take back with her.
et al
Furthermore, they will be available for advice and
consultation at a distance should there be any prob-
lems. For patients in the moderate- and high-risk
group, arrangements will be made for prenatal
visits and delivery at the clinic center.
The high-risk clinic is a long distance away from
the patient’s home. Would it not be more convenient
for the patient to be seen and delivered locally?
It would be more convenient but in the case of
moderate- and high-risk patients it cannot be
recommended. The benefits to mother and child
of the facilities and staff at the tertiary center’s high-
risk clinic outweigh any inconvenience because of
distance. Lack of insurance and financial support
can be a problem, especially in the United States,
and pose a dilemma to the patient and her family. In
those cases, close liaison with the local specialists on
a regular basis may replace the need for some of the
prenatal visits but one cannot overemphasize the
need to be delivered at the high-risk center.
What if the patient goes into premature labor?
There should already be a plan for dealing with
the situation of premature labor for any patient
who lives some distance away. Depending on the
progress of labor and the patient’s cardiac situa-
tion, urgent transfer may be possible, but early
communication with the high-risk clinic is impor-
tant. The patient should have copies of her notes
and plan with her because that facilitates the close
communication with the high-risk clinic special-
ists, that is crucial for the management of this
situation.
The patient has presented pregnant and wishes
to have a termination. Can that be carried out
locally?
Early first trimester termination usually can be
performed with minimal complications but the
factors that may make a patient moderate- to
high-risk for pregnancy carry the same types of
risks for termination. Patients deemed to be low
risk can have the procedure locally but it should
be performed in a hospital operating room.
Moderate- and high-risk patients should obtain
a prompt referral to have the procedure at the
tertiary center. Certainly any patient who requires
termination on medical grounds should be re-
ferred urgently to the high-risk center.
The patient wishes to have a tubal ligation. Can
that be carried out locally?
If the patient has already been seen by the team
for a previous assessment and has been considered
low risk, it would be appropriate to undertake
tubal ligation locally. If there are any doubts, she
needs to be referred to the clinic for a risk
 PREGNANCY AND CONGENITAL HEART DISEASE
assessment and plan, in light of the hemodynamic
perturbations that may occur during laparoscopy.
Similarly, it is important to discuss referral to the
clinic for consideration of IUD placement in
moderate- and high-risk patients.
What kind of counseling are these patients get-
ting at the high-risk clinic?
Depending on their needs, the patients, male
and female, are counseled on sexual functioning
and contraception. After full assessment, the
patients are counseled on their parental and
pregnancy risks. This counseling may involve the
cardiologist, the obstetrician, the clinical geneti-
cist, the psychologist or psychiatrist, or the
advanced practice nurse attached to the clinic.
Part of any counseling includes some discussion of
patient prognosis. Although there are data for
several lesions and clinical situations, new thera-
pies and changing therapies can make the de-
termination of prognosis difficult [5,66]. Any such
discussion should be prepared carefully before-
hand and presented in a frank but sensitive
manner.
The insurance company refuses to pay for any
visits to the high-risk clinic in another city because
they claim that the patient can be seen and delivered
locally. How can this situation be resolved?
Some insurance companies or care plans may
be reluctant to pay for a referral outside their
usual system. Documents such as this one can
persuade them that referral of these patients is not
just better for the patient and child but crucial for
their well being.
References
[1] Pradat P, Francannet C, Harris JA, et al. The epide-
miology of cardiovascular defects, part 1: a study
based on data from three large registries of congeni-
tal malformations. Pediatr Cardiol 2003;24:195–221.
[2] Gillum RF. Epidemiology of congenital heart dis-
ease in the United States. Am Heart J 1994;127:
919–27.
[3] Warnes CA. The adult with congenital heart disease:
born to be bad? J Am Coll Cardiol 2005;46:1–8.
[4] Hoyert DL, Mathews TJ, Menacker F, et al. Annual
summary of vital statistics: 2004. Pediatrics 2006;
117(1):168–83.
[5] Bhat AH, Sahn DJ. Congenital heart disease never
goes away, even when it has been ‘‘treated’’: the
adult with congenital heart disease. Curr Opin
Pediatr 2004;16:500–7.
[6] Canobbio MM. Health care issues facing adoles-
cents with congenital heart disease. J Pediatr Nurs
2001;16:363–70.
603
[7] Lane DA, Lip GYH, Millane TA. Congenital heart
disease: quality of life in adults with congenital heart
disease. Heart 2002;88:71–5.
[8] Kovacs AH, Sears SF, Saidi AS. Biopsychosocial
experiences of adults with congenital heart disease:
review of the literature. Am Heart J 2005;150:
193–201.
[9] Swan L, Lupton M, Anthony J, et al. Controversies
in pregnancy and congenital heart disease. Congenit
Heart Dis 2006;1:27–34.
[10] Whittemore R, Hobbins JC, Engle MA. Pregnancy
and its outcome in women with and without surgical
treatment of congenital heart disease. Am J Cardiol
1982;50:641–51.
[11] Shime J, Mocarski EJM, Hastings D, et al. Congen-
ital heart disease in pregnancy: short- and long-term
implications. Am J Obstet Gynecol 1987;156:
313–22.
[12] Abbas AE, Lester SJ, Connolly H. Pregnancy and
the cardiovascular system. Int J Cardiol 2005;98:
179–89.
[13] Siu SC, Sermer M, Colman JM, et al. Prospective
multicenter study of pregnancy outcomes in women
with heart disease. Circulation 2001;104:515–21.
[14] Head CEG, Thorne SA. Congenital heart disease in
pregnancy. Postgrad Med J 2005;81:292–8.
[15] Siu SC, Colman JM. Congenital heart disease: heart
disease and pregnancy. Heart 2001;85:710–5.
[16] Connolly HM, Warnes CA. Pregnancy and contracep-
tion. In: Gatzoulis MA, Webb GD, Daubeney PEF,
editors. Diagnosis and management of adult congeni-
tal heart disease. London: Churchill Livingstone; 2003.
[17] Hunter S, Robson SC. Adaptation of the maternal
heart in pregnancy. Br Heart J 1992;68:540–3.
[18] Varga I, Rigo J, Somos P, et al. Analysis of ma-
ternal circulation and renal function in physio-
logic pregnancies; parallel examinations of the
changes in the cardiac output and the glomerular
filtration rate. J Maternal-Fetal Med 2000;9:
97–104.
[19] Conrad KP. Mechanisms of renal vasodilation and
hyperfiltration during pregnancy. J Soc Gynecol
Investig 2004;11:438–48.
[20] Carbillon L, Uzan M, Uzan S. Pregnancy, vascular
tone, and maternal hemodynamics: a crucial adapta-
tion. Obstet Gynecol Surv 2000;55:574–81.
[21] Campos O. Doppler echocardiography during preg-
nancy: physiologic and abnormal findings. Echocar-
diography 1996;13:135–46.
[22] Klein LL, Galan HL. Cardiac disease in pregnancy.
Obstet Gynecol Clin N Am 2004;31:429–59.
[23] Duprez D, Kaufman J, Liu Y, et al. Increases in
atrial natriuretic peptide after delivery and in the pu-
erperium. Am J Cardiol 1989;64:674–5.
[24] Immer FF, Bansi AG, Immer-Bansi AS, et al. Aortic
dissection in pregnancy: analysis of risk factors and
outcome. Ann Thorac Surg 2003;76:309–14.
[25] Kelly BA, Bond BC, Poston L. Aortic adaptation
to pregnancy: elevated expression of matrix
604 KAFKA
metalloproteinanes-2 and -3 in rat gestation. Mol
Hum Reprod 2004;10:331–7.
[26] Edouard DA, Pannier BM, London GM, et al. Ve-
nous and arterial behavior during normal preg-
nancy. Am J Physiol 1998;274:H1605–12.
[27] Brenner B. Haemostatic changes in pregnancy.
Thromb Res 2004;114:409–14.
[28] Uebing A, Steer PJ, Yentis SM, et al. Pregnancy and
congenital heart disease. BMJ 2006;332:401–6.
[29] Khairy P, Ouyang DW, Fernandes SM, et al. Preg-
nancy outcomes in women with congenital heart dis-
ease. Circulation 2006;113:517–24.
[30] Niwa K, Siu SC, Webb GD, Gatzoulis MA. Pro-
gressive aortic root dilatation in adults after
repair of Tetralogy of Fallot. Circulation 2002;
106:1374–8.
[31] Steer PJ. Pregnancy and contraception. In:
Gatzoulis M, Swan L, Therrien J, Pantely G, editors.
Adult congenital heart disease. Oxford: Blackwell
Publishing; 2005. p. 16–35.
[32] Gowda RM, Khan IA, Mehta NJ, et al. Cardiac ar-
rhythmias in pregnancy: clinical and therapeutic
considerations. Int J Cardiol 2003;88:129–33.
[33] Shotan A, Ostrzega E, Mehra A, et al. Incidence of
arrhythmias in normal pregnancy and relation to
palpitations, dizziness, and syncope. Am J Cardiol
1997;79(8):1061–4.
[34] Widerhorn J, Widerhorn ALM, Rahimtoola SH,
et al. WPW syndrome during pregnancy: increased
incidence of supraventricular arrhythmias. Am
Heart J 1992;123:796–8.
[35] Romem A, Romem Y, Katz M, et al. Incidence and
characteristics of maternal cardiac arrhythmias dur-
ing labor. Am J Cardiol 2004;93:931–3.
[36] Campos O, Andrade JL, Bocanegra J, et al. Physio-
logic multivalvular regurgitation during pregnancy:
a longitudinal Doppler echocardiographic study.
Int J Cardiol 1993;40:265–72.
[37] Broberg CS, Yentis SM, Steer PJ, et al. Women and
congenital heart disease. In: Wenger N, Collins P,
editors. Women and heart disease. 2nd edition.
New York: Taylor & Francis; 2005. p. 454–71.
[38] De Swiet M, Nelson-Piercy C. Cardiac disease. In:
Lewis G, editor. Why mothers die 2000–2002: the sixth
report into maternal deaths in the United Kingdom.
London: RCOG Press; 2004 Available at:http://
www.cemach.org.uk/publications/WMD2000_2002/
content.htm. Accessed September 18, 2006.
[39] Siu SC, Colman JM, Sorensen S, et al. Adverse neo-
natal and cardiac outcomes are more common in
pregnant women with cardiac disease. Circulation
2002;105:2179–84.
[40] Avila WA, Rossi EG, Ramires JAF, et al. Pregnancy
in patients with heart disease: experience with 1000
cases. Clin Cardiol 2003;26:135–42.
[41] Presbitero P, Somerville J, Stone S, et al. Congenital
heart disease: pregnancy in cyanotic congenital heart
disease: outcome of mother and fetus. Circulation
1994;89:2673–6.
et al
[42] Georgieff MK. Intrauterine growth retardation and
subsequent somatic growth and neurodevelopment.
J Pediatr 1998;133:3–5.
[43] Brodszki J, Lanne T, Marsal K, et al. Pediatric car-
diology: impaired vascular growth in late adoles-
cence after intrauterine growth restriction.
Circulation 2005;111:2623–8.
[44] Barker DJ. Early growth and cardiovascular disease.
Arch Dis Child 1999;80:305–10.
[45] Romano-Zelekha O, Hirsh R, Blieden L, et al. The
risk for congenital heart defects in offspring of indi-
viduals with congenital heart defects. Clin Genet
2001;59:325–9.
[46] Burn J, Brennan P, Little J, et al. Recurrence risks in
offspring of adults with major heart defects: results
from first cohort of British collaborative study. Lan-
cet 1998;351:311–6.
[47] Lupton M, Oteng-Ntim E, Ayida G, et al. Cardiac
disease in pregnancy. Curr Opin Obstet Gynecol
2002;14:137–43.
[48] Renforth GL, Wilson DI. Adults with congenital
heart disease: a genetic perspective. In: Gatzoulis
MA, Webb GD, Daubeney PEF, editors. Diagnosis
and management of adult congenital heart disease.
London: Churchill Livingstone; 2003. p. 19–24.
[49] Nora JJ. From generational studies to a multilevel ge-
netic-environmental interaction. J Am Coll Cardiol
1994;23:1468–71.
[50] Bassett AS, Chow EWC, Husted J, et al. Clinical fea-
tures of 78 adults with 22q11 deletion syndrome. Am
J Med Genet 2005;138A:307–13.
[51] Carvalho JS. Early prenatal diagnosis of major con-
genital heart defects. Curr Opin Obstet Gynecol
2001;13:155–9.
[52] Canobbio MM, Perloff JK, Rapkin AJ. Gynecolog-
ical health of females with congenital heart disease.
Int J Cardiol 2005;98:379–87.
[53] Reid GJ, Irvine MJ, McCrindle BW, et al. Preva-
lence and correlates of successful transfer from pedi-
atric to adult health care among a cohort of young
adults with complex congenital heart defects. Pediat-
rics 2004;113:197–205.
[54] Kafka H, Uemura H, Gatzoulis MA. Surgical and
catheter intervention during pregnancy in patients
with heart disease. In: Steer P, Gatzoulis MA,
Baker P, editors. Cardiac disease and pregnancy.
London: RCOG Press; 2006. p. 95–128.
[55] Leonard H, O’Sullivan J, Hunter S. Family planning
requirements in the adult congenital heart disease
clinic. Heart 1996;76:60–2.
[56] ACOG Committee Opinion No. 299. Guidelines for
diagnostic imaging during pregnancy. Obstet Gyne-
col 2004;104:647–51.
[57] Leyendecker JR, Gorengaut V, Brown JJ. MR imag-
ing of maternal diseases of the abdomen and pelvis
during pregnancy and the immediate postpartum
period. Radiographics 2004;24:1301–16.
[58] James PR. Cardiovascular disease. Best Pract Res
Clin Obstet Gynaecol 2001;15:903–11.
 PREGNANCY AND CONGENITAL HEART DISEASE
[59] Elkayam U. Pregnancy through a prosthetic heart
valve. J Am Coll Cardiol 1999;33:1642–5.
[60] Bates SM, Greer IA, Hirsh J, et al. Use of antithrom-
botic agents during pregnancy. Chest 2004;126:
627S–44S.
[61] Qasqas SA, McPherson C, Frishman WH, et al. Car-
diovascular pharmacotherapeutic considerations
during pregnancy and lactation. Cardiol Rev 2004;
12:201–21.
[62] Thorne SA. Pregnancy in heart disease. Heart 2004;
90:450–6.
605
[63] Magee LA. Antihypertensives. Best Pract Res Clin
Obstet Gynaecol 2001;15:827–45.
[64] Ro A, Frishman WH. Peripartum cardiomyopathy.
Cardiol Rev 2006;14:35–42.
[65] Greenberg SS, Lancaster JR, Xie J, et al. Effects of
NO synthase inhibitors, arginine-deficient diet, and
amiloride in pregnant rats. Am J Physiol 1997;273:
R1031–45.
[66] Vonder Muhl I, Cumming G, Gatzoulis MA. Risky
business: insuring adults with congenital heart dis-
ease. Eur Heart J 2003;24:1595–600.
 Cardiol Clin 24 (2006) 607–618

The Role of the Psychologist in Adult

Congenital Heart Disease
Adrienne H. Kovacs, PhDa,*, Candice Silversides, MDb,
Arwa Saidi, MB, BChc, Samuel F. Sears, PhDd
a
Cardiac Psychology, Division of Cardiology, University Health Network, 399 Bathurst Street,
1-West-414, Toronto, ON M5T 2N2, Canada
b
Division of Cardiology, University Health Network, University of Toronto,
Toronto General Hospital, 585 University Avenue, 5N-521, Toronto, ON M5G 2N2, Canada
c
Departments of Pediatrics and Internal Medicine, University of Florida, PO Box 100296,
1600 SW Archer Road, Gainesville, FL 32610, USA
d
Department of Clinical and Health Psychology, College of Public Health and Health Professions,
University of Florida, Box 100165 UFHSC, 1600 SW Archer Road, Gainesville, FL 32610, USA

Due to diagnostic, surgical, interventional, and potential psychosocial consequences of growing

pharmacologic advances, it is estimated that up to
95% of infants born with congenital heart defects
will reach adulthood [1]. There are now more
adults than children living with congenital heart
disease [2], but there exists a less than ideal capac-
ity to manage adults in pediatric or adult cardiol-
ogy clinics. Adult congenital heart disease
(ACHD) patients are at an increased risk of late
cardiac complications including arrhythmias, en-
docarditis, congestive heart failure, and pulmo-
nary vascular disease. The effects of chronic
illness, multiple surgical admissions, and hospital-
ization in childhood, adolescence, and adulthood
can have a significant impact on the psychological
well-being of these patients. The evidence for psy-
chological distress in impacting health outcomes
and quality of life (QOL) across cardiac patient
populations is now considered by some as ‘‘clear
and convincing’’ [3]. Psychological distress is com-
mon following cardiac diagnosis, with approxi-
mately 20% to 50% of patients reporting
adjustment difficulties [4]. Therefore, in addition
to monitoring and treating the medical sequelae,
it is important to recognize and manage the
* Corresponding author.
E-mail address: adrienne.kovacs@uhn.on.ca
(A.H. Kovacs).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.003
up with congenital heart disease [5,6]. Interna-
tional working groups have emphasized the inclu-
sion of specialized mental health care for ACHD
[7–9]. The summary document of the 32nd Be-
thesda Conference (‘‘Care of the Adult with Con-
genital Heart Disease’’) included the following
statement: ‘‘The emotional health of adults with
[congenital heart disease] should be a priority in
the overall care of this patient population’’ [7].
Similarly, the Task Force on the Management of
Grown Up Congenital Heart Disease of the Euro-
pean Society of Cardiology stated, ‘‘The specialist
service for grown-ups with congenital heart dis-
ease must provide support for the many psychoso-
cial problems in this population’’ [9].
In this article, the authors review the evidence
that patients who have ACHD have special and
unique psychosocial needs and discuss ways in
which psychologists can be integrated into multi-
disciplinary ACHD care teams. The psychosocial
adjustment and QOL issues observed in adults
who have grown up with congenital heart defects
are reviewed. Three professional domains in
which clinical health psychologists can contribute
to an ACHD teamdprovision of clinical services,
multidisciplinary research, and professional edu-
cationdare discussed in detail. Finally, consider-
ations for introducing or increasing the presence
of psychology in ACHD teams are presented.
cardiology.theclinics.com
608 KOVACS
Psychosocial adjustment in adult congenital heart
disease
Psychological functioning
The increasing number of published studies
evaluating the psychological aspects of ACHD
reflects the growing recognition of the potential
emotional and behavioral implications of growing
up with a serious medical condition. In general,
ACHD patients are at greater risk of psycholog-
ical difficulties than adults who have not lived
with chronic medical conditions [10–13]. Based on
studies that employed clinical interview methodol-
ogies, it is estimated that 35% to 79% of ACHD
patients meet diagnostic criteria for a psychiatric
diagnosis [11,13]. Of note, this figure of 35% was
observed in a sample of ACHD patients deemed
‘‘well adjusted’’ by their medical teams. There
are, however, European data suggestive of favor-
able psychological adjustment among ACHD
patients compared with healthy peers [14–17].
Conflicting results might reflect variation in
measurement tools, study populations, and meth-
odologic approaches associated with self-report
versus interview assessments. ACHD patients
might under-report psychological symptoms on
self-report measures as a reflection of denial and
high achievement motivation [14,16]. Clinical in-
terviews likely offer the most accurate picture of
psychological adjustment. Reported differences
in psychological adjustment might also reflect so-
ciocultural differences in the long-term care of in-
dividuals who have chronic medical conditions [5].
Individuals from countries with universal access
and stronger mental health care systems might
report more positive health expectations or
experiences.
Compared with the general population, it
appears that the proportion of ACHD patients
experiencing psychological distress is elevated.
Risk factors related to psychological difficulties
in this population have been studied, and it has
been demonstrated that younger patients tend to
have more psychological problems than older
patients and women appear to have more psycho-
logical difficulties than men [18,19]. Results inves-
tigating the impact of medical variables are
inconsistent. Studies by Brandhagen and col-
leagues [10] and Utens and colleagues [14,15]
found that disease variables were unrelated to psy-
chological symptoms, whereas other studies found
that disease severity was related to psychological
symptoms [13,18]. Van Rijen and coworkers [18]
et al
found that significant medical predictors of pa-
tient-reported behavioral and emotional problems
included low maximal exercise capacity, physi-
cian-imposed restrictions, and negative personal
experiences with scarring. Popelova and col-
leagues [12] found that depressive symptoms
were associated with poorer functional state.
Collectively, these studies suggest that functional
status might be a stronger predictor of psycholog-
ical health than disease severity. Although psy-
chologists obviously cannot impact the sex or
age of a patient, other factors including depressive
symptoms, perception of scarring, and even func-
tional status (eg, reflecting improved exercise
capacity) are often amenable to change.
Quality of life
A complete understanding of the experience of
growing up with a congenital heart defect includes
not only traditional markers of psychological
maladjustment (ie, anxiety and depression) but
also health-related QOL. QOL has emerged as an
end point of interest because it reflects a patient’s
life satisfaction and ability to function in a variety
of life domains including physical, social, emo-
tional, and work-related. Some researchers believe
it is embodied in the approach of the biopsy-
chosocial model of health in which biologic,
psychological, and social functioning are interde-
pendent [20]. QOL, however, should not be equated
with health status. Moons and colleagues [21]
defined QOL as ‘‘the degree of overall life satisfac-
tion that is positively or negatively influenced by
individuals’ perception of certain aspects of life im-
portant to them, including matters both related and
unrelated to health.’’ Moons and coworkers [22]
described a ‘‘paradigm shift’’ in the conceptualiza-
tion of the QOL of ACHD patients, such that QOL
is most accurately understood through examina-
tion of domains deemed important to an individual
patient, rather than through the administration of
standardized questionnaires. Among ACHD pa-
tients, the most important determinants of an indi-
vidual’s QOL are family, job/education, friends,
health, and leisure time [22].
Although some researchers have concluded that
QOL in patients who have congenital heart disease
is poorer compared with healthy peers [23–26],
others have not [27–29]. Poor QOL in the ACHD
population has been linked with (1) clinical vari-
ables (cyanosis, orthopedic problems); (2) social
variables (lower level of educational attainment,
age, social impediments); and (3) psychological
 ROLE OF PSYCHOLOGISTS
variables (negative thoughts) [28,30,31]. Saliba and
colleagues [28] detected poorer QOL in cyanotic
patients compared with their noncyanotic peers.
Simko and McGinnis [32] found that individuals
who had acyanotic defects had better QOL than
those who had single-ventricle physiology. Ternes-
tedt and coworkers [33] noted that individuals who
had tetralogy of Fallot reported better QOL than
individuals who had atrial septal defects, an unex-
pected finding given that atrial septal defects are
generally considered a more benign cardiac lesion.
There are data, however, that suggest that indices
of functional status might be stronger predictors
of QOL than the severity of the initial cardiac diag-
noses [34]. For example, Rose and colleagues [26]
observed that objective cardiopulmonary function-
ing strongly impacted cardiac complaints and the
physical component of health-related QOL. Psy-
chological intervention can target functional status
through the promotion of healthy lifestyle behav-
iors including exercise, diet, and adherence to med-
ical regimens.
Social adjustment
Some individuals experience difficult familial
and peer interactions as a result of growing up with
congenital heart defects. Approximately one
fourth of ACHD patients describe parental over-
protection in childhood and adolescence [10,35].
ACHD patients are more likely to live with their
parents than their healthy peers [36]. ‘‘Feeling dif-
ferent’’ and body image concerns are common to
many adolescents and young adults living with
congenital heart disease [11,37,38]. Horner and col-
leagues [11] observed a pattern of pervasive denial
during their interviews with 29 ACHD patients. It
was common for their ACHD interviewees to re-
veal difficulties with childhood and adolescent so-
cial interactions only after extensive questioning.
Many of these interviewees also recalled coping
with negative childhood interactions by wearing
‘‘a happy face’’ and behaving as though ‘‘every-
thing was fine.’’ As adolescents who have congeni-
tal heart disease progress into adulthood, they are
presented with a new set of challenges. Difficulties
obtaining employment and health insurance can
be particularly frustrating [12,30,39]. In addition,
issues surrounding contraception, pregnancy, and
parenthood require extra consideration for many
ACHD patients [40]. Psychologists can provide
supportive and cognitive-behavioral therapy for
these difficult issues (eg, managing social anxiety,
managing stress, strategies for discussing sensitive
issues with one’s medical team).
609
Limitations of current research
There is a sample bias when studying a restricted
group of individuals. ACHD patients in these
studies are typically those who (1) receive follow-up
care for a congenital heart defect, (2) receive their
care at a facility with a program of research, (3) are
not excluded based on cognitive functioning, and
(4) consent to participate in research studies. It
is unfortunate that most individuals born with
heart defects do not receive specialized ACHD care
[41]; therefore, study results cannot necessarily be
generalized to the entire population of ACHD pa-
tients. We have no way of knowing whether the
psychosocial health of ACHD patients who are
lost to cardiology follow-up (or who present to
smaller cardiology clinics at which research is not
conducted) is better or worse than that of study
participants; however, the authors believe that
the conclusions can be generalized to patients
who present at ACHD specialty centers.
Need for specialized care
There is a growing recognition of the biopsy-
chosocial implications of growing up with congen-
ital heart disease. In addition to specialized
assessment and treatment of the biomedical se-
quelae of cardiac defects, the psychosocial well-
being of this population deserves specialized mental
health care. With assessment and psychotherapeu-
tic techniques, psychologists are able to provide
emotional support and specific strategies to maxi-
mize psychological well-being.
Psychologists typically have doctorates in phi-
losophy (PhD), education (EdD), or psychology
(PsyD). Psychologists receive specialized training
in assessment; case conceptualization; individual,
family, or group psychotherapy; professional
ethics; research design; and statistical analysis.
Strategies of a cognitive-behavioral psychologist
include worry management, thought restructuring,
relaxation training, behavioral activation, and
communication skills training. Psychologists who
have health psychology experience are aware of
the relationship between physical health and psy-
chological health and are able to tailor the afore-
mentioned strategies to assist individuals in
managing their illness experience. They are also
able to ‘‘normalize’’ patient experiences (versus
adopting a pathologic conceptualization of emo-
tional responses to illness). Psychiatrists, by com-
parison, are medical doctors and are more
strongly associated with a biologic approach to
mental health. Psychotropic medication can be an
important component of treatment for many
610 KOVACS
ACHD patients; however, it is not a replacement
for psychotherapy (‘‘talk therapy’’), and some
ACHD patients are reluctant to add to a complex
medication regimen. Collectively, psychology and
psychiatry can collaborate to provide optimal men-
tal health treatment for ACHD patients.
Roles of psychologists in adult congenital heart
disease
Provision of clinical services
The 32nd Bethesda Conference stated, ‘‘Appro-
priate screening and referral sources for [mental
health] treatment should be available at all re-
gional ACHD centers’’ [7]. Similarly, the Canadian
Consensus document stated that consultation with
psychology/psychiatry is indicated to ‘‘facilitate
optimal functioning,’’ to address clinical and sub-
clinical depression and anxiety, and to enhance
adjustment to illness exacerbations and hospi-
talizations [8]. The European Task Force also em-
phasized the importance of support for the
psychological difficulties of ACHD patients [9]. De-
spite these recommendations, there is underdiagno-
sis and undertreatment of psychosocial concerns in
ACHD patients [13]; therefore, clinical health psy-
chologists have clear potential to be important
members of ACHD teams. Psychologists are able
to diagnose psychiatric disorders and provide treat-
ment or to assist with referrals to other mental
health professionals when appropriate.
Symptoms of mood and anxiety disorders are
two of the more obvious indications for referral to
a psychologist. In a cardiology setting, such re-
ferrals are particularly warranted because psycho-
social factors are associated with the onset of
coronary artery disease and increased cardiac
mortality [42–44]. For example, depression follow-
ing myocardial infarction has been linked with a
2- to 2.5-fold increase in cardiovascular events or
mortality [45]. Additional specific indications for
referral to a psychologist working with an ACHD
team are provided in Box 1 and then discussed in
detail.
Psychological distress
Difficulties coping with medical condition. Al-
though there is a tendency for some patients to
consider themselves ‘‘cured,’’ the reality is that
many require lifelong medical follow-up [1]. Re-
search suggests that functional status is more
strongly related to QOL than cardiac defect sever-
ity [34]. It can be a significant psychological setback
when patients unexpectedly experience decrements
et al
Box 1. Common indications for referral
to psychology
Psychological distress
Difficulties coping with medical
condition
Cardiac anxiety
Patient mortality concerns
Medical treatment–related issues
Treatment decision making
Surgical preparation
Adjustment to cardiac devices
Psychosocial difficulties
Family and peer concerns
Difficult pediatric–adult transitions
Healthy lifestyle concerns
Behavioral modification
Maximizing adherence
Developmental/neurocognitive issues
in their functional status or learn that further
medical or surgical intervention is advised. Horner
and colleagues [11] reported that many of their in-
terviewees were angry when cardiac complications
developed in adulthood after years of relative
stability. Although this article focuses on the expe-
riences of adults who have grown up with congeni-
tal heart disease, some ACHD patients do not learn
of their cardiac defect until adulthood. These indi-
viduals may be ill prepared to cope with the impact
of this new diagnosis and its implications. Psychol-
ogists can provide emotional support and cogni-
tive-behavioral strategies to manage anger and
health-related stressors.
Cardiac anxiety. Heart-focused anxiety refers to
the fear of cardiac sensations and events (eg,
increased heart rate) because of presumed nega-
tive consequences [46]. Clinical experience sug-
gests that this is more common after individuals
are informed that their cardiac condition has
worsened or that they require follow-up proce-
dures or closer cardiac monitoring. Patients who
have higher cardiac anxiety tend to pay greater
attention to cardiac sensations and avoid certain
activities they fear will trigger cardiac symptoms
[46]. For example, among a sample of patients
who had implantable cardioverter defibrillators
(ICDs), patients who had panic disorder were
more likely to report attempts to pay attention
 ROLE OF PSYCHOLOGISTS
to bodily changes to avoid being surprised by their
next shocks and the belief that they could perceive
heartbeat irregularities [47]. Utens and colleagues
[15] hypothesized that increased somatic com-
plaints among ACHD patients might reflect their
heightened alertness to bodily symptoms. This in-
creased somatic hypersensitivity might correspond
to an increased risk of psychological stress, partic-
ularly anxiety. Rietveld and coworkers [48] found
that compared with healthy comparison peers,
ACHD patients were more likely to be distracted
by their heart beats during a concentration task
and were less accurate at estimating their heart
rate during another task. Unexpectedly, in this
sample, ACHD patients did not report more
anxiety associated with audible heartbeats than
healthy comparison peers. Psychologists can assist
individuals who have cardiac anxiety by way of
psychoeducation and the provision of strategies
selected to reduce excessive self-monitoring of
cardiac symptoms and to promote engagement
in avoided activities.
Patient mortality concerns. The psychosocial con-
cerns of ACHD patients do not remain static
throughout their lifetimes. Individuals who have
chronic medical conditions develop an ‘‘illness
career’’ that varies in response to several factors
including changes in health and interactions with
health care professionals [49]. Certain situations
result in increased patient awareness of mortality
(ie, changes in cardiac status, including but not
limited to being informed of terminal status). In
this difficult stage of life, it is important that
patients, if they choose, are able to clearly discuss
their concerns and fears with supportive individ-
uals. Patients may be reluctant to disclose their
concerns to physicians for a variety of reasons, in-
cluding not wanting to ‘‘burden’’ their physicians
with nonmedical concerns. Patients might also be
cautious to reveal their fears and concerns to fam-
ily members who are already distressed by the sit-
uation. Psychologists can assist these individuals
during this difficult time by providing emotional
support, discussing anticipatory grief, and sharing
strategies to communicate most effectively with
medical teams and loved ones.
Medical treatment–related issues
Treatment decision making. Cardiac patients are
often asked to make difficult medical decisions.
For example, they might be recommended to
undergo invasive cardiac procedures, many of
which have relatively high risks of complications.
611
They may have difficulties understanding medical
recommendations and respond with exaggerated
fears. Patients may benefit from additional sup-
port or counseling to be able to commit to these
complex care decisions. For any difficult decision,
it is important that depression or anxiety does not
distort patient judgment [50,51]. Psychologists can
evaluate the psychological functioning of individ-
uals who are making difficult decisions. They can
also assist patients in the decision-making process
(eg, with a decisional balance exercise in which the
pros and cons of options are thoroughly
reviewed).
Surgical preparation. Greater presurgical anxiety
is associated with more complicated and slower
surgical recovery [52]. Psychological factors can
influence surgical recovery through endocrine
and immune functioning and behavioral mecha-
nisms (including smoking and poor diet) [52].
Meta-analytic studies indicate that psychological
preparation for surgery and psychoeducational
interventions can positively impact outcomes
including psychological distress, pain, and surgi-
cal recovery [53,54]. Psychologists can offer surgi-
cal preparation techniques (including stress
management and imagery) that have proved effec-
tive in the reduction of anxiety for cardiac patients
undergoing surgical procedures [55].
Adjustment to cardiac devices. The use of ICDs is
an increasing treatment for ACHD patients at risk
for sudden cardiac death [56]; however, this life-
saving technology is not without psychosocial im-
pact. Although in general, ICDs appear equal or
better than antiarrhythmic medications in terms
of their impact on patient QOL [57–60], they
lead to significant psychological distress for
many device recipients. Approximately 13% to
38% of recipients experience diagnosable levels
of anxiety, and significant depression is observed
in 24% to 33% of recipients [61]. Younger age
and greater frequency of ICD firings appear to
be the strongest risk factors for psychological dis-
tress [61]. The provision of ICD-specific education
and cognitive-behavioral strategies by psycholo-
gists can enhance the confidence and QOL of
ICD recipients [62].
Psychosocial difficulties
Family and peer concerns. As outlined earlier,
parental overprotection and difficult peer interac-
tions are common among ACHD patients. Fur-
thermore, childhood and adolescent experiences
with peers and family are not always easily
612 KOVACS
forgotten in adulthood. Some ACHD patients
benefit from social skills training. Others benefit
from learning communication strategies to exert
increased independence with family members or to
be more involved in their health care. These are two
specific examples in which psychologists can pro-
vide interventions to enhance social well-being.
Difficult pediatric–adult transitions. The transition
from pediatric to adult health care can be tumul-
tuous for many ACHD patients (see the article by
Webb and Reiss found elsewhere in this issue).
Many adolescents who have congenital heart
disease do not successfully transition to adult
care [63]. Although successful transition begins
in the pediatric setting, psychologists working in
ACHD can provide individual consultation to pa-
tients experiencing a difficult transition (eg, offer-
ing strategies to empower patients to assume
increased control over their cardiac management).
In addition, psychologists can contribute to a for-
malized transition program. For example, the
Toronto Congenital Cardiac Centre for Adults
currently offers transition nights for 17- and 18-
year-old pediatric patients before their first visit
to the center. A psychologist attends these meet-
ings to proactively discuss transition issues with
patients and families.
Healthy lifestyle concerns
Behavioral modification. A healthy lifestyle in-
cluding a well-balanced diet and an individually
appropriate level of physical activity is important
for all individuals but even more so for those who
have cardiac conditions. Dusseldorp and col-
leagues [64] conducted a meta-analysis of psycho-
education programs for individuals who had
coronary artery disease and concluded that such
interventions had a positive impact on healthy
lifestyle behaviors (eg, body weight, diet, exercise
habits) and resulted in reduced cardiac mortality
and recurrence of myocardial infarction. Al-
though there are no studies examining the impact
of psychoeducation targeting heart-healthy be-
haviors for ACHD patients, it is assumed that
such education is beneficial. It is unfortunate
that ACHD patients are usually less likely to en-
gage in physical activity [65]. Psychologists can
provide behavioral strategies for the adoption
and maintenance of a physician-approved physi-
cal activity regimen that empowers patients to ex-
ercise control over their condition.
Maximizing adherence. Kulkarni and colleagues
[66] noted that CAD patients often stop
et al
medications within 1 year of their prescription.
They also found that poorer adherence to a cardio-
vascular medication regimen was linked to a
number of factors including lower educational
attainment, older age, greater number of medica-
tions prescribed, and poorer mental health [66].
Depression is strongly linked to nonadherence to
medical treatments [67] and has been associated
with nonadherence to the prescribed behavioral
and lifestyle modifications in postmyocardial in-
farction patients [68]. In a sample of heart failure
patients, depression has also been associated with
patient-reported difficulty taking medications as
directed by physicians [69]. Psychologists can pro-
vide strategies that target depressive symptoms
and nonadherence concerns concurrently.
Developmental/neurocognitive issues
Neuropsychologic assessment in individuals
who have congenital heart disease can help to
detect cerebral dysfunction in the absence of anat-
omic alterations in the brain and to evaluate the
cognitive impact of disease [70]. Neurocognitive
deficits are associated with certain genetic disorders
(eg, Down syndrome) and sequelae from ischemia/
hypoxia and pediatric cardiopulmonary bypass.
A review of the impact of chronic or intermittent
hypoxia on cognition in childhood indicated
adverse impacts on cognitive, behavioral, and
academic outcomes [71]. The impact of cardiac
defects on cognitive and academic performance
may persist even after cardiac surgery [72]. In
a study of individuals who were operated on during
childhood for tetralogy of Fallot, the most common
cognitive deficits were exhibited in tasks that de-
manded executive functioning, problem solving,
and planning (rather than memory, learning, or
attention); many of these patients also reported
a lower than normal level of academic achievement
[73]. In another study, approximately one fourth of
ACHD patients had participated in special educa-
tion [16]. A thorough review of the neurocognitive
deficits associated with specific congenital heart de-
fects is beyond the scope of this article, but excellent
summaries are available [70,74]. Neuropsycholo-
gists can prove to be especially valuable in the
assessment of neurocognitive functioning of chil-
dren, adolescents, and adults who have congenital
heart defects.
Multidisciplinary research
As stated earlier, psychologists receive exten-
sive training in research design and statistical
analysis during their graduate training and are
 ROLE OF PSYCHOLOGISTS
therefore able to contribute to the overall research
agenda of an ACHD team. Box 2 indicates poten-
tial areas of collaborative research. In all areas of
ACHD patient research (biomedical or psychoso-
cial), multicenter studies are recommended. To
truly increase our understanding of ACHD
patient experiences, collaboration between health
care teams at different centers is crucial.
Development of adult congenital heart
disease–specific measures
Within the cardiac psychology and nursing
literature, there are specific measures targeting
different subgroups of patients, including those
who have angina, heart failure, ICDs, and so
forth. Therefore, it is generally accepted that the
patient experience can differ across cardiac con-
ditions. American and European guidelines call
for the creation of psychosocial measures specifi-
cally developed for the ACHD patient population
[7,9]. Kamphuis and colleagues [75] developed and
validated a specific assessment tool for ACHD pa-
tients that has three subscales pertaining to prob-
lems associated with physical symptoms,
cardiac monitoring and treatment (eg, echocardi-
ography, taking blood, hospital admission), and
worries. Additional ACHD-specific measures
should be developed that address the unique
concerns of this patient population. For exam-
ple, just as the left ventricular ejection fraction
is generally less meaningful for ACHD patients,
psychosocial measures designed for adults who
have acquired cardiac disease do not capture
the full ACHD patient experience (eg, transition
from pediatric to adult cardiology care, chronic
cyanosis, and so forth). Health professionals (in-
cluding psychologists) who have knowledge of
the unique ACHD patient experiences are best
able to create and evaluate such measures.
Longitudinal psychosocial assessment
There are few prospective studies examining
the psychosocial adjustment of ACHD patients;
Box 2. Areas of research
Development of ACHD-specific
measures
Longitudinal psychosocial assessment
Evaluation of psychological interventions
Evaluation of medical interventions
Relationship between mental and
physical health
613
most research has used a cross-sectional design.
Longitudinal studies, by definition, would allow
us to monitor the psychosocial functioning of
ACHD over time to better understand the impact
of the transition from pediatric to adult care,
changing cardiac status, and general aging expe-
riences of ACHD patients.
Evaluation of psychological interventions
Despite an increasing awareness of the psycho-
social concerns of the ACHD patient population,
there are no published trials of psychological
interventions for ACHD patients [76]. One practi-
cal consideration is the fact that many individuals
attending ACHD outpatient clinics travel quite
a distance to attend clinics. For this reason, ‘‘cre-
ative solutions for counseling patients in groups
or those who live a far distance away should be
developed’’ [7]. Interventions provided by tele-
phone or over the Internet might be most realistic
for many ACHD patients.
Evaluation of medical interventions
The impact of surgical and interventional pro-
cedures is not limited to cardiac anatomy. Suc-
cessful outcomes include improvements in
functional status and health-related QOL. Bio-
medical and psychosocial outcomes should be
investigated to demonstrate the effectiveness of
medical treatment. The summary document of the
32nd Bethesda Conference recommended investi-
gating the physical and psychosocial impact of
regular follow-up [7], which illustrates the impor-
tance of multidisciplinary research such that phy-
sician, nursing, and psychology staff collaborate
to best understand the biopsychosocial experience
of medical treatment.
Relationship between mental and physical health
In the previous paragraph, research examining
the impact of medical intervention on psychoso-
cial functioning was recommended. It is equally
important to examine the influence of psycholog-
ical factors on physical functioning. Coronary
artery disease literature suggests that the relation-
ship between psychosocial status and cardiac
outcome might be mediated by physiologic mech-
anisms including alteration of catecholamine
activity, inflammatory responses, and impaired
cardiovascular reactivity. Behavioral pathways
linking psychosocial health and cardiac outcome
include adherence to prescribed treatments and
the adoption of healthy lifestyle behaviors. It is
now time to investigate ways in which psychoso-
cial factors might influence medical outcomes of
614 KOVACS
ACHD patients (eg, surgical recovery, heart func-
tioning, disease progression, and so forth). Again,
this line of research demands multidisciplinary
collaboration between medicine, nursing, and
psychology.
Professional education
The third area in which psychologists can
contribute to ACHD care teams involves the
education of medical trainees (including nursing
trainees) and student psychologists. As defined
by current guidelines, competency as a level 2 or
level 3 ACHD specialist includes knowledge of
psychosocial aspects of adolescence, the transition
to adulthood, experience with lifestyle counseling,
and advocacy [7,9]. These training guidelines
suggest that exposure to other professionals includ-
ing psychologists is important during training.
Education for cardiology residents and fellows,
staff cardiologists, and nurses can occur for-
mally (eg, presentation at professional meetings,
rounds, case conferences) and informally (eg, ca-
sual questions/interactions). Box 3 presents the
more formal teaching topics for medical trainees
and psychologists.
Increased psychosocial awareness
Kovacs and colleagues [5] provided a list of six
clinical strategies to maximize psychosocial care
of ACHD patients. This list included the increased
psychosocial awareness of cardiologists working
within ACHD. Psychologists are the most quali-
fied individuals to educate physicians and nursing
staff regarding the psychosocial experiences of
ACHD due to their knowledge of relevant re-
search and clinical experiences; case studies can
often be more poignant and memorable than
data summaries. Psychologists emphasize the di-
versity of the patient experience. Not all ACHD
patients experience significant psychological dis-
tress, and triggers of distress vary greatly between
individuals.
Box 3. Formal teaching topics
Increased psychosocial awareness
Indications for referral to a mental health
practitioner
Physician–patient communication skills
training
Training psychologists to work in ACHD
et al
Indications for referral to a mental health
practitioner
Psychology referrals might be patient initiated
or physician initiated. ACHD patients occasion-
ally inquire directly about mental health treat-
ment; at other times, physicians broach this topic
with their patients. In such situations, it is
imperative that referral to psychology (or psychi-
atry) be made with the patient’s knowledge. Re-
ferrals should not be automatic when a patient
bursts into tears during a clinic appointment; it is
important for the physician or nurse to inquire
whether this reflects ongoing psychological dis-
tress or is a transient reaction to a stressful
medical situation. It is also important to clarify
appropriate referral indications with the particu-
lar psychologist or psychologists working with an
ACHD team. These referral indications might
include those presented earlier in this article;
however, there might be situations in which it is
more appropriate to refer to other mental health
professionals including neuropsychologists, psy-
chiatrists, and mental health professionals who
have experience in eating disorders, family ther-
apy, and specific psychiatric disorders. For exam-
ple, individuals who have 22q11 deletion
syndrome have significantly higher rates of
psychotic disorders and should be evaluated by
a psychiatrist [77]. Substance abuse is another ex-
ample of a problem that is best treated by individ-
uals who specialize in this area. Among ACHD
patients, substance abuse is of particular concern
because it has been linked to unsuccessful transi-
tions from pediatric to adult health care [63].
Physician–patient communication skills training
Patients consistently report that they want to be
heard by their medical team. Therefore, improving
physician–patient communication skills can be
very beneficial. A number of situations exist in
which such communication can be especially tricky,
including delivering ‘‘bad news’’ and proactive
discussions of challenging issues such as pregnancy
and medical prognosis. Patient participation in
medical consultations can be enhanced through
physician behaviors [78]. For interested readers,
Street and colleagues [78] provided a sample of phy-
sician statements intended to build partnerships
with patients and increase patient participation.
Psychologists can educate physician and nursing
staff with regard to the importance of effective
communication and patient participation in their
medical care. They can also present strategies for
improving physician–patient communication.
 ROLE OF PSYCHOLOGISTS
Training psychologists to work in ACHD
It is important that psychologists working in
ACHD strive to educate interested student psy-
chologists at various stages in their training
(including graduate school practica, predoctoral
internships, and postdoctoral fellowships). The
32nd Bethesda Conference emphasized the impor-
tance of developing training programs for non-
physician staff including psychologists [7].
Similarly, the 1996 Canadian Consensus Confer-
ence on Adult Congenital Heart Disease suggested
that an ideal supraregional ACHD referral center
would establish and evaluate ongoing training cri-
teria for cardiologists, surgeons, and associated
staff including psychologists [8]. Psychologists
working with ACHD patients should strive to edu-
cate student psychologists, with the ultimate goal of
increasing the number of psychologists who have
ACHD-specific training.
Considerations for integrating psychology
A number of factors impact the level of
involvement of a psychologist in an ACHD team:
The overall commitment of an adult congenital
heart disease team to the psychosocial concerns
of their patients
It is difficult to ignore the research findings that
suggest that ACHD patients are at increased risk
of psychosocial difficulties compared with indi-
viduals who did not grow up with lifelong medical
conditions. Recognition of psychosocial concerns,
however, does not imply a commitment to address
these concerns. It can require significant effort and
financial support to introduce a psychologist into
the ACHD care team.
The size of the adult congenital heart disease
program
A larger ACHD program with a number of
physicians and nurses may be better able to
support a more involved psychologist. Such a pro-
gram might support a ‘‘clinic within a clinic’’
model in which psychologists see patients during
the outpatient cardiology clinics and are available
for ‘‘on-the-spot consultation’’ [5]. For a smaller
program, however, it may be more reasonable to
identify psychologists in the medical facility or
community for clinical referrals or research col-
laboration. Regardless of the level of involvement,
it is important to establish a plan for integration
of the psychologist into the care team (eg,
615
education seminars for trainees and staff, atten-
dance at clinical rounds).
The availability of psychologists who have
specialized cardiac experience
Ideally, it is desirable to work with a psychol-
ogist who has specialized ACHD experience;
however, few psychologists currently have this
experience. It is more realistic to seek connections
with psychologists who have experience in health
psychology and, preferably, specialized training in
a cardiac setting.
Reimbursement for psychology
In the United States, many ACHD patients do
not have health insurance that covers psycholog-
ical services. In the Canadian province of Ontario,
outpatient psychology is not covered in the pro-
vincial health plan. Thus, in North America, it is
very unlikely that psychologists working with
ACHD patients will be adequately reimbursed
under a fee-for-service system. Similarly, the
Psychosocial Working Group of the Association
for European Pediatric Cardiology has acknowl-
edged difficulties with reimbursement for mental
health professionals and reported that psychoso-
cial professionals are typically funded through
mixed resources [79]. It is therefore likely that the
inclusion of a psychologist into an ACHD team
would require a strong financial commitment
from the ACHD center.
Summary
The recognition and management of psycho-
social difficulties among ACHD patients deserves
increased attention. Working groups from Europe
and North America have emphasized the inclu-
sion of specialized mental health care for ACHD
patients. There are a number of ways in which
psychologists can be integrated into multidisci-
plinary ACHD care teams. These include the
provision of clinical services, collaboration in
a multidisciplinary research ream, and the pro-
vision of professional education to trainees in
cardiology, nursing, and psychology. Barriers
exist to the integration of psychologists, including
the availability of individuals who have special-
ized cardiac psychology experience and concerns
relating to financial reimbursement. Integration of
psychologists requires a strong commitment from
an ACHD center but would positively impact the
lives of ACHD patients.
616 KOVACS
References
[1] Warnes CA. The adult with congenital heart dis-
ease: born to be bad? J Am Coll Cardiol 2005;46(1):
1–8.
[2] Webb G. Improving the care of Canadian adults
with congenital heart disease. Can J Cardiol 2005;
21(10):833–8.
[3] Rozanski A, Blumenthal JA, Kaplan J. Impact of
psychological factors on the pathogenesis of cardio-
vascular disease and implications for therapy. Circu-
lation 1999;99(16):2192–217.
[4] Carney RM, Freedland KE, Sheline YI, et al. De-
pression and coronary heart disease: a review for
cardiologists. Clin Cardiol 1997;20(3):196–200.
[5] Kovacs AH, Sears SF, Saidi A. Biopsychosocial
experiences of adults with congenital heart disease:
review of the literature. Am Heart J 2005;150(2):
193–201.
[6] Moons P, De Geest S, Budts W. Comprehensive care
for adults with congenital heart disease: expanding
roles for nurses. Eur J Cardiovasc Nurs 2002;1(1):
23–8.
[7] Care of the Adult with Congenital Heart Disease.
Presented at the 32nd Bethesda Conference. Bethesda,
Maryland, October 2–3, 2000. J Am Coll Cardiol
2001;37:1161–98.
[8] Connelly MS, Webb GD, Somerville J, et al. Cana-
dian Consensus Conference on Adult Congenital
Heart Disease 1996. Can J Cardiol 1998;14(3):
395–452.
[9] The Task Force on the Management of Grown Up
Congenital Heart Disease, European Society of Car-
diology, ESC Committee for Practice Guidelines.
Management of grown up congenital heart disease.
Eur Heart J 2003;24(11):1035–84.
[10] Brandhagen DJ, Feldt RH, Williams DE. Long-
term psychological implications of congenital heart
disease: a 25-year follow-up. Mayo Clin Proc 1991;
66(5):474–9.
[11] Horner T, Liberthson R, Jellinek MS. Psychosocial
profile of adults with complex congenital heart dis-
ease. Mayo Clin Proc 2000;75(1):31–6.
[12] Popelova J, Slavik Z, Skovranek J. Are cyanosed
adults with congenital cardiac malformations de-
pressed? Cardiol Young 2001;11(4):379–84.
[13] Bromberg JI, Beasley PJ, D’Angelo EJ, et al. De-
pression and anxiety in adults with congenital heart
disease: a pilot study. Heart Lung 2003;32(2):
105–10.
[14] Utens EM, Verhulst FC, Erdman RA, et al. Psycho-
social functioning of young adults after surgical cor-
rection for congenital heart disease in childhood:
a follow up study. J Psychosom Res 1994;38(7):
745–58.
[15] Utens EM, Bieman HJ, Verhulst FC, et al. Psycho-
pathology in young adults with congenital heart
disease: follow up results. Eur Heart J 1998;19(4):
647–51.
et al
[16] Van Rijen EH, Utens EM, Roos-Hesselink JW, et al.
Psychosocial functioning of the adult with congeni-
tal heart disease: a 20–33 years follow up. Eur Heart
J 2003;24(7):673–83.
[17] Cox D, Lewis G, Stuart G, et al. A cross-sectional
study of the prevalence of psychopathology in adults
with congenital heart disease. J Psychosom Res
2002;52(2):65–8.
[18] van Rijen EH, Utens EM, Roos-Hesselink JW, et al.
Medical predictors for psychopathology in adults
with operated congenital heart disease. Eur Heart J
2004;25(18):1605–13.
[19] van Rijen EH, Utens EM, Roos-Hesselink JW, et al.
Longitudinal development of psychopathology in an
adult congenital disease cohort. Int J Cardiol 2005;
99(2):315–23.
[20] Engel GL. The need for a new medical model: a chal-
lenge for biomedicine. Science 1977;196(4286):
129–36.
[21] Moons P, Marquet K, Budts W, et al. Validity, reli-
ability and responsiveness of the ‘‘Schedule for the
Evaluation of Individual Quality of Life-Direct
Weighting’’ (SEIQoL-DW) in congenital heart dis-
ease. Health Qual Life Outcomes 2004;2:27.
[22] Moons P, Van Deyk K, Marquet K, et al. Individual
quality of life in adults with congenital heart disease:
a paradigm shift. Eur Heart J 2005;26(3):298–307.
[23] Kamphuis M, Ottenkamp J, Vliegen HW, et al.
Health related quality of life and health status in
adult survivors with previously operated complex
congenital heart disease. Heart 2002;87(4):356–62.
[24] Lane DA, Lip GY, Millane TA. Quality of life in
adults with congenital heart disease. Heart 2002;
88(1):71–5.
[25] Simko LC, McGinnis KA. Quality of life experi-
enced by adults with congenital heart disease.
AACN Clin Issues 2003;14(1):42–53.
[26] Rose M, Kohler K, Kohler F, et al. Determinants of
the quality of life of patients with congenital heart
disease. Qual Life Res 2005;14(1):35–43.
[27] Gersony WM, Hayes CJ, Driscoll DJ, et al. Second
natural history study of congenital heart defects.
Quality of life of patients with aortic stenosis, pul-
monary stenosis, or ventricular septal defect. Circu-
lation 1993;87(2 Suppl):I52–65.
[28] Saliba Z, Butera G, Bonnet D, et al. Quality of life
and perceived health status in surviving adults with
univentricular heart. Heart 2001;86(1):69–73.
[29] Immer FF, Althaus SM, Berdat PA, et al. Quality of
life and specific problems after cardiac surgery in ad-
olescents and adults with congenital heart diseases.
Eur J Cardiovasc Prev Rehabil 2005;12(2):138–43.
[30] Fekkes M, Kamphuis RP, Ottenkamp J, et al.
Health-related quality of life in young adults with
minor congenital heart disease. Psychol Health
2001;16(2):239–50.
[31] Rietveld S, Mulder BJ, van Beest I, et al. Negative
thoughts in adults with congenital heart disease.
Int J Cardiol 2002;86(1):19–26.
 ROLE OF PSYCHOLOGISTS
[32] Simko LC, McGinnis KA. What is the perceived
quality of life of adults with congenital heart disease
and does it differ by anomaly? J Cardiovasc Nurs
2005;20(3):206–14.
[33] Ternestedt BM, Wall K, Oddsson H, et al. Quality of
life 20 and 30 years after surgery in patients operated
on for tetralogy of Fallot and for atrial septal defect.
Pediatr Cardiol 2001;22(2):128–32.
[34] Moons P, Van Deyk K, De Geest S, et al. Is the se-
verity of congenital heart disease associated with
the quality of life and perceived health of adult
patients? Heart 2005;91:1193–8.
[35] McMurray R, Kendall L, Parson JM, et al. A life less
ordinary: growing up and coping with congenital
heart disease. Coron Health Care 2001;5(1):51–7.
[36] Kokkonen J, Paavilainen T. Social adaptation of
young adults with congenital heart disease. Int
J Cardiol 1992;36(1):23–9.
[37] Tong EM, Sparacino PS, Messias DK, et al. Grow-
ing up with congenital heart disease: the dilemmas of
adolescents and young adults. Cardiol Young 1998;
8(3):303–9.
[38] Claessens P, Moons P, de Casterle BD, et al. What
does it mean to live with a congenital heart disease?
A qualitative study on the lived experiences of adult
patients. Eur J Cardiovasc Nurs 2005;4(1):3–10.
[39] Kamphuis M, Vogels T, Ottenkamp J, et al. Em-
ployment in adults with congenital heart disease.
Arch Pediatr Adolesc Med 2002;156(11):1143–8.
[40] Somerville J. The woman with congenital heart dis-
ease. Eur Heart J 1998;19(12):1766–75.
[41] Williams RG, Pearson GD, Barst RJ, et al. Report
of the National Heart, Lung, and Blood Institute
working group on research in adult congenital heart
disease. J Am Coll Cardiol 2006;47(4):701–7.
[42] Yusuf S, Hawken S, Ounpuu S, et al, on behalf of
the INTERHEART Study Investigators. Effect of
potentially modifiable risk factors associated with
myocardial infarction in 52 countries (the INTER-
HEART Study): case-control study. Lancet 2004;
364:937–52.
[43] Rozanski A, Blumenthal JA, Davidson KW, et al.
The epidemiology, pathophysiology, and manage-
ment of psychosocial risk factors in cardiac practice:
the emerging field of behavioral cardiology. J Am
Coll Cardiol 2005;45(5):637–51.
[44] Zellweger MJ, Osterwalder RH, Langewitz W, et al.
Coronary artery disease and depression. Eur Heart
J 2004;25(1):3–9.
[45] van Melle JP, de Jonge P, Spijkerman TA, et al.
Prognostic association of depression following myo-
cardial infarction with mortality and cardiovascular
events: a meta-analysis. Psychosom Med 2004;66(6):
814–22.
[46] Eifert GH, Zvolensky MJ, Lejuez CW. Heart-
focused anxiety and chest pain: a conceptual and
clinical review. Clin Psychol Sci Prac 2000;7:403–17.
[47] Pauli P, Wiedemann G, Dengler W, et al. Anxiety in
patients with an automatic implantable cardioverter
617
defibrillator: what differentiates them from panic
patients? Psychosom Med 1999;61(1):69–76.
[48] Rietveld S, Karsdorp PA, Mulder BJ. Heartbeat sen-
sitivity in adults with congenital heart disease. Int
J Behav Med 2004;11(4):203–11.
[49] Price B. Illness careers: the chronic illness experi-
ence. J Adv Nurs 1996;24(2):275–9.
[50] Mueller PS, Hook CC, Hayes DL. Ethical analysis
of withdrawal of pacemaker or implantable cardi-
overter-defibrillator support at the end of life.
Mayo Clin Proc 2003;78:959–63.
[51] Quill TE, Barold SS, Sussman BL. Discontinuing
an implantable cardioverter defibrillator as a life-
sustaining treatment. Am J Cardiol 1994;74(2):
205–7.
[52] Kiecolt-Glaser JK, Page GG, Marucha PT, et al.
Psychological influences on surgical recovery: per-
spectives from psychoneuroimmunology. Am Psy-
chol 1998;53(11):1209–18.
[53] Devine EC. Effects of psychoeducational care for
adult surgical patients: a meta-analysis of 191 stud-
ies. Patient Educ Couns 1992;19(2):129–42.
[54] Johnston M, Vogele C. Benefits of psychological
preparation for surgery: a meta-analysis. Ann Beh
Med 1993;15:245–56.
[55] Seskevich JE, Crater SW, Lane JD, et al. Beneficial
effects of noetic therapies on mood before percutane-
ous intervention for unstable coronary syndromes.
Nurs Res 2004;53(2):116–21.
[56] Gatzoulis K, Frogoudaki A, Brili S, et al. Implant-
able defibrillators: from the adult cardiac to the
grown up congenital heart disease patient. Int J Car-
diol 2004;97:117–22.
[57] Herbst JH, Goodman M, Feldstein S, et al. Health
related quality of life assessment of patients with
life-threatening ventricular arrhythmias. Pacing
Clin Electrophysiol 1999;22:915–26.
[58] Schron EB, Exner DV, Yao O, et al. Quality of life in
the antiarrhythmics versus implantable defibrillators
trial: impact of therapy and influence of adverse
symptoms and defibrillator shocks. Circulation
2002;105(5):589–94.
[59] Sears SF, Conti JB. Quality of life and psychological
functioning of ICD patients. Heart 2002;87(5):
488–93.
[60] Irvine J, Dorian P, Baker B, et al. Quality of life in
the Canadian Implantable Defibrillator Study
(CIDS). Am Heart J 2002;144(2):282–9.
[61] Sears SF, Todaro JF, Lewis TS, et al. Examining the
psychosocial impact of implantable cardioverter de-
fibrillators: a literature review. Clin Cardiol 1999;
22(7):481–9.
[62] Sears SF, Kovacs AH, Azzarello L, et al. Innova-
tions in health psychology: the psychosocial care of
adults with implantable cardioverter defibrillators.
Prof Psychol Res Pract 2004;35(5):520–6.
[63] Reid GJ, Irvine MJ, McCrindle BW, et al. Preva-
lence and correlates of successful transfer from pedi-
atric to adult health care among a cohort of young
618 KOVACS
adults with complex congenital heart defects. Pediat-
rics 2004;113:e197–205.
[64] Dusseldorp E, van Elderen T, Maes S, et al. A meta-
analysis of psychoeducational programs for coro-
nary heart disease patients. Health Psychol 1999;
18(5):506–19.
[65] Reybrouck T, Mertens L. Physical performance and
physical activity in grown-up congenital heart dis-
ease. Eur J Cardiovasc Prev Rehabil 2005;12(5):
498–502.
[66] Kulkarni SP, Alexander KP, Lytle B, et al. Long-
term adherence with cardiovascular drug regimens.
Am Heart J 2006;151(1):185–91.
[67] DiMatteo MR, Lepper HS, Croghan TW. Depres-
sion is a risk factor for noncompliance with medical
treatment. Meta-analysis of the effects of anxiety
and depression on patient adherence. Arch Intern
Med 2000;160(14):2102–7.
[68] Ziegelstein RC, Fauerbach JA, Stevens SS, et al. Pa-
tients with depression are less likely to follow recom-
mendations to reduce cardiac risk during recovery
from a myocardial infarction. Arch Intern Med
2000;160(12):1818–23.
[69] Morgan AL, Masoudi FA, Havranek EP, et al. Dif-
ficulty taking medications, depression, and health
status in heart failure patients. J Card Fail 2006;
12(1):54–60.
[70] Daliento L, Mapelli D, Volpe B. Measurement of
cognitive outcome and quality of life in congenital
heart disease. Heart 2006;92(4):569–74.
[71] Bass JL, Corwin M, Gozal D, et al. The effect of
chronic or intermittent hypoxia on cognition in
et al
childhood: a review of the evidence. Pediatrics
2004;114(3):805–16.
[72] Wray J, Sensky T. Congenital heart disease and car-
diac surgery in childhood: effects on cognitive func-
tion and academic ability. Heart 2001;85(6):687–91.
[73] Daliento L, Mapelli D, Russo G, et al. Health re-
lated quality of life in adults with repaired tetralogy
of Fallot: psychosocial and cognitive outcomes.
Heart 2005;91:213–8.
[74] Griffin KJ, Elkin TD, Smith CJ. Academic outcomes
in children with congenital heart disease. Clin
Pediatr 2003;42(5):401–9.
[75] Kamphuis M, Zwinderman KH, Vogels T, et al.
A cardiac-specific health related quality of life mod-
ule for young adults with congenital heart disease:
development and validation. Qual Life Res 2004;
13(4):735–45.
[76] Lip GY, Lane DA, Millane TA, et al. Psychological
interventions for depression in adolescent and adult
congenital heart disease. Cochrane Database Syst
Rev 2003;3:CD004394.
[77] Bassett AS, Chow EW, Husted J, et al. Clinical fea-
tures of 78 adults with 22ql Deletion Syndrome. Am
J Med Genet 2005;138A:307–13.
[78] Street RL, Gordon HS, Ward MM, et al. Patient
participation in medical consultations: why some
patients are more involved than others. Med Care
2005;43(10):960–9.
[79] Salzer MU. Highlights of the meeting of the Psycho-
social Working Group of the Association for Euro-
pean Paediatric Cardiology. Cardiol Young 2005;
15(1):111–3.
 Cardiol Clin 24 (2006) 619–629

Transition and Transfer from Pediatric to Adult

Care of the Young Adult with Complex
Congenital Heart Disease
Alison Knauth, MD, PhDa,*, Amy Verstappen, MEdb,
John Reiss, PhDc, Gary D. Webb, MDd
a
Boston Adult Congenital Heart Program, Children’s Hospital Boston, Brigham and Women’s Hospital, and Harvard
Medical School, 300 Longwood Avenue, Boston, MA 02115, USA
b
Adult Congenital Heart Association, 6757 Greene Street, Philadelphia, PA 19119, USA
c
Institute for Child Health Policy, University of Florida, PO Box 100147, Gainesville, FL 32610, USA
d
Philadelphia Adult Congenital Heart Center, Children’s Hospital of Philadelphia and the Hospital of the University
of Pennsylvania, 6 Penn Tower, 3400 Spruce Street, Philadelphia, PA 19104, USA

Many children afflicted with complex child-
hood illnesses, which historically caused early
death, are now surviving childhood with the
potential for meaningful and productive adult
lives. Although most of these patients eventually
transfer their care from a pediatric to an adult
health care environment, the process of transition-
ing, or preparing the patient and family for this
transfer, faces many obstacles. As a result transi-
tion is neither easy to orchestrate nor simple to
perform. An absent or inadequate transition pro-
cess may result in delayed and inappropriate care,
improper timing of the transfer of care, and undue
emotional and financial stress for the patients,
their families, and the health care system. At its
worst, commonly patients are lost to appropriate
follow-up. To avoid these hazards, physicians
managing these young patients must recognize
the importance of the transition process, an
educational and experiential process that prepares
patients to take responsibility for their own health
care. This transition is an important and appro-
priate process for these young people and marks
a significant period in their development as
autonomous patients. It ends by seeing the young
patients transfer to an appropriate adult health
care setting, one that is better suited to addressing
their evolving adult needs and that must continue
to do so for the rest of their lives. Performed
correctly, the process is challenging but rewarding
for these young people and provides them with the
opportunity to maximizing their future physical
and psychosocial well-being.
This article focuses first on the process of
transition (Box 1) and then on the transfer of
care for young adults who have chronic childhood
illness in general and complex congenital heart
disease in specific. It defines the transition process
and briefly discusses its history. It then reviews the
important aspects of transition, outlines the key
elements of a successful transition program, and
provides a general curriculum appropriate for
the young adult with congenital heart disease.
Last, it identifies barriers to the transfer of care,
discusses the importance of a policy on the timing
of transfer, outlines briefly the components of
adult provider services that may be needed, and
reviews the steps to an orderly transfer process.

* Corresponding author. Department of Cardiology, Definition of transition and transfer

Children’s Hospital Boston, 300 Longwood Avenue,
Boston, MA 02115. Transition of care, as referred to in this article,
E-mail address: alison.knauth@cardio.chboston.org is the process by which adolescents and young
(A. Knauth). adults who have chronic childhood illnesses are
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.010 cardiology.theclinics.com
620 KNAUTH
Box 1. Key elements of a transition
and transfer program
 Development of program for patient
preparation and education
 Well-structured curriculum
 Choice of multiple educational
modalities
 Clinic visits
 Educational seminars
 Peer support groups
 Identification of obstacles to transfer of
care
 Patients
 Families
 Pediatric caregivers
 Adult caregivers
 Establishment of (flexible) policy on
timing of transfer
 Identification of capable and available
adult providers
 Plan for a coordinated transfer
 Coordinator
 Written health summary
 Direct communications between
caregivers
prepared to take charge of their lives and their
health in adulthood. It is an educational process
that ideally begins before children reach adoles-
cence and continues until they are capable of
taking full responsibility for their care. Most
importantly, young patients who have complex
congenital heart disease are taught that, although
they have the potential to live healthy and pro-
ductive lives, they have been ‘‘repaired’’ and not
‘‘cured’’; they require lifelong cardiac surveillance,
and they are responsible for obtaining appropriate
care.
Transfer defines an event or series of events
through which adolescents and young adults who
have chronic physical and medical conditions
move their care from a pediatric to an adult
health care environment. It is important that
transfer be deferred in the absence of appropriate
adult resources. If appropriate adult care is avail-
able, however, transfer is likely to be beneficial to
patients. Although a pediatric care model is ideal
for young children, as it focuses on growth and
et al
development and directs education toward the
parents or guardians, it is no longer appropriate
for adult patients. The adult care model, in
contrast, employs a visit dynamic that is a more
of a partnership with education directed toward
the patients, thus encouraging independence, re-
sponsibility, and self-reliance. The goal is a smooth
transfer following a structured transition program
provided by pediatric and adolescent providers.
A brief history of transition
Health care transition for young people who
have chronic health conditions has long been
recommended [1–3]. In the United States, multiple
invitational conferences and task forces have iden-
tified the problems faced by the growing popula-
tion of adults who had chronic childhood
illnesses and their need for formal transition pro-
grams. In 2003, the American Academy of Pediat-
rics, American Academy of Family Physicians,
and American College of Physicians presented
a consensus statement on health care transition
for young adults who have special health care
needs [4]. This statement was endorsed by the So-
ciety for Adolescent Medicine in its 2003 position
paper [5]. Similarly, in the United Kingdom, the
transition of young adults who have chronic child-
hood illnesses to the adult health care system was
addressed explicitly by the Royal College of Pedi-
atrics and Child Health in recently released guide-
lines on health care for adolescents [6].
Specialty professional organizations also have
weighed in on the need for transition programs
with editorials, practice guidelines, and position
statements related to the transition of young
people who have specific health concerns. In
particular, the needs of adults who have congen-
ital heart disease and recommendations for tran-
sition and transfer of care have been formalized
by a number of expert task forces including the
32nd Bethesda Conference convened by the
American College of Cardiology [7–11], the Cana-
dian Cardiovascular Society Consensus Confer-
ence [12–14], and the British Cardiac Society
[6,15]. Perhaps even more compelling is that pa-
tients, most notably the Adult Congenital Heart
Association, are asking for transition programs
by organizing work groups to address these issues
and to promote the formation of programs across
the country.
Despite the recognized need for transition
programs, most adolescent patients still do not
have the opportunity to participate in such
 TRANSITION AND TRANSFER IN CHD
a process. It is hoped that this situation will
change as the endorsement of health care transi-
tion grows, broader ranges of transition experi-
ence are described, and steps are taken to evaluate
such programs formally [16–22].
Goals of the transition process
Few would argue with the appropriateness of
and benefits associated with preparing young
people to take charge of their own health and
lives. Nevertheless, the smooth transition to this
state remains a challenge for everyone involved.
With this challenge in mind, formal structured
transition programs have been proposed [5,6,22–
24]. The goals of such programs are several. First
and foremost, they aim to provide uninterrupted
health care that is patient centered, age and devel-
opmentally appropriate, flexible, and comprehen-
sive. Young people are coached to make a smooth
transfer from a pediatric to an adult health care
environment without becoming lost to follow-up.
Transition programs are committed to educating
young people about their medical conditions and
promoting skills in communication, decision mak-
ing, self-care, and self-advocacy. As a result, pa-
tients become capable of personal and medical
independence and gain a greater sense of control
over their health, health care decisions, and psy-
chosocial environment. The ultimate goal of tran-
sition programs is to optimize the quality of life
and future potential of young patients [2].
Transition needs of adults who have congenital
heart disease
Adolescents who have congenital heart disease
constitute a growing population of individuals
who have such complex medical issues that a well-
planned and well-executed transition process can
be valuable. Because of advances in pediatric
cardiovascular surgery, percutaneous interven-
tional therapies, intensive care, and medical man-
agement, many children who have complex
congenital heart disease are now surviving into
adulthood. As a result, the number of adults who
have congenital heart disease in the United States
is rising steadily and currently nears 1 million.
Half of these patients have complex congenital
heart conditions for which most adult cardiolo-
gists have not been trained. In addition, they often
present with complex comorbidities and psycho-
social challenges. These patients need lifelong care
621
by specialized practitioners who are fluent in the
complex anatomy and physiology of congenital
heart disease and who are well versed in the lexis
of adult internal medicine [7–11,25].
Models of transition
A variety of transition models have been pro-
posed, including generic and disease-based
models, and transition recommendations have
been elaborated and published. The generic tran-
sition model employs adolescent medicine services
to run generic transition programs designed to
address general adolescent and transition issues
while relying on subspecialty programs to handle
specific medical issues. The disease-based model
carries out the transition process within a program
that specializes in the medical needs of a particular
type of childhood illness. This model has the
benefit of being able to tailor the transition
process to the individual patient [23,26]. The dis-
ease-based model may be best for young adults
who have congenital heart disease, given the het-
erogeneity of the population and the complexity
of their medical conditions. In the absence of
such a program, much of the responsibility for
transition falls on pediatric providers using a less
formal approach.
Key elements of transition
Regardless of the model, fundamental princi-
ples of transition have achieved nearly universal
endorsement [6,19,23]. These principles provide
a framework for individual programs and institu-
tions whose goal is to improve the transition expe-
rience for their young adult patients who have
chronic health conditions. The following discus-
sion outlines what the authors believe are the
key elements that must be incorporated into
a transition program. These key elements begin
with patient preparation and education using
a well-structured curriculum, require the recogni-
tion of the obstacles to transition and transfer
that must be proactively addressed, and depend
on a policy regarding the timing of transfer, the
identification of appropriate adult provider ser-
vices, and a coordinated transfer process.
Preparation and education
Although a formal transition curriculum is
most appropriate during late childhood and
adolescence, preparation for successful transition
622 KNAUTH
should begin early and continue throughout child-
hood. A major potential barrier to successful
transition is the mistaken perception of families
and patients that a ‘‘cure’’ has resulted from
successful cardiac surgery. Such misperceptions
should be solicited regularly and addressed di-
rectly. Patients and families must understand that,
although in many cases their children are likely to
have a normal or near-normal life, they do not
have a normal heart and will require lifelong
surveillance. Furthermore, they must be taught
that congenital heart disease is vastly different
from other forms of heart disease, making it
imperative that they follow up with providers
specializing in adult congenital heart disease.
From early in the course of care, families should
have the expectation that their children will be
independent in managing their own medical care
by young adulthood. Families should be encour-
aged to help their children reach this goal. Pro-
viders must help families build their children’s
knowledge of their heart disease using age-appro-
priate strategies [23,27,28]. Core concepts should
be revisited regularly, because medical concepts
are complex, and educational research has shown
that repetition greatly enhances retention and un-
derstanding [29]. Ideally, using such a strategy of
early preparation, the patients and families will
enter the transition period with an understanding
of their heart disease and the need for lifelong
specialized care.
Building on this preparation, the transition
curriculum should ensure that these young people
understand their diagnosis and medical history.
Ideally, they must understand the normal heart
(basic anatomy and physiology), how the heart they
were born with differs from normal, and what
surgical or catheter-based interventions were un-
dertaken to ‘‘repair’’ their heart. Because these
young people already understand that ‘‘repair’’
does not mean ‘‘cure,’’ they will be ready to learn
about the risks of residual hemodynamic burdens
and arrhythmic complications. They should be
taught to recognize important signs and symptoms
that may alert them to such complications. The
transitioning adolescents should have a good un-
derstanding of the rationale for previous therapies
and options for future medical, surgical, and
catheter-based therapies. Finally, they need to be
prepared to navigate the adult health care system:
how to access an adult congenital heart specialist,
how often they need such follow-up, how to access
routine health care, how to access emergency health
care, and how to navigate the insurance process.
et al
The transition curriculum
A standard core educational program (Box 2)
is an important component of an organized tran-
sition process. The goals of this curriculum can be
achieved using a variety of educational modalities
and should be tailored to individual patients. The
curriculum should be of appropriate breadth and
depth but concise enough to be completed before
patients move into the adult congenital heart dis-
ease clinic. Completion of the curriculum may re-
quire two clinic visits for mature, well-adjusted
patients who have mild disease, mild functional
limitations, and a strong support system or
many visits with intense involvement of a peer
support group for a young person who has learn-
ing disabilities, complex disease, significant func-
tional limitations, and no support system.
Flexibility is a key to success.
The elements of a transition curriculum include
both medical and nonmedical issues and are
outlined below. The curriculum goals can be
accomplished in a number of different ways.
Much of the curriculum can be addressed at
routinely scheduled clinic visits with the partici-
pation of physicians, midlevel practitioners, and
support services. Additionally, a transition pro-
gram can design introductory seminars for pa-
tients and families or use peer support groups and
informational seminars to address some of the
components of the transition curriculum.
Residual hemodynamic considerations
Most patients who have repaired congenital
heart disease have residual hemodynamic bur-
dens. Honest discussions are required regarding
these residual hemodynamic issues, their implica-
tions for the future, and the signs and symptoms
that should raise concern. Patients should be told
about future diagnostic testing required to keep
these matters under appropriate surveillance. In
addition, they should be educated about the
possible medical, catheter-based, or surgical op-
tions for treating these problems if or when they
are needed.
It also is important to discuss new advances in
cardiovascular medicine including advances in
medical management, cardiac catheterization,
and cardiovascular surgery. Such discussions
often can be accomplished effectively in peer
support groups and informational seminars.
Through this education, young people can de-
velop a positive and hopeful outlook for the
future despite the knowledge of potential future
 TRANSITION AND TRANSFER IN CHD 623

Box 2. Transition curriculum topics  Communication

 Decision making
 Residual hemodynamic considerations
 Creative problem solving
 Hemodynamic issues
 Assertiveness
 Symptoms and how to respond
 Self-care
 Diagnostic tools in follow-up
 Self-advocacy
 Management options
 Contraception and pregnancy
planning complications. Giving patients tools to learn
 Contraceptive options and risks about such therapies on their own and encourag-
 Risks of pregnancy to mother and to ing them to seek new knowledge continuously
fetus enables them to feel a sense of control over their
 Management of pregnancy plan health and medical management.
 Endocarditis considerations
 Risks, implications, recognition, and Contraception and pregnancy planning
response It is important that discussions about contra-
 Prevention ception and pregnancy occur before patients reach
 Arrhythmia considerations adolescence. It is common for young people to
 Risks transfer abruptly to the adult health care system
 Signs and symptoms as a result of an unplanned pregnancy. Pregnancy,
 Screening tools especially in a woman who has complex congen-
 Diagnostic tools ital heart disease, should be well planned to
 Management options minimize risks to the mother and unborn child.
The young person should have the opportunity to
 Noncardiac surgery considerations
make an informed decision about if and when to
 Risks
become pregnant. For this reason, contraceptive
 Location of surgery
options should be discussed with patients before
 Knowledge and skills of surgical and
they become sexually active. During these conver-
ICU team
sations, it must be emphasized that the various
 Noncardiac medical problems contraceptive options carry both benefits and
 Access to appropriate care risks. For example, estrogen-containing oral con-
traceptive pills increase the risk of thromboembo-
 Career and vocational planning
lism and may not be recommended in certain
 Lifestyle issues patients. Beyond pregnancy planning or preven-
 Marriage and family planning tion, information about barrier methods for pre-
 Education vention of sexually transmitted diseases always
 Employment should be provided.
 Life and health insurance Pregnancy is feasible for many adult patients
 Learning disabilities who have congenital heart disease, but the poten-
 Anxiety and depression tial risks to the mother and unborn child should
 High-risk behaviors be discussed before conception. Frank discussions
 Healthy eating should be had regarding the severity of the
 Physical fitness patients, underlying disease and relative risk of
 Salt and fluid restriction complications during pregnancy, delivery, and
(if warranted) postpartum, as well as possible risks to the fetus
 Relative safety of exercise and such as poor growth, premature labor and deliv-
hobbies ery, and the chance of congenital heart disease.
Fetal echocardiography and genetic counseling
 End-of-life decisions should be offered in selected cases. Finally, the de-
 Skills training velopment of a realistic plan regarding the timing
of and most appropriate number of pregnancies
may be desirable.
624 KNAUTH
Endocarditis considerations
Few adolescents who have congenital heart
disease understand what endocarditis is and why
they are more likely than the general population
to suffer this complication [27]. As a standard part
of a transition curriculum, patients should be ed-
ucated about the implications of endocarditis.
They should learn how to reduce the risk of endo-
carditis through attention to oral health and the
use appropriate antibiotic prophylaxis. Finally,
they should be able to recognize the signs and
symptoms of endovascular infection so they
know to seek medical attention promptly.
Arrhythmia considerations
Arrhythmias are a relatively common compli-
cation experienced by patients who have congen-
ital heart disease, and patients should be educated
about them when relevant. It is important that
patients understand what atrial and ventricular
arrhythmias are, their causation, and the relative
risk or hazard of each (tailored to each individual
patient). An explanation and understanding of
which arrhythmias are simply an annoyance,
which may be life threatening, and how to
recognize the signs and symptoms that should
raise concern are necessary. Patients should un-
derstand the importance of screening tools and
the potential need for other diagnostic tests if
concerning signs or symptoms arise. Finally,
arrhythmia management, including medications,
radiofrequency catheter ablation, surgical abla-
tion, and device therapies (pacemakers and de-
fibrillators), should be discussed as needed.
Noncardiac surgery considerations
Patients need to be advised about the issues
regarding noncardiac surgery. Important consid-
erations include the need for the treating team to
be informed fully about the patients, heart condi-
tion. Except for the simplest outpatient proce-
dures, noncardiac surgeries in complex patients
should be performed in a tertiary hospital where
cardiac anesthesiologists, as well as other mem-
bers of the surgical and critical care teams, are
comfortable with the care of adults who have
congenital heart disease. Knowledge about these
issues allows patients to serve as their own
advocates.
Noncardiac medical problems
Discussions about noncardiac medical prob-
lems must be tailored to the individual patient.
et al
Patients should be aware of the implications of
these issues and know how to access appropriate
care.
Career, vocational, and insurance planning
Many adolescents who have congenital heart
disease will be able to plan their education and
careers in a normal fashion. The futures of some
will include physical or other limitations, how-
ever, and they should be advised to pursue
education and careers that are sustainable even
if physical limitations develop. It is important that
this part of the process begin before patients reach
high school, when they will make academic
choices that can limit future options. In addition,
it is important to teach these young patients that
adults who have complex congenital heart disease
are likely to have difficulties securing health
insurance independently, and therefore obtaining
a health plan through an employer may be
optimal. Life insurance poses similar obstacles,
and often early and disciplined self-investment is
desirable.
Lifestyle issues
Because the goal of all medical care is to
improve the quality of life for the patient,
a transition curriculum must at some point depart
from the medical realm and include discussions
about lifestyle and quality of life. These include
marriage and family planning, education, employ-
ment, and life and health insurance. Other chal-
lenges faced by young people who have congenital
heart disease, such as learning disabilities, anxiety
and depression, and high-risk behaviors, including
use of tobacco, alcohol, and other drugs of abuse,
should be acknowledged. It is important that
these discussions be held in private, in a safe and
comfortable environment. The importance of
healthful eating and physical fitness, the risks of
poor nutrition and obesity, and the need for salt
and fluid restriction (if warranted) should be
addressed. Education regarding appropriate diet
and exercise can be accomplished effectively in
informational seminars and peer support groups.
Finally, the relative safety of exercise and other
hobbies should be discussed.
End-of-life decisions
Anyone may have to make end-of-life de-
cisions. Young people who have chronic child-
hood illnesses may face these decisions sooner
than their healthy peers. It is well known that
 TRANSITION AND TRANSFER IN CHD
these decisions are made most effectively during
a time of health rather than urgently at a time of
acute illness or worsening chronic disease. For this
reason discussions about these decisions should be
included in the transition process. In addition,
informational seminars and peer support groups
may be effective ways for young people to discuss
their ideas, fears, and plans.
Skills training
The transition program curriculum should in-
clude skills training in communication, decision
making, creative problem solving, assertiveness,
self-care, self-determination, and self-advocacy.
Much of this training can be accomplished in
informational seminars and peer support groups
and reinforced at clinic visits.
Obstacles to transfer of care
Although many barriers exist to the successful
transfer of congenital heart patients out of pedi-
atric cardiac care, only a few young adults who
have complex congenital heart problems continue
to receive care from their pediatric cardiologists.
Currently, the majority of adults who have
congenital heart disease experience unplanned
transfer without transition; they discontinue care
with their pediatric team at some point in adoles-
cence or young adulthood and later transfer to an
adult health care environment without prepara-
tion, education, or a plan. Research suggests that
the majority of adults who have complex congen-
ital heart disease are either lost from cardiac care
entirely or are cared for by cardiologists without
specialized training [30,31].
For patients still being seen in the pediatric
care system, there are many obstacles to a smooth
transfer process, both intrinsic and extrinsic to the
physician–patient team, and a successful transfer
must address these issues [1,23,32]. Obstacles that
lie outside the patient–physician relationship are
often the most obvious. These obstacles include
problems accessing competent adult care that
can match the quality of care and range of services
available in the pediatric environment as well as
financial and insurance issues.
Less obvious, and often more difficult to
address, are obstacles that lie within the physi-
cian–patient axis. First, the adolescents them-
selves may impede their own progress.
Transferring care may be traumatic for the young
person. The challenge of accepting their new
625
responsibility for their own health and lives can
be overwhelming, and adolescents may subcon-
sciously see the transfer of care as a danger rather
than an opportunity. Moreover, at this time of
personal adolescent struggle, they are forced to
give up their relationships with trusted and
respected pediatric care providers and are ex-
pected to develop new relationships with an un-
known group. They may be expected to transfer
from a pediatric facility within which they feel
comfortable to an adult facility that is anything
but coddling. Finally, the self-image of young
people who have chronic illness is often infantil-
ized, promoting dependence. The transfer of care
to the adult health care system challenges patients
to become independent and may be resisted.
Similarly, patients’ families can be unwitting
obstacles to the transfer of their children’s care.
Adult care models usually take an individual
approach rather than a family approach, and
although this approach encourages patient auton-
omy and self-dependence, the families of these
patients often feel excluded from this new dy-
namic. Some parents perceive this difference in
approach as a loss of control and thus resist any
change in the status quo, rather than viewing
transfer as a positive step toward greater auton-
omy and self-sufficiency for their maturing ado-
lescent. In addition, families are often concerned
that their children will not be capable of caring for
their illness independently, further contributing to
resistance to the transfer. Finally, like their
adolescent children, families are reluctant to
change from the known and trusted pediatric
providers and facilities to the unknown adult
system. They often are concerned that the care
will be impersonal or less skilled. For these
reasons, a successful transition program must
explicitly acknowledge and address the expected
shift in family dynamics and be prepared to assist
and counsel families who have particular difficulty
in stepping back and letting go of their previous
intimate role in their children’s care. It must be
emphasized to the families that the transfer of care
is ultimately a positive step for their children
toward reaching their full potential.
The pediatric provider also can be an obstacle
to the transition process. Too often the provider is
unable to relinquish the intense and important
relationship that has developed with these patients
and families. The impact that the loss of such
relationships imposes on the pediatric providers is
often underestimated. In addition, pediatric pro-
viders may be skeptical about the knowledge,
626 KNAUTH
skills, and experience of the adult providers and
may communicate this attitude to their young
patients in explicit or implicit ways, making them
wary of transfer. The point should be emphasized
again: important though transfer may be, it
always should be to competent adult care. It
would be unacceptable to put patients at risk by
transferring their care to unskilled providers.
Finally, the adult service may pose a significant
obstacle. Adult providers may be intimidated by
the complexity of these patients and may recog-
nize the care of these patients as a financial
liability. When willing to take on such complex
patients, adult providers may wish to embark on
an extensive reevaluation or a prompt change in
management, which can be unsettling to patients
and families. The transfer of medical records and
communication between pediatric and adult pro-
viders is not always well done and can compound
these problems. Anticipation and avoidance of
these obstacles to transfer is preferable to strug-
gling with them as they arise.
The timing of transfer
Ideally, transfer of care from a pediatric to an
adult health care system occurs at the successful
completion of a thoughtful transition process.
When deciding on the timing of transfer, two
important points need to be considered. First,
there should be a policy on timing to ensure that
transition and transfer actually occur and occur in
a predictable manner [23]. Second, despite the pol-
icy on timing, a transition program must be flexi-
ble and should tailor the transfer process and its
timing to the developmental and psychosocial sta-
tus of each young person [26,32].
Each institution requires a policy on the age at
which its patients will transfer to appropriate
adult care. It is important for care providers,
patients, and families to have an explicit target
age. With such a policy, patients and families
recognize that transition and transfer will occur,
appreciate their active involvement in the de-
cisions about timing, and are prepared for the
ultimate transfer of care. Additionally, when
transition is the rule, young people see the process
not as something they as individuals are forced to
go through alone but rather a natural process that
everyone experiences: much as high school follows
junior high school, so adult care follows pediatric
care. Finally, a policy on timing ensures that
everyone involved in the process has the time and
et al
opportunity to take full advantage of the entire
transition curriculum.
Uncompromising structure frequently is only
marginally better than chaos, and neither has
a place in a well-organized transition process. It
therefore is important to remember that the
presence of structure in this process should never
preclude rational plasticitydthe timing of transfer
must be flexible. Important decisions about timing
must be individualized to each young person and
should occur in a developmental context. Tran-
sitioning should be introduced early, as previously
emphasized, and should be completed only after
the young patients have accomplished the de-
velopmental tasks of adolescence and have dem-
onstrated the ability to manage their health care
independently of their families and pediatric
providers.
Early introduction of the concepts of transition
and transfer to patients and their families is
critical [16,17] and should occur before the child
enters the developmental stage of adolescence.
At the initiation of these discussions, it may be
helpful for the patients and families to write an in-
dividualized transition plan that includes a time-
line. This written document can be reviewed and
revised at intervals. In this way, the patients and
families participate actively in the transition pro-
cess and are aware and prepared for the ultimate
transfer of care.
Adolescents have many difficult tasks to
accomplish before leaving this stage in their
development, namely, to develop a personal sense
of identity and adjust to this new physical sense of
self; to adjust to new intellectual abilities and the
increased cognitive demands placed on them at
school and in society in general; to develop
educational and vocational goals for themselves
and a plan to meet these goals; to establish
emotional and psychologic independence from
their parents and begin to adopt a personal value
system; to learn to manage their sexuality and
begin to develop intimate relationships; and to
develop increased impulse control and behavioral
maturity. Chronic illness, physical disability, and
cognitive limitations can disrupt the usual de-
velopmental trajectories. This interruption makes
the accomplishment of such tasks more difficult,
and thus the process of transition may be more
challenging [24,26].
Furthermore, the sometimes overprotective
attitudes of pediatric providers and patients’
families can diminish self-esteem and hinder the
development of self-sufficiency. In general, the
 TRANSITION AND TRANSFER IN CHD
adult health care system demands a higher level of
personal responsibility and autonomy than the
pediatric system. Thus, to be effective health care
consumers in the adult health care environment,
young people should demonstrate the ability to
meet their health care needs independently of their
families and pediatric providers before transfer.
They need to have the ability to make their own
appointments, meet independently with their
health care providers, administer their own med-
ications and other treatments, understand their
medical history, and recognize signs of clinical
deterioration.
Finally, transfer of care during medical crises
or periods of psychosocial disequilibrium should
be avoided [23]. All too often patients are trans-
ferred when they become pregnant or after they
suffer their first ‘‘adult’’ complication. Transfer
during these times precludes maximal use of the
transition curriculum and imposes a tremendous
psychologic burden on the existing crisis. A struc-
tured policy on timing of transfer helps ensure
that the process is begun before many of these
crises have the opportunity to develop and the
flexibility to defer transfer until acute issues are
addressed; thus a structured policy reduces the
psychosocial burden for patients.
Adult provider services
Although the needs of some patients are
relatively straightforward, other patients have
complex needs and may require the involvement
of a variety of consultants, including an electro-
physiologist, psychiatrist, obstetrician, gynecolo-
gist, gastroenterologist, hepatologist, nephrologist,
pulmonologist, neurologist, or oncologist. Other
medical and nonmedical professionals, including
midlevel providers, social workers, and those who
specialize in vocational and educational issues, are
also critical to successful transition [8,10,23].
In addition, young people who have chronic
health conditions require capable primary health
care providers willing to coordinate and manage
the complex care they need [6]. Young people who
have chronic conditions share many of the same
health issues and concerns as their peers, namely,
growth and development; sexuality; depression
and other mental health disorders; substance
use; and other high-risk behaviors. The primary
care providers who take on this population of pa-
tients must be prepared to address these issues
themselves or to ensure they are being addressed
in other settings. The specialist often will not
627
address other medical issues; thus the role of pri-
mary care providers should be defined explicitly.
Communication between specialists and primary
care providers is paramount.
Coordinated transfer process
A coordinated transfer process is the final
component of a successful transition process
[17,22,23]. As the time for transfer approaches,
preparations need to be made to ensure the pro-
cess goes smoothly. By now, the patients and their
families have been prepared through the transi-
tion program and are ready for the transfer of
care. A carefully prepared health summary allows
seamless transfer of care and provides a blueprint
for the new health care team, especially early after
transfer. A comprehensive summary should in-
clude a complete medical history including diag-
noses and previous interventions, a medication
list, laboratory values, and other diagnostic stud-
ies, as well as information about the patient’s
functional status, the tempo of disease progres-
sion, and the impact of other comorbidities. Psy-
chosocial concerns, end-of-life preferences, the
extent of family involvement, and adherence is-
sues should also be communicated if available.
A comprehensive summary avoids the need to
reinvent the therapeutic wheel and helps prevent
errors being made or mistakes being repeated. In
addition to the formal medical summary, it is vital
that the pediatric providers, including the special-
ist physicians, physical and occupational thera-
pists, dietitians, social workers, nurses, and other
health-care professionals, communicate directly
with their counterparts in the adult health care
system.
It is important that the adult providers respect
the therapeutic plan established by the pediatric
providers and communicated to the young pa-
tients and their families. As explained earlier, the
immediate reevaluation and drastic change in
management that many new adult providers are
tempted to make after care transfer often can be
overwhelming to the patients and diminish their
trust. Nonetheless, one must also take advantage
of the unique opportunity this transition process
provides to take a fresh look at the patients as
evolving adults and to reassess therapeutic op-
tions in the light of new technologies. This process
can be done in conjunction with the pediatric
providers and after a period of introduction and
relationship development in the transition
628 KNAUTH
process. In this way trust is not lost, and the best
possible plan for the future is made.
Many transition tools have been suggested to
aid in a smooth transition process. For example,
a central care provider who can assume responsi-
bility for progress through the process of transition
and transfer has been shown to be helpful [17,22].
This coordinator can share membership on both
the pediatric and adult teams and serve as a liaison
between them, as well as a reassuring presence and
advocate for families. In addition, there are a num-
ber of ways to help address the inevitable uncer-
tainty felt by young people and their families
about impending transfer. Many programs have
created a transfer package with information and
details about the adult program, including its
providers and procedures [17]. Patients who have
already transferred to the adult program some-
times can play a part in welcoming young people
who are newly transferring their care.
Summary
Because increasing numbers of young people
who have complex congenital illnesses are surviv-
ing into adulthood, there is an urgent need for
programs designed to facilitate their smooth
movement from pediatric to adult health care
environments. This article has identified the im-
portant constituents of the transition process and
has provided guidelines to the successful transfer
of the patient to the adult health care environ-
ment. In the near future, it is hoped, transition
programs will become the standard of care,
making it more likely that patients who have
complex congenital heart disease can achieve their
full potential under appropriate medical surveil-
lance and live meaningful and productive lives.
References
[1] Schidlow DV, Fiel SB. Life beyond pediatrics. Tran-
sition of chronically ill adolescents from pediatric to
adult health care systems. Med Clin North Am 1990;
74(5):1113–20.
[2] Blum RW, Garell D, Hodgman CH, et al. Transition
from child-centered to adult health-care systems for
adolescents with chronic conditions. A position
paper of the Society for Adolescent Medicine.
J Adolesc Health 1993;14(7):570–6.
[3] Rosen DS. Transition from pediatric to adult-
oriented health care for the adolescent with
chronic illness or disability. Adolesc Med 1994;
5(2):241–8.
et al
[4] A consensus statement on health care transitions for
young adults with special health care needs. Pediat-
rics 2002;110(6 Pt 2):1304–6.
[5] Rosen DS, Blum RW, Britto M, et al. Transition to
adult health care for adolescents and young adults
with chronic conditions: position paper of the Soci-
ety for Adolescent Medicine. J Adolesc Health
2003;33(4):309–11.
[6] Grown-up congenital heart (GUCH) disease: cur-
rent needs and provision of service for adolescents
and adults with congenital heart disease in the UK.
Heart 2002;88(Suppl 1):i1–14.
[7] Warnes CA, Liberthson R, Danielson GK, et al.
Task force 1: the changing profile of congenital heart
disease in adult life. J Am Coll Cardiol 2001;37(5):
1170–5.
[8] Foster E, Graham TP Jr, Driscoll DJ, et al. Task
force 2: special health care needs of adults with con-
genital heart disease. J Am Coll Cardiol 2001;37(5):
1176–83.
[9] Child JS, Collins-Nakai RL, Alpert JS, et al.
Task force 3: workforce description and educa-
tional requirements for the care of adults with
congenital heart disease. J Am Coll Cardiol
2001;37(5):1183–7.
[10] Landzberg MJ, Murphy DJ Jr, Davidson WR Jr,
et al. Task force 4: organization of delivery systems
for adults with congenital heart disease. J Am Coll
Cardiol 2001;37(5):1187–93.
[11] Skorton DJ, Garson A Jr, Allen HD, et al. Task
force 5: adults with congenital heart disease: ac-
cess to care. J Am Coll Cardiol 2001;37(5):
1193–8.
[12] Therrien J, Dore A, Gersony W, et al. Canadian
Cardiovascular Society consensus conference 2001
update: recommendations for the management of
adults with congenital heart disease. Part I. Can J
Cardiol 2001;17(9):940–59.
[13] Therrien J, Gatzoulis M, Graham T, et al. Canadian
Cardiovascular Society consensus conference 2001
update: recommendations for the management of
adults with congenital heart diseasedpart II. Can
J Cardiol 2001;17(10):1029–50.
[14] Therrien J, Warnes C, Daliento L, et al. Canadian
Cardiovascular Society consensus conference 2001
update: recommendations for the management of
adults with congenital heart disease: part III. Can J
Cardiol 2001;17(11):1135–58.
[15] British Cardiac Society and the Royal College of
Physicians of London. The future of paediatric car-
diology in the United Kingdom. Br Heart J 1992;
68(6):630–3.
[16] Boyle MP, Farukhi Z, Nosky ML. Strategies for im-
proving transition to adult cystic fibrosis care, based
on patient and parent views. Pediatr Pulmonol 2001;
32(6):428–36.
[17] Madge S, Bryon M. A model for transition from pe-
diatric to adult care in cystic fibrosis. J Pediatr Nurs
2002;17(4):283–8.
 TRANSITION AND TRANSFER IN CHD
[18] Kipps S, Bahu T, Ong K, et al. Current methods of
transfer of young people with Type 1 diabetes to
adult services. Diabet Med 2002;19(8):649–54.
[19] Fernandes SM, Landzberg MJ. Transitioning the
young adult with congenital heart disease for life-
long medical care. Pediatr Clin North Am 2004;
51(6):1739–48 [xi.].
[20] Reid GJ, Irvine MJ, McCrindle BW, et al. Preva-
lence and correlates of successful transfer from pedi-
atric to adult health care among a cohort of young
adults with complex congenital heart defects. Pediat-
rics 2004;113(3 Pt 1):e197–205.
[21] Williams WG, McCrindle BW. Practical experi-
ence with databases for congenital heart disease:
a registry versus an academic database. Semin
Thorac Cardiovasc Surg Pediatr Card Surg Annu
2002;5:132–42.
[22] Viner R. Barriers and good practice in transition
from paediatric to adult care. J R Soc Med 2001;
94(Suppl 40):2–4.
[23] Viner R. Bridging the gaps: transition for young peo-
ple with cancer. Eur J Cancer 2003;39(18):2684–7.
[24] Hergenroeder AC. The transition into adulthood for
children and youth with special health care needs.
Tex Med 2002;98(2):51–8.
[25] Murphy DJ Jr, Foster E. ACCF/AHA/AAP recom-
mendations for training in pediatric cardiology.
Task force 6: training in transition of adolescent
care and care of the adult with congenital heart dis-
ease. J Am Coll Cardiol 2005;46(7):1399–401.
629
[26] Bryon M, Madge S. Transition from paediatric to
adult care: psychological principles. J R Soc Med
2001;94(Suppl 40):5–7.
[27] Moons P, De Volder E, Budts W, et al. What do
adult patients with congenital heart disease know
about their disease, treatment, and prevention of
complications? A call for structured patient educa-
tion. Heart 2001;86(1):74–80.
[28] Webb GD. Challenges in the care of adult patients
with congenital heart defects. Heart 2003;89(4):
465–9.
[29] Bransford JDBA, Cocking RR. How people
learn: brain, mind, experience, and school. Com-
mittee on Developments in the Science of Learn-
ing, Commission on Behavioral and Social
Sciences and Education, National Research
Council. Washington (DC): National Academy
Press; 1999. p. 39–67.
[30] Dore A, de Guise P, Mercier LA. Transition of care
to adult congenital heart centres: what do patients
know about their heart condition? Can J Cardiol
2002;18(2):141–6.
[31] Wacker A, Kaemmerer H, Hollweck R, et al.
Outcome of operated and unoperated adults
with congenital cardiac disease lost to follow-up
for more than five years. Am J Cardiol 2005;
95(6):776–9.
[32] Rosen D. Between two worlds: bridging the cultures
of child health and adult medicine. J Adolesc Health
1995;17(1):10–6.
 Cardiol Clin 24 (2006) 631–639
Determinants and Assessment of Pulmonary
Regurgitation in Tetralogy of Fallot: Practice
and Pitfalls
Andrew N. Redington, MD
Division of Cardiology, Hospital for Sick Children, 555 University Avenue,
Toronto, Ontario M5G 1X8, Canada
During the last two decades, a remarkable
turnaround has occurred in understanding the
effects of pulmonary regurgitation after the repair
of tetralogy of Fallot. Previously thought to be
a benign consequence of outflow tract surgery, it
is now recognized as the single most important
determinant of late outcome. Indeed, deteriora-
tion in right ventricular function, exercise perfor-
mance, and the propensity to ventricular arrhythmia
can all be traced back to the secondary effects of
pulmonary regurgitation.
The goal of this article is not to describe the
secondary effects but rather the primary determi-
nants of pulmonary regurgitation. Furthermore,
the methods by which pulmonary regurgitation is
measured and some of the pitfalls therein are
highlighted.
Determinants of pulmonary regurgitation
There are four determinants of semilunar valve
regurgitation (Fig. 1): (1) the size of the regurgi-
tant orifice, (2) the hydraulic impedance/afterload,
(3) the ventricular diastolic compliance, and (4)
the diastolic filling time. In aortic regurgitation,
the difference between hydraulic impedance/after-
load and ventricular compliance is an overwhelm-
ing and relatively fixed determinant of regurgitant
flow; however, because of the low pressure, low im-
pedance nature of the pulmonary vascular bed,
a ‘‘freely’’ regurgitant outflow tract can be tolerated
for decades. Nonetheless, relatively subtle changes
E-mail address: andrew.redington@sickkids.ca
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.007
in these variables can lead to a major change in
the absolute volume of pulmonary incompetence.
For a full understanding of the amount of pulmo-
nary incompetence in a given patient, each of these
variables must be taken into account.
Regurgitant valve orifice
The normal right ventricular outflow tract is
a complex structure [1]. The pulmonary valve is
a free-standing structure separate from the fibrous
skeleton of the heart that provides a supporting
superstructure to the tricuspid, mitral, and aortic
valves. Sitting on a complete muscular infundibu-
lum, the integrity of the pulmonary valve leaflets
themselves will vary not only with the integrity
of the valve but also with the dynamic nature of
the muscular infundibulum and the size of the im-
mediate supravalve area. Gross right heart dilata-
tion, regardless of the cause, can lead to important
pulmonary regurgitation as the muscular outflow
tract dilates. There is no anatomic fibrous ring
supporting the pulmonary valve leaflets; rather,
there is a muscular to fibrous connection at the
point of the ventriculoarterial junction. The pul-
monary valve leaflets are arranged in semilunar
fashion, traversing the muscular and fibrous
elements, as in the aortic valve [1]. Thus, it is easy
to imagine that supravalve or subvalve dilatation
(iatrogenic or secondary) could undermine the
function of the semilunar valve leaflets, even if their
integrity was preserved at the time of repair.
Rarely does a patient with tetralogy of Fallot
not require enlargement of the outflow tract in
some way (beyond resection of the outlet septum);
therefore, surgery to the sinuses and supravalve
cardiology.theclinics.com
632
Pulmonary
afterload
Regurgitant
orifice
size
Diastolic
compliance
Fig. 1. Schematic demonstrating the determinants of
pulmonary regurgitation. The pulmonary artery after-
load, regurgitant orifice, and diastolic compliance of
the right ventricle are all independent variables in indi-
vidual patients.
area, enlargement of the ventriculoarterial junc-
tion, enlargement of the subpulmonary infund-
ibulum, or a combination of one or more
procedures in the form of a transannular patch
are variable elements of the repair. Perhaps
because of a lack of understanding of the late
sequelae of pulmonary regurgitation and the early
reports of the adverse effects of residual stenosis at
the time of primary repair [2], surgeons have erred
on the side of ‘‘complete’’ relief of stenosis at the
expense of regurgitation. Indeed, resection of pul-
monary valve leaflets in association with the trans-
annular patch was de rigueur in the 1960s and
1970s. Subsequently, there has been an increasing
trend toward preservation of the pulmonary out-
flow tract at the time of primary repair.
The paranoia regarding residual outflow tract
obstruction seems to be less relevant in the con-
temporary era. Surgical mortality is extremely low
as is interval mortality in the first years after sur-
gery, despite degrees of residual stenosis considered
intolerable in previous eras. It also recognized that
the diminutive outflow tract may grow, even if
residual stenosis remains at the end of the primary
procedure. Antunes and coworkers [3] showed
clearly that residual pressure gradients generally
REDINGTON
fell over the first 12 months after surgery in a series
in which transannular patches were avoided in the
vast majority of patients. More recently, the
Toronto group has demonstrated the early benefits
and the high rate of success with valve-sparing pro-
cedures [4]. Over 70% of unselected patients were
able to avoid transannular patches when a protocol
including deferred repair to the age of 3 to 6 months
was instituted. Modest residual gradients tended
not to progress, and many regressed during the first
years after surgery. There was a small risk for the
need of reoperation. Nevertheless, there is an in-
creasing trend toward early primary repair at the
time of diagnosis in the neonatal period. Although
this procedure can now be performed with an
extremely low mortality, the consequence of this
approach is a much higher transannular patch
rate, often greater than 70% [5]. It remains to be
seen whether the combination of early surgical
repair, with its stated aim of preserving right ven-
tricular diastolic function, in combination with
a transannular patch might lead to more long-
term problems than more conservative approaches.
This is a complex question because, for the reasons
outlined previously, even the more conservative
approaches, such as isolated outflow tract and
supra-annular patching, may ultimately provide
the substrate for free pulmonary regurgitation.
One should remember that even the circum-
stances of free pulmonary incompetence do not
equate directly to ‘‘severe’’ pulmonary incompe-
tence. This extremely important concept ties into
the nature of semilunar valve regurgitation. The
volume of regurgitation will vary markedly, even
across a potentially freely regurgitant valve orifice,
depending on the balance of afterload in the
pulmonary vascular bed and right ventricular di-
astolic function.
Pulmonary vascular bed
It has long been known anecdotally that the
combination of a transannular patch and distal
pulmonary artery stenosis is a particularly malig-
nant combination in terms of the evolution of
right ventricular dysfunction [6]. Intuitively, this
observation is unsurprising, but it is poorly docu-
mented in terms of the influence of pulmonary re-
gurgitation. To test this premise more directly, my
colleagues and I performed a series of studies in
children and young adults at the time of cardiac
catheterization [7]. Using a conductance catheter,
pulmonary regurgitation was measured beat-
by-beat during balloon dilation and stenting of
 PULMONARY REGURGITATION IN TETRALOGY OF FALLOT
peripheral pulmonary artery stenosis. Following
successful relief of stenosis, subsequent balloon in-
flation in the branch pulmonary artery was associ-
ated with a linear increase in the amount of
pulmonary regurgitation. Although this simulated
branch pulmonary artery stenosis led to the most
profound changes in pulmonary regurgitation, the
study also showed that the subtle waxing and
waning in mean airway pressure associated with
positive pressure ventilation also led to significant
changes in pulmonary incompetence, independent
of any pulmonary artery stenosis. Subtle pressure
transients in the pulmonary vascular bed may lead
to significant changes in the volume of pulmonary
incompetence.
This phenomenon is further illustrated by
a recent study from the Toronto group [8]. The to-
tal volume of regurgitation across the outflow
tract was measured using cardiac MR in a series
of patients late after repair. The individual branch
pulmonary artery contribution to this total regur-
gitant flow was also assessed (Fig. 2). Interest-
ingly, regurgitation from the left pulmonary
artery predominated. Indeed, in some individuals,
over 90% of the total regurgitant volume at the
outflow tract was contributed by regurgitation
from the left pulmonary artery. There are many
potential explanations for this phenomenon.
Purely anatomic reasons could not be invoked,
because there was no relationship between size
of the pulmonary artery and its regurgitant
100
P<0.0001
RPA
LPA
PR(%)
50
0 branches (Fig. 3), throughout respiration pre-
RPA LPA
Fig. 2. Graph demonstrating differential right (RPA)
and left pulmonary arterial (LPA) regurgitation as a per-
centage of the total regurgitant volume at the outflow
tract. In some patients, over 80% of the total regurgitant
volume is derived from the left pulmonary artery.
(Adapted from Kang IS, Redington AN, Benson LN,
et al. Differential regurgitation in branch pulmonary
arteries after repair of tetralogy of Fallot: a phase-
contrast cine magnetic resonance study. Circulation
2003;107(23):2940; with permission.)
633
volume. Although speculative, more subtle
changes in the pulmonary vascular bed of the
left lung may be responsible for the greater vol-
ume of regurgitation. It is easy to imagine a situa-
tion whereby some pulmonary incompetence leads
to dilatation of the right ventricle, which increases
the heart size, compressing the left lung. This in-
creases its vascular impedance, generating more
pulmonary incompetence and setting up a vicious
spiral. The heart can be considered to be a ‘‘space-
occupying lesion’’ under these circumstances.
Ventricular diastolic performance
Regurgitation into the ventricle will also be
affected by its relaxation properties in early
diastole, its compliance in late diastole, and,
ultimately, its absolute capacitance. All of these
properties may be abnormal in patients after
repair of tetralogy of Fallot, and, clearly, all are
subject to change with varying hemodynamic
loads. There is a wide variability in right ventric-
ular size (a reasonable surrogate of the total right
ventricular volume load), even in the presence of
a wide open outflow tract. Perhaps surprisingly,
some patients, despite a transannular patch, fail to
demonstrate significant right ventricular dilata-
tion, suggesting the total hemodynamic burden as
a result of free pulmonary incompetence is
limited. Conversely, the majority of patients with
identical outflow tract and pulmonary artery
morphology will experience progressive right
heart dilation under these circumstances.
The reason for these widely disparate re-
sponses must lie in the right ventricular diastolic
performance. Indeed, my colleagues and I showed
some years ago that the presence of restrictive
right ventricular physiology limited the adverse
effects of pulmonary incompetence [9]. The pres-
ence of antegrade diastolic flow, demonstrable
by pulsed wave Doppler echocardiography in
the distal pulmonary artery and pulmonary artery
dicted a smaller right ventricular size, improved
exercise performance, and lowered the propensity
to arrhythmia [10]. Interestingly, this physiology
when present in the early postoperative period
was associated with worse outcome [11]. In the
latter group, restrictive physiology was character-
ized by a lower cardiac output, a greater propen-
sity to fluid retention, and a slower postoperative
recovery. It remains unclear whether early postop-
erative restrictive physiology is the cause of late
postoperative restrictive physiology, but an
634
Fig. 3. Pulsed wave Doppler recording in the distal main pulmonary artery in a patient with restrictive right ventricular
physiology. Note the antegrade late diastolic flow coincident with the p wave on the electrocardiogram (ECG). A micro-
manometer-tipped right ventricular pressure (RVp) recording is also displayed. Note the raised right ventricular end-
diastolic pressure.
interesting study by Norgard and coworkers [12]
showed that restrictive physiology in the early
postoperative period was the only independent
predictor of restriction at midterm follow-up.
The Doppler echocardiographic manifestations
of restrictive right ventricular physiology are
entirely qualitative. In essence, the presence of
antegrade diastolic flow in the pulmonary artery
reflects any situation in which the hydraulic
impedance to blood flow into the pulmonary artery
is lower than the resistance to filling of the right
ventricle at end diastole. It is exquisitely variable
with small changes in hemodynamics, not least the
changes throughout the respiratory cycle. It may
also be true that antegrade diastolic flow in the
pulmonary artery appears when the right ventricle
has dilated so much that its ultimate capacitance
has been reached. Although highly significant as
a group, not all patients with restrictive physiology
have a small right ventricle. Indeed, there was
considerable overlap in terms of right ventricular
size and cardiothoracic ratio with the nonrestric-
tive group in our original series [10]. This finding
might also explain some of the differences observed
when ‘‘restrictive physiology’’ is diagnosed by
Doppler echocardiography compared with cardiac
MR imaging. Although several Doppler studies
have confirmed our findings [13,14], some MR
studies have suggested that the presence of restric-
tive physiology does not predict right ventricular
size, function, or outcome. Furthermore, it has
REDINGTON
recently been suggested that antegrade diastolic
flow measured by magnetic resonance is associated
with a worse outcome [15]. Although the reason for
this disparity is not entirely clear, it should be
remembered that MR measures antegrade diastolic
flow as it is averaged throughout many respiratory
and cardiac cycles. Furthermore, most measure-
ments are made at the putative ventriculoarterial
junction rather than the distal pulmonary artery.
The orientation of the patient (lying flat on his or
her back) may influence the pattern of flow in the
proximal right ventricular outflow tract. All of
these factors may conspire to make the interpreta-
tion of antegrade diastolic flow different when
measured under these circumstances.
Regardless of the implications of antegrade
diastolic flow, whether measured by Doppler
echocardiography or MR imaging, the ultimate
arbiter of outcome seems to be the amount of
pulmonary incompetence and its effects on right
ventricular size and performance. The measure-
ment of pulmonary incompetence is crucial to the
understanding of outcomes in these patients.
Techniques for the measurement of pulmonary
incompetence
Clinical surrogates
Although the presence of overt right heart
dilatation clinically, a loud diastolic murmur,
 PULMONARY REGURGITATION IN TETRALOGY OF FALLOT
and signs of congestive cardiac failure represent
one end of the hemodynamic spectrum, the assess-
ment of pulmonary incompetence by clinical evalu-
ation and auscultation cannot provide the detailed
evaluation required for the longitudinal evaluation
of these patients. Some useful simple outpatient
investigations may provide a guide to severity. The
utility of a chest radiograph should not be under-
estimated. In the absence of other hemodynamic
contributors, a change of heart size on the chest
radiograph may be a useful indicator of the presence
of continued ventricular and atrial dilatation. In our
original description of restrictive right ventricular
physiology, patients with the smallest right ventricles
by echocardiography had the smallest cardiotho-
racic ratios by chest radiography [9]. The cardiotho-
racic ratio is also a good generic predictor of overall
functional performance. In an outstanding study of
exercise performance in tetralogy of Fallot per-
formed by the Chicago group [16], the cardiotho-
racic ratio was an important independent predictor
of exercise performance. The smaller the heart size
on the radiograph, the better the exercise perfor-
mance in children and adolescents after surgery.
The electrocardiogram is also a reasonable sur-
rogate measure of changes in right ventricular size.
Reflecting the ‘‘mechano-electric’’ relationship, my
colleagues and I showed some years ago that the
QRS duration on the electrocardiogram is loosely
related to right ventricular size measured by echo-
cardiography [10]. Subsequently, this relationship
has been confirmed in more elegant studies using
MR assessment of right ventricular volume [17].
There can be no absolute guidelines regarding the
relationship between QRS duration and right
ventricular size because each patient has his or her
own baseline QRS morphology and trajectories of
change. Nonetheless, a significant change in the
QRS duration is likely to reflect an important
change in right ventricular size. Furthermore, a rap-
idly progressive prolongation of QRS duration is
a harbinger of adverse clinical outcome in the form
of ventricular arrhythmia and sudden death [18].
Although these surrogate measures may have
some use in the outpatient assessment of patients,
they clearly cannot be used for more sophisticated
analysis of outcomes. For this alternative, more
sensitive methods are required.
Cardiac catheterization: conductance assessment
of right ventricular volumes
The subjective assessment of pulmonary in-
competence from pulmonary arteriography and
635
right ventriculography is clearly a thing of the
past. Nonetheless, initial measurements of right
ventricular pressure-volume relationships using
digitized ventriculograms provide an improved
understanding of the relationship between pulmo-
nary incompetence and right ventricular hemody-
namics [19]. Subsequently, the author and others
have used conductance catheterization to assess
right ventricular hemodynamics on a beat-by-
beat basis [20,21]. This technique relies on the
time varying change in blood pool resistance
that is proportional to right ventricular volume
as it changes throughout the cardiac cycle. A spe-
cialized catheter incorporating multiple electrodes
is used to set up a series of electric fields within the
ventricle. If the resistivity of blood is known, right
ventricular size can be calculated from the com-
posite change in blood conductance. This tech-
nique has a remarkably high temporal resolution
(200 Hz) and can be combined with micromano-
metric pressure recordings to construct ventricular
pressure-volume relationships from which the vol-
ume of pulmonary incompetence occurring during
isovolumic relaxation can be measured. In pa-
tients with pulmonary regurgitation, right ventric-
ular volume begins to increase well before the
point of tricuspid valve opening. Assuming there
is no residual ventricular septal defect or other
source of volume increase in the ventricle during
that period, the source must be via the pulmonary
valve. The increase in volume between minimum
(end systolic) volume and that at the time of tri-
cuspid valve opening can be expressed as an index
of pulmonary regurgitant volume [7,18]. The re-
gurgitant volume measured in this way has been
shown to relate to right ventricular end diastolic
and end systolic size [7], to reflect differences in ex-
ercise performance [19], and, as mentioned previ-
ously, has been used to show the instantaneous
change in pulmonary incompetence as right ven-
tricular afterload increases [18]. It remains a useful
experimental technique for the analysis of pulmo-
nary incompetence in patients and in experimental
models if instantaneous changes need to be mea-
sured. It would be foolish to suggest that this tech-
nique could be used clinically, because there are
clearly far more satisfactory methods for general
screening and longitudinal assessment.
Cross-sectional echocardiography: Doppler
assessment
In general, the assessment of pulmonary in-
competence by ultrasound techniques has been
636 REDINGTON

qualitative rather than quantitative. A variety of Recently, an attempt at quantitation with

methods have been proposed [22], including the
measurement of jet widths, the propagation of re-
gurgitant color flow maps, and determination of
the extent of regurgitation into the distal pulmo-
nary arteries. Although all of these methods
show some relationship to the true regurgitant
volume, none can be used to assess small changes
longitudinally. Indeed, it is likely that the experi-
enced echocardiographer incorporates all of these
patterns into a qualitative assessment of the sever-
ity of pulmonary incompetence. The size of the
right ventricle, the appearance of the outflow tract
and pulmonary arteries, and the Doppler patterns
all are incorporated into the grading of severity.
comparison using MR imaging as a gold standard
has been proposed. The pulmonary regurgitation
index is a measure of pulmonary regurgitation
severity based on the duration of the regurgitant
signal by spectral Doppler study compared with
the diastolic filling time. When this ratio is less
than 0.77, Li and coworkers [23] suggest 100%
sensitivity and 84% specificity for identifying pa-
tients with a pulmonary regurgitant fraction of
greater than approximately 25% by MR imaging.
Fig. 4 shows two patients with a similar pulmo-
nary regurgitation index; both indices are signifi-
cantly lower than the 0.77 cut-off suggested by
Li and coworkers. Patient 1 had a normal right

Fig. 4. Composite illustrations taken from two patients with a similar pulmonary regurgitation (PR) index (see text for
details). In both cases, the index is predictive of severe PR; however, patient 1 (A) has a normal right ventricular size and
little PR by MR imaging despite ‘‘free’’ pulmonary incompetence as a result of a transannular patch placed 20 years
previously. This patient has restrictive right ventricular physiology that limits PR and prevents secondary dilation.
Patient 2 (B) does have severe right ventricular volume overload and severe PR (and, paradoxically, a less severe PR
index). These cases emphasize the need to be cautious with Doppler-derived surrogates of pulmonary incompetence
in the individual patient.
 PULMONARY REGURGITATION IN TETRALOGY OF FALLOT 637

ventricular size and only a small pulmonary regur-
gitant volume. This patient had a severely restric-
tive right ventricle. Patient 2 complied with the
hypothesis that a low pulmonary regurgitation in-
dex is associated with severe pulmonary incompe-
tence. MR imaging confirmed a huge right
ventricle with a large pulmonary regurgitant vol-
ume. These findings suggest that the pulmonary
regurgitant index should be used with caution.
There are two important pitfalls regarding use
of the index. First, the presence of restrictive
physiology, particularly if there is a variable or
prolonged pulmonary regurgitant interval, will
clearly influence this measurement in a counterin-
tuitive way. Second, there is an increasing
tendency to express pulmonary incompetence as
a fraction or ratio of the total stroke volume.
Indeed, the pulmonary regurgitant fraction is used
in most of the MR-based studies of pulmonary
incompetence nowadays. The use of the pulmo-
nary regurgitant fraction is fundamentally flawed.
Despite the relationship between the pulmonary
regurgitant fraction and the absolute pulmonary
regurgitant volume (which, in turn, will influence
right ventricular size, hemodynamics, and so on),
this relationship is variable. Patients who have
relatively normal right ventricles and a relatively
modest absolute volume of pulmonary regurgitant
may have an identical regurgitant fraction to
patients with a huge regurgitant volume and
a proportionately larger right ventricular stroke
volume.
My colleagues and I have recently looked at
the relationship between the pulmonary regurgi-
tant fraction and absolute pulmonary regurgitant
volumes. In the graph shown in Fig. 5, it can be
seen that, for the same pulmonary regurgitant
fraction, the absolute volume of regurgitation
may vary two or three fold. Conversely, for the
same regurgitant volume, the pulmonary regurgi-
tant fraction may vary markedly. Given that the
measurement of absolute volume is one of the
strengths of cardiac MR, it is disappointing that
its application to the measurement of pulmonary
incompetence has been simplified in this way. It
is likely that absolute volume, when corrected
for body surface area, for example, will prove to
be a more robust parameter than the regurgitant
fraction in the future.
Cardiac magnetic resonance
With the caveats expressed previously, cardiac
MR is clearly the technique of choice when
assessing pulmonary incompetence and right ven-
tricular hemodynamics. Nonetheless, there are
significant limitations to its use, and the lack of
consistent protocols makes assessment of data
from different groups difficult. Differences in
sequencing and the types of signal generation
affect the measurement of right ventricular vol-
ume and must be taken into account, even when
following an individual patient longitudinally, as
local equipment and protocols change.

Fig. 5. Plot showing MR imaging–derived pulmonary regurgitant volume (corrected for body surface area) versus the
pulmonary regurgitant fraction in individual patients. The patients highlighted in boxes have a similar regurgitant vol-
ume (2.5 L/min/m2) but a regurgitant fraction of 28% and 47%, respectively. The circled patients have a similar regur-
gitant fraction of 42%, but one has double the pulmonary incompetence volume than the other (4 versus 2 L/min/m2).
The pulmonary regurgitant fraction does not adequately describe the preload on the right ventricle. (Unpublished data
obtained in collaboration with Drs. Shi Joon Yoo and Lars Grosse-Wortmann, 2006.)
638
Furthermore, there remains a subjective element
to right ventricular boundary recognition in terms
of assessing the blood pool to myocardial in-
terface and the right ventricular outflow tract. The
latter is of particular significance. Frequently,
these patients have grossly dilated, aneurysmal,
and even dyskinetic segments of the outflow tract
[24]. The junction between the ventricle and pul-
monary artery is poorly defined and may vary
from study to study and observer to observer. Fre-
quently, the myocardial activity in the inlet and
apical portions of the ventricle is far more vigorous
than in the outflow tract. Whether independent
measurements of function in the different areas
may be of more utility, for example, in predicting
reverse remodeling, remains to be seen. Despite
the ability to measure directly pulmonary regurgi-
tant volume for over a decade, we remain uncer-
tain as to the clinical implications and utility of
such measurements in terms of clinical decision
making.
Summary
Pulmonary incompetence is the single most
important factor in the midterm outcome of
patients after repair of tetralogy of Fallot. The
pulmonary regurgitant volume and, hence, the
burden on the right ventricle, is dependent on
a subtle balance of orifice size, pulmonary after-
load, and right ventricular diastolic physiology.
Although the ability to measure pulmonary
incompetence has improved significantly, there
remain important pitfalls to its assessment in the
individual patient, and our interpretation of its
implications, regardless of which method is used,
remains unsophisticated.
References
[1] Anderson RH, Freedom RM. Normal and abnor-
mal structure of the ventriculo-arterial junctions.
Cardiol Young 2005;15(Suppl 1):3–16.
[2] Kirklin JW, Blackstone EH, Pacifico AD, et al. Risk
factors for early and late failure after repair of tetral-
ogy of Fallot, and their neutralization. Thorac Car-
diovasc Surg 1984;32(4):208–14.
[3] Antunes MJ, Castela E, Sanches MF, et al. Pre-
servation of the pulmonary annulus in total
correction of tetralogy of Fallot: decreasing transan-
nular gradients in the early follow-up period. Eur J
Cardiothorac Surg 1991;5(10):528–32.
[4] Van Arsdell G, Yun TJ. An apology for primary re-
pair of tetralogy of Fallot. Semin Thorac Cardiovasc
Surg Pediatr Card Surg Annu 2005;128–31.
REDINGTON
[5] Pigula FA, Khalil PN, Mayer JE, et al. Repair of
tetralogy of Fallot in neonates and young infants.
Circulation 1999;100(19 Suppl):II157–61.
[6] Sunakawa A, Shirotani H, Yokoyama T, et al.
Factors affecting biventricular function following
surgical repair of tetralogy of Fallot. Jpn Circ J
1988;52(5):401–10.
[7] Chaturvedi RR, Kilner PJ, White PA, et al.
Increased airway pressure and simulated branch
pulmonary artery stenosis increase pulmonary re-
gurgitation after repair of tetralogy of Fallot: real-
time analysis with a conductance catheter technique.
Circulation 1997;95(3):643–9.
[8] Kang IS, Redington AN, Benson LN, et al. Differential
regurgitation in branch pulmonary arteries after repair
of tetralogy of Fallot: a phase-contrast cine magnetic
resonance study. Circulation 2003;107(23):2938–43.
[9] Gatzoulis MA, Clark AL, Cullen S, et al. Right ven-
tricular diastolic function 15 to 35 years after repair
of tetralogy of Fallot: restrictive physiology predicts
superior exercise performance. Circulation 1995;
91(6):1775–81.
[10] Gatzoulis MA, Till JA, Somerville J, et al. Mecha-
noelectrical interaction in tetralogy of Fallot: QRS
prolongation relates to right ventricular size and pre-
dicts malignant ventricular arrhythmias and sudden
death. Circulation 1995;92(2):231–7.
[11] Cullen S, Shore D, Redington A. Characterization
of right ventricular diastolic performance after com-
plete repair of tetralogy of Fallot: restrictive physiol-
ogy predicts slow postoperative recovery.
Circulation 1995;91(6):1782–9.
[12] Norgard G, Gatzoulis MA, Josen M, et al. Does re-
strictive right ventricular physiology in the early
postoperative period predict subsequent right ven-
tricular restriction after repair of tetralogy of Fallot?
Heart 1998;79(5):481–4.
[13] Sachdev MS, Bhagyavathy A, Varghese R, et al.
Right ventricular diastolic function after repair of
tetralogy of Fallot. Pediatr Cardiol 2006;27(2):
250–5.
[14] Eroglu AG, Sarioglu A, Sarioglu T. Right ventricular
diastolic function after repair of tetralogy of Fallot: its
relationship to the insertion of a ‘‘transannular’’
patch. Cardiol Young 1999;9(4):384–91.
[15] Helbing WA, Niezen RA, Le Cessie S, et al. Right
ventricular diastolic function in children with pul-
monary regurgitation after repair of tetralogy of Fal-
lot: volumetric evaluation by magnetic resonance
velocity mapping. J Am Coll Cardiol 1996;28(7):
1827–35.
[16] Wessel HU, Cunningham WJ, Paul MH, et al. Exer-
cise performance in tetralogy of Fallot after intracar-
diac repair. J Thorac Cardiovasc Surg 1980;80(4):
582–93.
[17] Abd El Rahman MY, Abdul-Khaliq H, Vogel M,
et al. Relation between right ventricular enlarge-
ment, QRS duration, and right ventricular function
in patients with tetralogy of Fallot and pulmonary
 PULMONARY REGURGITATION IN TETRALOGY OF FALLOT
regurgitation after surgical repair. Heart 2000;84(4):
416–20.
[18] Gatzoulis MA, Balaji S, Webber SA, et al. Risk fac-
tors for arrhythmia and sudden cardiac death late af-
ter repair of tetralogy of Fallot: a multicentre study.
Lancet 2000;356(9234):975–81.
[19] Redington AN, Oldershaw PJ, Shinebourne EA,
et al. A new technique for the assessment of pulmo-
nary regurgitation and its application to the assess-
ment of right ventricular function before and after
repair of tetralogy of Fallot. Br Heart J 1988;60(1):
57–65.
[20] Dickstein ML, Yano O, Spotnitz HM, et al. Assess-
ment of right ventricular contractile state with the
conductance catheter technique in the pig. Cardio-
vasc Res 1995;29(6):820–6.
[21] Lambermont B, Ghuysen A, Kolh P, et al. Effects of
endotoxic shock on right ventricular systolic
639
function and mechanical efficiency. Cardiovasc Res
2003;59(2):412–8.
[22] Mulhern KM, Skorton DJ. Echocardiographic eval-
uation of isolated pulmonary valve disease in adoles-
cents and adults. Echocardiography 1993;10(5):
533–43.
[23] Li W, Davlouros PA, Kilner PJ, et al. Doppler-echo-
cardiographic assessment of pulmonary regurgita-
tion in adults with repaired tetralogy of Fallot:
comparison with cardiovascular magnetic resonance
imaging. Am Heart J 2004;147(1):165–72.
[24] Davlouros PA, Kilner PJ, Hornung TS, et al. Right
ventricular function in adults with repaired tetralogy
of Fallot assessed with cardiovascular magnetic res-
onance imaging: detrimental role of right ventricular
outflow aneurysms or akinesia and adverse right-
to-left ventricular interaction. J Am Coll Cardiol
2002;40(11):2044–52.
 Cardiol Clin 24 (2006) 641–660

Exercise Intolerance in Adults with Congenital

Heart Disease
Konstantinos Dimopoulos, MDa,b,*, Gerhard-Paul Diller, MDa,b,
Massimo F. Piepoli, MD, PhDb, Michael A. Gatzoulis, MD, PhD, FACCa
a
Adult Congenital Heart Programme, Department of Cardiology, Royal Brompton Hospital,
Sydney Street, London, UK
b
Department of Clinical Cardiology, National Heart and Lung Institute, Imperial College School of Medicine,
London, SW3 6NP, UK

Exercise intolerance is a common feature of
congenital and acquired heart disease. It affects
approximately one third of adults with congenital
heart disease (CHD) and is present in all congen-
ital groups, including those with simple lesions [1].
Its clinical manifestations are exertional dyspnea
or exertional fatigue, and it is one of the most fre-
quent causes of reduced quality of life in adult
CHD (ACHD) [2–4].
CHD shares important features with heart
failure in patients with acquired heart disease,
both in the clinical presentation and systemic
manifestations. CHD is, in fact, a cause of the so-
called heart failure syndrome. Traditionally, heart
failure was identified with left ventricular systolic
dysfunction of idiopathic or ischemic origin. Re-
cent American College of Cardiology/American
Heart Association (ACC/AHA) and European
Society of Cardiology guidelines define heart fail-
ure mainly on the basis of symptoms of exercise
intolerance and recognize it can be caused by
various cardiac disorders beyond left ventricular
systolic dysfunction [5,6]. It has become apparent
K.D. has been supported by the European Society
of Cardiology. G-P.D. is supported by Actelion, UK.
The Royal Brompton Adult Congenital Heart Program
and the Department of Clinical Cardiology have re-
ceived support from the British Heart Foundation and
the Clinical Research Committee, Royal Brompton
Hospital, London.
* Corresponding author.
E-mail address: k.dimopoulos@ic.ac.uk
(K. Dimopoulos).
0733-8651/06/$ - see front matter Ó 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.ccl.2006.08.002
that heart failure is a clinical syndrome with numer-
ous systemic manifestations and multiple potential
causes, including CHD.
This article describes the ways to assess exer-
cise capacity in CHD and the impact of exercise
intolerance in the ACHD population. It also
discusses the likely pathogenesis of exercise in-
tolerance in CHD, the similarities between ACHD
and acquired heart failure, and potential thera-
peutic options.
Assessment of exercise intolerance in the adult
congenital heart disease population
Subjective assessment
The most popular means of quantification of
physical impairment in ACHD is the New York
Heart Association (NYHA) classification [3,4]. It
is an established semiquantitative means of sub-
jective assessment of the degree of exercise intoler-
ance in chronic heart failure. The NYHA
classification has been adopted by ACHD health
providers, because it is easy to use and is familiar
to health workers outside the area of ACHD [7]. It
is also often reported (inappropriately) as a mea-
sure of health-related quality of life in ACHD de-
spite the existence of other specific scoring systems
of ACHD quality of life [3,4,8].
Objective assessment
Cardiopulmonary exercise testing
Cardiopulmonary exercise testing is a powerful
tool for the simultaneous objective evaluation of
cardiology.theclinics.com
642 DIMOPOULOS
the cardiovascular, respiratory, and muscular
systems under conditions of controlled metabolic
stress [9]. It is used widely for assessing patients
with acquired heart failure, and, like the NYHA
classification, it has been adopted by ACHD phy-
sicians, gradually becoming part of the routine
clinical assessment of ACHD patients. Incremen-
tal protocols are used to assess important prog-
nostic indices such as peak oxygen consumption
(peak VO2), ventilatory response to exercise, an-
aerobic threshold, and heart rate response [10].
Peak VO2 is the highest measured value of
oxygen uptake (usually expressed in mL/kg/min)
and approximates the maximal aerobic power of
each individual (Fig. 1) [11,12]. It is a surrogate
of the functional status of the pulmonary,
cardio–circulatory, and muscular systems and
represents the upper limit of oxygen use. It is the
most frequently reported exercise parameter
because of its simplicity and prognostic power
both in acquired heart disease and in ACHD.
Peak VO2 approximates maximal exercise capac-
ity to the extent that the patient is able and deter-
mined to exercise to exhaustion. When patients
fail to exercise to the maximum of their capacity,
measured peak VO2 will not reflect their true
exercise capacity. Moreover, as it is derived only
A 2.0
VO2
1.5 VCO2
1.0
Start
0.5
0 200 400
B 70
VE/VO2
60 VE/VCO2
Start
50
40
30
0 200 400
Fig. 1. Recordings from a maximal cardiopulmonary exercise test using a modified Bruce protocol in a patient after
Fontan-type operation. (A) Oxygen consumption (VO2) and carbon dioxide production (VCO2 in mL/min) as related
to time (in seconds). (B) The ratio VE/VO2 and VE/VCO2 against time. Abbreviations: AT, anaerobic threshold;
Peak, peak exercise; RC, respiratory compensation point; Start, start exercise; VE, ventilation.
et al
from few measurements at peak exercise, it can
be prone to technical error [13].
The anaerobic threshold is the level of oxygen
consumption above which aerobic metabolism is
supplemented by anaerobic processes [10,11,13].
Above the anaerobic threshold, lactate accumu-
lates and is buffered by plasma bicarbonate,
causing an increase in carbon dioxide (CO2) pro-
duction (VCO2). Recognition of the anaerobic
threshold is achieved by observing the VCO2 ver-
sus VO2 slope (the V-slope method) or the ventila-
tion versus VO2 and VCO2 slopes (the ventilatory
equivalent method) [11]. The anaerobic threshold
carries important pathophysiologic and prognos-
tic information in acquired heart failure [14]. Nev-
ertheless, in more symptomatic patients, its
recognition can be difficult, despite the develop-
ment of automated algorithms [13].
The VE/VCO2 slope is the slope of the regres-
sion line between minute ventilation (VE) and
VCO2 (Fig. 2) and has been proposed as an exer-
cise parameter that is independent of achieving
maximal exertion. It is a simplification of the com-
plex relation between ventilation and VCO2 and
appears to contain important physiological and
prognostic information [14,15]. It is easy to
calculate, is highly reproducible, and is a marker
Peak
RC
AT
600 800 1000
time
Peak
AT RC
600 800 1000
time
 EXERCISE INTOLERANCE IN ACHD
60
50
VE (l/min)
40
30
20
10
500 1000 1500 2000
VCO2 (ml/min)
Fig. 2. Plot of ventilation (VE) versus VCO2 from the
same patient as Fig. 1. The VE/VCO2 slope is the slope
of the regression line between ventilation (VE) and
VCO2 (continuous line). This can be calculated either
over the entire duration of exercise (as shown in this
figure), or after exclusion of the portion following the
respiratory compensation point [16]. In this case, using
data from the entire period of exercise, the VE/VCO2
slope is 28, suggesting that no ventilatory inefficiency is
present.
of exercise intolerance strongly related to peak
VO2 and to mortality in acquired heart failure
[14,15]. The mechanism behind the abnormal ven-
tilatory response to increments in CO2 production
is not yet understood fully. Several mechanisms
have been proposed, such as pulmonary hypoper-
fusion or vasoconstriction with increased physio-
logical dead space and ventilation/perfusion
mismatch and stimulation of the respiratory cen-
ters by triggers other than CO2 [16]. It is generally
recommended that the nonlinear portion of the
VE versus VCO2 relation at the start of exercise
and after the respiratory compensation point
(the point at which hyperventilation occurs in re-
sponse to increasing lactate production) should be
excluded when calculating the VE/VCO2 slope.
Nevertheless, numerous studies that have estab-
lished VE/VCO2 as an important exercise param-
eter have not used this theoretically appealing
convention [15,17,18]. One study on patients
with acquired heart failure has suggested that
the VE/VCO2 slope calculated using the entire ex-
ercise period is prognostically stronger [19]. In the
authors’ center, the VE/VCO2 slope is routinely
calculated as an adjunct to peak VO2 in assessing
exercise capacity and prognosis in the ACHD
population (Fig. 3). Values of VE/VCO2 slope
643
exceeding 33 are considered abnormal. In noncya-
notic patients who have ACHD, a VE/VCO2
slope of 38 or above identifies those at a 10-fold
increased risk of mortality at a mean follow-up
of 20 months (Fig. 4) [16].
6-minute-walk test
The 6-minute walk test is a submaximal timed
distance test often employed in patients who have
ACHD. The main response variable in this type of
exercise testing is the distance that individuals are
able to cover at their own pace in 6 minutes. A
normal 40-year-old subject covers approximately
600 m, decreasing by 50 m per decade [20]. Oxy-
gen saturations by portable pulse oximetry can
also be recorded.
The 6-min walk test reflects ordinary daily
activities and is easy to perform. It is a submaxi-
mal test in healthy individuals and mildly im-
paired patients but can be a maximal test in highly
symptomatic patients. Indeed, the 6-minute walk-
ing test distance correlated well to peak VO2 in
a study of highly symptomatic patients [21]. It is
commonly used to assess the response to
advanced therapies in patients with pulmonary
hypertension and is the only test approved by
the US Food and Drug Administration (FDA)
as an endpoint for prospective clinical trials in
this population [22,23]. Correlation to peak VO2
appears to be best when distance is adjusted for
patient weight (distance  weight) [24].
It may be inappropriate to use the 6-min walk
test in patients with mild impairment of exercise
capacity, because it could mask improvement
after an intervention [25]. Reproducibility and
comparability of repeated 6-min walking tests
also depend on adequate standardization of the
protocol used. An important learning effect has
also been described and should be considered
when comparing the first with subsequent tests
[26].
Exercise intolerance in adult patients
with congenital heart disease
The Euro Heart Survey, a large European
registry of ACHD patients reported that approx-
imately one third of ACHD patients complain of
symptoms of exercise intolerance (Fig. 5) [1].
Patients with cyanotic lesions and those with
univentricular hearts have the highest prevalence
of exercise intolerance, whereas patients with
aortic coarctation are usually asymptomatic
[1,27,28]. Eisenmenger patients have by far the
644 DIMOPOULOS et al

Fig. 3. Example of the form used at the authors’ center for reporting cardiopulmonary exercise tests. Peak VO2, the VE/
VCO2 slope, blood pressure and heart rate response, and oxygen saturation at rest and during exercise are reported
routinely.

highest degree of subjective impairment, suggesting
a synergistic detrimental effect of cyanosis and pul-
monary hypertension. In one cohort of 188 patients
with this condition, 68% were symptomatic by age
9, rising to 84% at age 28 [29]. Marked exercise in-
tolerance also has been reported in patients who
have congenitally corrected (l-) transposition of
the great arteries [30,31]. Similarly, patients after
atrial switch for transposition of the great arteries
are symptomatic with exertional dyspnea, whereas
those with repaired tetralogy of Fallot are generally
less limited in their exercise capacity [32–34].
Significant objective impairment in exercise
capacity has been described in various ACHD
groups using cardiopulmonary exercise testing
[18,35–43]. In the authors’ laboratory, routine car-
diopulmonary exercise testing applied to a large
cohort of ACHD patients provided estimates of ex-
ercise capacity across the spectrum of CHD (Fig. 6)
[44]. Peak VO2 was depressed in all ACHD groups
compared with healthy subjects of similar age, but
varied according to underlying anatomy. Peak
heart rate, forced expiratory volume (%FEV1),
pulmonary arterial hypertension, cyanosis, gender,
and body mass index were independent predictors
of peak VO2 in this population. Systemic and pul-
monary ventricular function were not predictive of
peak VO2 on multivariate analysis, confirming in
ACHD what is already known in acquired heart
failure, that exercise capacity is independent of
resting cardiac function [45,46].
The subjective assessment of exercise intoler-
ance using the NYHA classification appears to
underestimate the severity of functional
 EXERCISE INTOLERANCE IN ACHD 645

A 3.0 100

Ventilation(l/min)
Predicted Peak VO2
80
VO2 6
2.0 24.
l/min VCO2 pe=
60 slo
O2
98% of predicted /VC
peak VO2 40 VE
1.0

20
0.0
0 200 400 600 800 1000 1200 0.5 1.0 1.5 2.0 2.5 3.0
time VCO2 (l/min)

B 100
8

Ventilation(l/min)
Predicted Peak VO2
3.0 =3
80 pe
slo
38% of predicted peak VO2 2
VO2 CO
l/min

2.0 60 /V
VCO2 VE
40
1.0
20
0.0
0 200 400 600 800 1000 1200 0.5 1.0 1.5 2.0 2.5 3.0
time VCO2 (l/min)

C 1.5 100
Ventilation(l/min)

Predicted Peak VO2

0
=9
80

pe
1.0 43% of predicted peak VO2

slo
VO2
l/min

60
VCO2 O
2
/VC

0.5 40
VE

20
0.0
0 200 400 600 800 1000 1200 0.5 1.0 1.5 2.0 2.5 3.0
time VCO2 (l/min)

Fig. 4. Examples of cardiopulmonary exercise tests. (A) An asymptomatic (NYHA class I) 40-year-old patient with re-
paired tetralogy of Fallot underwent cardiopulmonary exercise testing as part of routine clinical assessment. Peak VO2
was 29 mL/kg/min. This was calculated as the mean of the two highest consecutive values of 15-second averages of VO2
divided by the patient’s weight (in this case 2450 mL/min divided by 84 kg of weight) and was 98% of predicted for age,
weight, height, and gender. The VE/VCO2 slope was 24.6, within normal limits. He had an adequate heart rate (90 to 171
bpm) and blood pressure response (118/78 to 158/92 mm Hg). No desaturation was observed. In this case, cardiopul-
monary exercise testing confirmed an optimal exercise tolerance. (B) Mildly symptomatic (NYHA class II) 20-year-
old man after Fontan-type operation. He exercised for 8.6 min and reached a peak VO2 of 16.8 mL/kg/min, which
was 38% of predicted. The VE/VCO2 slope was 38, which is abnormally raised. Blood pressure and heart rate response
were blunted (80 to 133 bpm and 100/64 to 132/72mm Hg). No desaturation was observed, making right-to-left shunting
(commonly observed in Fontan patients) unlikely. The results of the exercise test suggest that this patient had an exercise
capacity that was impaired significantly, more than what was suggested by his symptoms. This patient was found to have
moderate-to-severe systemic ventricular dysfunction. Moreover, a peak VO2 of less than 15.5 mL/kg/min and VE/VCO2
slope of 38 would identify this patient, in retrospect, at high risk of hospitalization and death, which was indeed the case
[16,44]. The patient died 6 months after the index exercise test. (C) A mildly symptomatic 52-year-old woman with Ei-
senmenger physiology. She exercised for 6.3 minutes and reached a peak VO2 of 11.2 mL/kg/min, which was 43% of
predicted. The VE/VCO2 slope was 90, which is severely raised. Her resting saturations were 90% and decreased to
45% at peak exercise, when she became light-headed and had to stop. Heart rate response was appropriate (89 to
151 bpm), but blood pressure response was suboptimal (128/82 to 142/80mm Hg). In this case, the cardiopulmonary
exercise test revealed significant exercise intolerance and severe ventilatory inefficiency. Despite near-normal resting ox-
ygen saturations, severe desaturation occurred and was likely the cause of termination of exercise. Moreover, the signif-
icant right-to-left shunt likely caused a significant ventilation-perfusion (V/Q) mismatch and stimulation of peripheral
and central chemoreceptors, leading to an increase in ventilation, disproportionate to CO2 production.
646 DIMOPOULOS et al

100%

90%

80%

70%

60%
Patients

Symptomatic (NYHA II-IV)

50% Asymptomatic (NYHA I)

40%

30%

20%

10%

0%
OVERALL

COA

MARFAN

VSD

TOF

TGA

ASD

FONTAN

CYANOTIC
Fig. 5. Symptomatic (NYHA class II or more) versus asymptomatic (NYHA class I) ACHD patients according to
underlying anatomy. (Data from Engelfiet P, Boersma E, Oechslin E, et al. The spectrum of adult congenital heart
disease in Europe: morbidity and mortality in a 5-year follow-up period. The Euro Heart Survey on adult congenital
heart disease. Eur Heart J 2005;26(21):2325–33.)

impairment in ACHD. A linear relationship exists
between NYHA class and peak VO2 across the
spectrum of ACHD, but many asymptomatic
ACHD patients show dramatically lower peak
VO2 values compared with normal controls. Use
of a subjective measure like NYHA class, which is
based on the perception of ordinary everyday
activity in a population of patients with inborn
heart defects, understandably leads to underesti-
mation of the real degree of impairment. This
appears to be an important point of disparity
between ACHD and acquired heart failure, in
which the difference in peak VO2 between normal
controls and asymptomatic heart failure patients
is much less. Peak VO2 values of ACHD patients
in the authors’ study approximated those of older
patients with acquired heart failure and similar
NYHA class. Subjective assessment using the
NYHA classification thus seems to adequately cat-
egorize ACHD patients into classes of decreasing
exercise capacity but cannot be used to assess the
magnitude of exercise intolerance in this popula-
tion. This underlines the importance of cardiopul-
monary exercise testing for the routine clinical
assessment of patients with ACHD [44,47].
The authors recently described abnormally
high mean values of the VE/VCO2 slope in all
major ACHD groups compared with normal
controls [16]. The Eisenmenger and complex
diagnoses groups were found to have the highest
VE/VCO2 slopes. Cyanosis was the strongest in-
dependent predictor of the VE/VCO2 slope in
this cohort and appeared to have a significant im-
pact on the ventilatory response to exercise. The
VE/VCO2 slope showed a linear relation to
NYHA class, suggesting a link between the venti-
latory response to exercise and the occurrence of
symptoms. The VE/VCO2 slope was raised signif-
icantly even among asymptomatic patients, fur-
ther underscoring the importance of objective
assessment of exercise capacity in ACHD.
Mechanisms of reduced exercise capacity
in adult congenital heart disease
The components of exercise: the lungs, the heart,
the blood, the vessels, and the skeletal muscles
During exercise, skeletal muscles must regen-
erate energy continuously in the form of ATP [48].
Aerobic metabolism is by far the most efficient
means of ATP production, but it requires ade-
quate coupling of external respiration (oxygen
delivery) to cellular respiration (muscle require-
ments). Anaerobic glycolysis becomes active
when oxygen delivery is inadequate. It requires
no oxygen but produces smaller amounts of
ATP and has lactic acid as a by-product, which
 EXERCISE INTOLERANCE IN ACHD 647

A Peak VO2 (ml/kg/min) Exercise capacity largely depends on adequate

60 coupling of muscle and external respiration (ie, on
an adequate delivery of oxygen to meet the needs
50 of the active muscles and sustain the aerobic
regeneration of ATP). Adequate oxygen supply
40
depends on a chain of events: ventilation, lung
Units

30 perfusion, and the cardiovascular system. Abnor-

malities in any of these elements (lungs, heart,
20 blood, blood vessels) can be detrimental and limit
10
exercise capacity. CHD can affect all components
of the cardiorespiratory chain, resulting in a sig-
0 nificant impairment of exercise capacity.
MVO2
Normal controls
Coarctation
Fallot
ASD
Mustard

cc-TGA
VSD
Ebstein
Fontan
AVSD
Complex
PPH
Valvular

Eisenmenger The heart

The heart plays a key role in metabolic–re-
spiratory coupling during exercise. An adequate
increase in systemic cardiac output is essential to
meet the metabolic demands of the working
B VE/VCO2
105 muscles. The heart increases its output by in-
95
crements in stroke volume and heart rate. The left
ventricle is directly responsible for the increase in
85
systemic cardiac output but can only achieve this
75 in the presence of an adequate increment in right
Units

65 ventricular output. At the start of exercise,

55 pulmonary vascular resistance drops, allowing
45 the right ventricle to increase pulmonary perfu-
35
sion and blood to return to the left ventricle.
Moreover, significant ventricular interaction oc-
25
curs in CHD, especially in cases of right ventric-
15
ular overload such as Ebstein’s anomaly [49].
VE/VCO2 slope
Thus, conditions that affect the left or the right
Normal controls
Coarctation
Fallot
ASD
Mustard

cc-TGA
VSD
Ebstein
Fontan
AVSD
Complex
PPH
Valvular

Eisenmenger

ventricle can have a great impact on exercise ca-

Fig. 6. Peak VO2 and VE/VCO2 slope in various ACHD
groups. Data on 933 ACHD patients and controls from
the Royal Brompton Hospital cardiopulmonary exercise
testing laboratory. ASD, atrial septal defect; AVSD,
atrioventricular septal defect; cc-TGA, congenitally cor-
rected (l-) transposition of the great arteries; Complex,
complex anatomy (double inlet or outlet right and left
ventricle or complex pulmonary atresia); PPH, idio-
pathic pulmonary arterial hypertension; VSD, ventricu-
lar septal defect.
is buffered by bicarbonate, generating CO2. The
hydrolysis of phosphocreatine is another anaero-
bic mechanism of ATP production, which be-
comes transiently active at the beginning of
exercise, when the circulation has not had time
to respond to the activity of the muscles [9,48].
pacity [50].
Most types of CHD impart a hemodynamic
load (pressure or volume overload) upon one or
both ventricles. This load is, by definition in
congenital heart disease, long-standing and can
lead to severe ventricular dysfunction. Significant
impairment of systemic ventricular function was
found 10 to 30 years after surgery in patients after
atrial repair of transposition of the great arteries,
congenitally corrected transposition of the great
arteries, and patients with Fontan-like circula-
tions [27,30,31,51–57].
Right ventricular systolic dysfunction is typi-
cally present in Ebstein’s anomaly and in patients
who have arrhythmogenic right ventricular dys-
plasia [58,59]. Moreover, significant pulmonary
regurgitation after repair of tetralogy of Fallot
or of isolated pulmonary stenosis can also lead
to right ventricular dysfunction, as can other
causes of right ventricular volume or pressure
overload (atrial septal defects, atrioventricular
648 DIMOPOULOS
septal defects with tricuspid regurgitation, pulmo-
nary stenosis, or Eisenmenger physiology) [50,58].
Finally, it has been observed that repeated opera-
tions involving protracted cardiopulmonary
bypass, especially those performed in previous
eras, and right ventricular outflow tract patch
augmentation, can affect right ventricular func-
tion adversely [50,58,60].
Arrhythmias
Patients who have ACHD have an increased
propensity to arrhythmias. This can be attributed
to several factors such as the congenital cardiac
lesion itself, long-standing hemodynamic factors,
and scarring from previous surgery. Arrhythmias
in ACHD will often cause cardiac dysfunction
and may become life-threatening, especially when
very fast, when of ventricular origin, or in the
presence of significant underlying disease. Never-
theless, even supraventricular tachycardias with
a mild increase in ventricular rate can cause
a reduction in cardiac output because of loss of
atrioventricular synchrony or when long-stand-
ing. Li and colleagues showed that patients who
have ACHD with atrial flutter (mean ventricular
rate 106 plus or minus 21bpm) developed a dra-
matically decreased exercise duration and lower
peak blood pressure and systemic ventricular
function [61]. The authors have reported that
patients in sinus rhythm during cardiopulmonary
exercise testing have a higher peak heart rate and
a higher peak VO2 compared with those not in
sinus rhythm [44]. Tachycardia management in
ACHD remains challenging because of complex
and variable hemodynamic profiles, extensive
scarring from prior surgery, and anatomical
variance posing ongoing difficulties to electrical
therapies.
Chronotropic incompetence
Cardiac output is the product of stroke volume
and heart rate. Stroke volume increases at the
beginning of exercise, whereas further increments
in cardiac output occur through increases in heart
rate. Chronotropic incompetence is the inability
of the heart rate to increase appropriate to the
metabolic demands during exercise and can be
a cause of exercise intolerance [62]. Even though
the concept of chronotropic incompetence ap-
pears straight-forward, a definition of normal or
expected heart rate response to exercise is elusive.
Numerous formulas have been proposed, the
most widely used being 220 minus age or the As-
trand formula [63].
et al
In acquired heart failure, chronotropic incom-
petence relates to NYHA class and peak VO2 [64].
Moreover, chronotropic incompetence is a predic-
tor of adverse outcome in patients with ischemic
heart disease and in the general population
[65–67]. In the ACHD population, chronotropic
incompetence is also prevalent and relates to exer-
cise capacity [37,68,69]. The authors recently re-
ported that heart rate reserve (HRR equals peak
heart rate minus resting heart rate) is a strong
predictor of mortality, especially in patients who
have complex lesions, Fontan-type surgery, and
repaired tetralogy of Fallot [70]. Whether the
prognostic power of HRR is the result of its
relationship to autonomic imbalance and exercise
capacity needs to be elucidated [71,72].
Pacing
Pacemaker therapy has an important role for
managing patients who have ACHD. Sinus node
dysfunction is present in 13% to 44% of patients
after Fontan operation and in 50% of patients
after atrial switch repair for transposition of the
great arteries [73]. Atrioventricular block used to
be common immediately after surgical repair of
perimembranous or infundibular ventricular sep-
tal defects or after right ventricular muscle bundle
resection. It can also develop later in life, irrespec-
tive of surgery, in patients who have atrioventric-
ular septal defects and congenitally corrected (l-)
transposition [73]. Right ventricular pacing can
cause ventricular asynchrony, and in the noncon-
genital population, it has been shown to cause left
ventricular dysfunction and reduced exercise ca-
pacity [74–76]. Dual-chamber pacemakers are
being implanted increasingly, maintaining atrio-
ventricular synchrony in an attempt to optimize
ventricular function and cardiac output [77].
Although rate-responsive pacemakers are used al-
most universally, rate responsiveness at higher
levels of exercise may be inadequate to maintain
an adequate cardiac output. In the authors’ expe-
rience, ACHD patients with permanent pace-
makers have significantly lower values of peak
heart rate and a trend toward lower peak VO2
compared with those without a pacemaker [44].
Myocardial perfusion abnormalities
Myocardial perfusion is a major determinant
of ventricular function. Abnormal myocardial
perfusion has been suggested as an underlying
cause of ventricular dysfunction in CHD.
Reversible and fixed perfusion defects with
 EXERCISE INTOLERANCE IN ACHD
concordant regional wall motion abnormalities
and impaired myocardial flow reserve often occur
in the right ventricle of patients with transposition
of the great arteries long after atrial repair [53,78].
Myocardial perfusion defects are also common in
patients after an arterial switch operation
[79–81]. These perfusion abnormalities correlate
well to impaired ventricular function and are
seen more frequently in older patients with longer
follow-up. Although the genesis of these abnor-
malities remains obscure, it is likely that myocar-
dial perfusion defects may be a sensitive predictor
of systemic ventricular impairment and a target
for future therapies [52,57,82–86].
Pericardial disease
Previous cardiac surgery for congenital or
acquired disease is a known risk factor for de-
veloping constrictive pericarditis. This is a rare
disease, occurring in approximately 0.2% of
operated patients [87]. Constrictive pericarditis
can have dramatic effects, causing significant exer-
cise intolerance and signs of congestive heart fail-
ure. Diagnosis requires a high degree of suspicion
and has to be part of the differential diagnosis of
unexplained exercise intolerance, especially when
dealing with the ACHD population in which
most patients have undergone cardiac surgery
[87,88].
Partial and total absence of the pericardium
can also cause significant changes in cardiac
anatomy and compromise cardiac function.
Loss of the pericardial constraint can result in
ventricular dilation in response to small increases
in preload [89,90]. Disruption of the anatomy of
the tricuspid valve can lead to severe tricuspid
regurgitation and significant right ventricular
dilation [91–93]. Pericardial absence can affect
systemic venous return, and, to a minor extent,
pulmonary venous return [94]. Finally, total
absence of the pericardium is associated with
abnormal interventricular septal movement,
which relates to left ventricular rotation and
motion abnormality [95].
Medication
Beta-blockers, calcium antagonists, and anti-
arrhythmic drugs with negative inotropic and
chronotropic effects are often used in ACHD
and can affect ventricular performance and exer-
cise capacity. In the authors’ experience ACHD
patients on beta-blocker therapy have a lower
peak heart rate during maximal exercise testing
and lower peak VO2 than patients not on
649
beta-blockers [44]. Underlying conduction system
disease is often latent in patients who have
ACHD, and medication can unmask sinus node
dysfunction, atrioventricular block, and chrono-
tropic incompetence [96]. Particular care and close
follow-up are thus needed when such therapies are
prescribed in this population.
Other conditions that can affect exercise capacity
Cyanosis, pulmonary vascular disease,
and Eisenmenger physiology
Cyanosis and pulmonary hypertension signifi-
cantly affect exercise capacity. Objective data
confirm that Eisenmenger and complex patients
with cyanosis are impaired significantly in their
ability to exercise [1,17]. The effect of right-to-left
shunting is obvious from the onset of exercise,
when VO2 fails to increase because of the inability
to sufficiently increase pulmonary blood flow
[8,97]. In these patients, an increase in cardiac out-
put is obtained through shunting, at the expense
of further systemic desaturation. An abrupt, exag-
gerated increase in ventilation occurs almost at
the onset of exercise, resulting in alveolar hyper-
ventilation, a rise in VCO2 and a drop in VO2
[97]. This accounts for the transient increase in
the respiratory quotient (the ratio between
VCO2 and VO2) often encountered in these pa-
tients at the start of exercise.
Although ventilation is increased, ventilatory
efficiency is impaired significantly, as suggested by
very high values of the VE/VCO2 slope in cya-
notic patients [16]. Pulmonary hypoperfusion, an
increase in physiological dead-space through
right-to-left shunting, and enhanced ventilatory
reflex sensitivity are potential mechanisms con-
tributing to the ventilatory inefficiency and the
failure to meet oxygen requirements [16]. Al-
though the inefficiency of the ventilatory response
to exercise in cyanotic patients suggested by the
VE/VCO2 slope could lead to the early onset of
dyspnea and exercise limitation, the exaggerated
ventilatory response in these patients appears
appropriate from a chemical point of view. In
fact, hyperventilation succeeds in maintaining
near-normal arterial PCO2 and pH levels in
the systemic circulation, at least during mild-
to-moderate exertion, despite significant right-
to-left shunting [40].
The effect of cyanosis on exercise capacity and
ventilation is difficult to distinguish from that of
pulmonary hypertension. Significant ventilatory
650 DIMOPOULOS
inefficiency and a hyperventilatory response to
exercise have been described in patients with
primary pulmonary hypertension in the absence
of right-to-left shunting [98,99]. The authors re-
cently reported that patients with significant pul-
monary hypertension and cyanosis have
significantly higher VE/VCO2 slopes compared
with those without cyanosis, suggesting an addi-
tive effect of cyanosis on ventilation over that of
pulmonary hypertension alone. Moreover, Fon-
tan patients with cyanosis have higher VE/VCO2
slopes compared with those without, suggesting
a significant effect of cyanosis on ventilation
even in the absence of pulmonary arterial hyper-
tension [16].
Anemia and secondary erythrocytosis
Adequate blood oxygen carrying capacity is
fundamental for the normal response to exercise.
Hemoglobin carries oxygen and regulates the
amount released at the tissue level [8,86]. Anemia
results in reduced oxygen carrying capacity and
a premature shift to anaerobic metabolism during
exercise. In acquired chronic heart failure, anemia
relates to exercise capacity and is a predictor of
outcome. It is also known to precipitate heart fail-
ure, by affecting myocardial function and volume
overload. Anemia in ACHD can occur as the
result of pregnancy or menorrhagia, gastrointesti-
nal blood loss, recent surgery or intervention,
hemolysis caused by prosthetic valves, intracar-
diac patches or endocarditis, spontaneous bleed-
ing from arteriovenous fistulas in the gut or
lungs, hemoptysis in patients with severe pulmo-
nary hypertension, or the use of anticoagulants.
Moreover, anemia can occur because of chronic
renal failure or other chronic disease.
Anemia, as conventionally defined, is rarely
encountered in cyanotic patients. Chronic hypoxia
results in an increase in erythropoietin production
and an isolated rise in the red blood cell count
(secondary erythrocytosis) [100]. This is a compen-
satory mechanism aimed at augmenting the
amount of oxygen delivered to the tissues in the
presence of significant right-to-left shunting. Rela-
tive anemia (ie, an inadequate rise in hemoglobin
levels despite chronic cyanosis) can occur as a re-
sult of iron deficiency and is often detrimental to
these delicate patients [101]. Unfortunately, there
is no accepted algorithm for the calculation of ap-
propriate hemoglobin levels, although it generally
is accepted that a patient with a resting oxygen
saturation of less than 85% should have
et al
a hemoglobin concentration of at least 18 g/dL.
The diagnosis of relative anemia remains intuitive,
and serum ferritin and transferrin saturation
should be monitored [101].
Pulmonary function
There are data to suggest that pulmonary
function in ACHD is impaired and could affect
exercise capacity. Subnormal forced vital capacity
has been reported in patients with Ebstein’s
anomaly, tetralogy of Fallot, congenitally cor-
rected (l-) transposition, Fontan operation, atrial
repair of transposition, and atrial septal defects
[37,102,103]. In the authors’ experience, pulmo-
nary function expressed as percent of %FEV1
and forced vital capacity (%FVC) was depressed
significantly in patients who have ACHD.
%FEV1 was a powerful predictor of exercise
capacity in this population [44]. Possible mecha-
nisms for the abnormal pulmonary function
observed in patients who have ACHD include
prior surgery with lung scarring, atelectasis, chest
deformities, diaphragmatic palsy, pulmonary vas-
cular disease with loss of distensibility of periph-
eral arteries, significant cardiomegaly, and
reduced blood supply to portions of the lung
that become hypoplastic such as in the case of
scimitar syndrome and pulmonary atresia.
Endothelial dysfunction
Endothelial dysfunction appears to play an
important role in the pathophysiology of exercise
intolerance in acquired heart failure [104,105]. It
correlates to the degree of impairment and to
brain natriuretic peptide levels. It is a predictor
of outcome in patients with advanced heart failure
and improves after heart failure treatment
[45,106,107]. It is postulated that endothelial dys-
function has a detrimental effect on myocardial
and skeletal muscle function. Evidence of endo-
thelial dysfunction in CHD is available for pediat-
ric patients after a Fontan-type operation and in
those with Kawasaki disease [108–110]. In the
Fontan group, endothelial function was related
to exercise capacity, supporting the role of the en-
dothelium in exercise adaptation [100]. Oechslin
and colleagues recently provided evidence of sys-
temic endothelial dysfunction and reduced basal
nitric oxide availability in ACHD patients with
cyanosis [111].
 EXERCISE INTOLERANCE IN ACHD 651

Neurohormonal activation in adult congenital Sharma and colleagues reported high levels of
heart disease cytokine expression in patients who have ACHD
[122]. Tumor necrosis factor (TNF) levels corre-
Although cardiac dysfunction is the primary
lated well to disease severity expressed as
cause of the heart failure syndrome, the periphery
NYHA class. Patients with peripheral edema
plays a fundamental role in the pathophysiology
had significantly higher endotoxin levels, a potent
of heart failure [112]. The reduction in blood flow
stimulus for inflammatory cytokine production
to organs such as the kidney and the skeletal mus-
[123]. Cyanotic patients in particular were found
cle appears to cause hyperactivation of the neuro-
to have significantly increased levels of inflamma-
endocrine system and of peripheral reflexes that
tory cytokines and a trend toward higher endo-
control blood pressure and the ventilatory and
toxin levels [109]. Hypoxia is, in fact, a known
chronotropic response to exertion. These, in
trigger for inflammatory cytokine release in pa-
turn, may be responsible for the onset of symp-
tients with heart failure [124]. Although TNF re-
toms limiting exercise capacity and for perpetuat-
ceptor (TNFR) concentrations in the serum were
ing and aggravating heart failure itself. Modern
not significantly higher in patients compared
heart failure treatment is based on agents such
with controls, a significant correlation was seen
as angiotensin-converting enzyme inhibitors,
between levels of TNFR-1 and systemic ventricu-
beta adrenergic blockers, and aldosterone, which
lar function.
are aimed at reversing these changes.
Similar studies on immune activation in CHD
Elevated levels of the atrial natriuretic peptide
have focused on the pediatric population, espe-
were first described in children with CHD by Lang
cially in relation to surgery [125–127]. Lequier and
and colleagues [113]. Thereafter, neurohormone
colleagues noted that 40% of infants with congen-
levels were proposed as a means of assessing he-
ital heart defects were significantly endotoxemic,
modynamic status, the adequacy of medical treat-
with an adverse impact on clinical outcome
ment, and the effect of surgery in such patients
[128]. Surgery, nevertheless, represents a strong
[114–116]. Bolger and colleagues described ele-
stimulus for the production of inflammatory cyto-
vated levels of the natriuretic peptides, endothe-
kines, and the results of these studies need to be
lin-1, norepinephrine, renin, and aldosterone in
interpreted with caution before making inferences
patients who had ACHD, even when asymptom-
to ACHD patients in general.
atic [117]. Natriuretic peptides and endothelin
levels were related to functional (NYHA) class,
peak VO2 and systemic ventricular function, and
ECG and radiographic parameters. Activation Heart rate variability, baroreflex,
of the natriuretic peptides occurs in a spectrum and chemoreflex sensitivity
of congenital and noncongenital cardiac diseases
Deranged cardiac autonomic nervous activity
and remains elevated for years after surgical re-
is associated with adverse outcome in heart failure
pair of even relatively simple lesions [118]. Natri-
and coronary artery disease [129,130]. Heart rate
uretic peptide levels, therefore, appear to be
variability and baroreflex sensitivity, both
a sensitive marker of cardiac dysfunction in
markers of cardiac autonomic nervous activity,
ACHD and in acquired heart failure [119].
are strong independent predictors of death after
myocardial infarction and in chronic heart failure
[106,131–133]. Evidence of autonomic dysfunc-
tion has been observed in ACHD [109,134,135].
Cytokine activation
Markedly deranged cardiac autonomic nervous
Activation of cytokines also appears to play control late after repair of tetralogy of Fallot
a role in the pathophysiology of heart failure, and in patients after a Fontan-type operation
especially in advanced stages. Levine and col- has been reported by the authors’ group
leagues showed that chronic heart failure is [136,137]. Autonomic imbalance was related to
associated with cytokine activation [120]. Rauch- the propensity for arrhythmias and to hemody-
haus and colleagues extended this concept by namic parameters such as the degree of pulmo-
demonstrating that raised inflammatory cytokine nary regurgitation, right ventricular end-systolic
and cytokine receptor levels are strong prognostic pressures, left and right ventricular ejection frac-
markers in heart failure and seem to have direct tions, and QRS duration in patients with tetralogy
toxic effects on the myocardium [121]. of Fallot. Even though cardiac denervation
652 DIMOPOULOS
secondary to surgery could play a role in cardiac
autonomic derangement, a clear relationship be-
tween autonomic indices and hemodynamic status
was present. The association of abnormal heart
rate variability to both a history of arrhythmias
and known markers of ventricular tachycardia
and sudden cardiac death suggests a possible
prognostic impact of autonomic parameters in
ACHD.
Chemoreflex and peripheral metaboreflex (er-
goreflex) hypersensitivity are recognized features
of chronic heart failure and important predictors
of reduced exercise tolerance and adverse prog-
nosis [138,139]. Moreover, periodic breathing,
characterized by oscillations in the depth of venti-
lation associated with oscillations in heart rate
and blood pressure, is a powerful predictor of
poor outcome in acquired heart failure [140].
Few data exist on the chemoreflex and ergoreflex
sensitivity in ACHD. A small study by Brassard
and colleagues reported that patients after a
Fontan-type operation had excessive contribution
of the ergoreflex to blood pressure but not to ven-
tilatory variations. Moreover, Fontan patients
had abnormal skeletal muscle function expressed
as time to fatigue. After an 8-week training pro-
gram, abnormal ergoreflex hyperactivation was
reduced [141]. Georgiadou and colleagues re-
ported two cases of periodic breathing and abnor-
mal chemoreflex sensitivity in previously repaired
tetralogy of Fallot [142].
The prognostic value of parameters of exercise
intolerance
Cardiopulmonary exercise testing is a useful
tool for identifying patients at risk of hospitaliza-
tion and death, and thus in need of particular
attention and additional resources. Peak oxygen
consumption and the VE/VCO2 slope are estab-
lished prognostic markers for the acquired heart
failure population [1,2]. The authors’ group re-
ported that peak VO2 was an independent predic-
tor of outcome in the medium term (median
follow-up 304 days) expressed as death or hospi-
talization [44]. Patients with a peak VO2 less
than 15.5 mL/kg/min demonstrated a 2.9-fold in-
creased risk of hospital admission or death com-
pared with patients with a peak VO2 above 15.5
mL/kg/min. Moreover, peak VO2 was related to
the frequency and duration of hospitalization ad-
justed for follow-up time, even after accounting
for NYHA class, age, age at surgery, and gender.
et al
The VE/VCO2 slope is also a strong predictor of
outcome in ACHD and is stronger than peak
VO2 in predicting death in patients without cya-
nosis [16]. A VE/VCO2 slope above 38 identified
noncyanotic patients at higher risk of death in
the mid-term. In the overall ACHD population,
peak VO2 remained the strongest independent
predictor of outcome.
Improving exercise tolerance in adult congenital
heart disease
Surgical and interventional options
Because cardiac dysfunction is a major de-
terminant of exercise intolerance in ACHD,
improved cardiac hemodynamics should be the
first target in the effort to improve exercise
capacity. Potential therapeutic options include
surgical or interventional relief of obstructive
lesions, repair of valve abnormalities, and closure
of shunts. Improvement in symptoms has been
reported after interventions such as Fontan-type
operations, tetralogy of Fallot repair, relief of
congenital aortic stenosis, and transcutaneous
closure of ASD [143–147].
Resynchronization therapy has a role for
treating patients who have left ventricular dys-
function and intraventricular conduction delay
[148–150]. Preliminary studies of ventricular re-
synchronization in selected congenital cohorts
have provided promising results. In biventricular
hearts with significant intraventricular delay,
Janousek and colleagues and Zimmerman and
colleagues reported an increase in systolic blood
pressure and cardiac index with biventricular pac-
ing in the postoperative period [151,152]. Dubin
and colleagues showed an acute increase in car-
diac output and right ventricular systolic perfor-
mance in patients who had intraventricular delay
[153]. Recently, a multi-center registry reported
the results of biventricular pacing on 103 pediatric
patients, 73 of whom had CHD [154]. Overall,
a significant increase in systemic ventricular ejec-
tion fraction and a decrease in QRS duration
were observed. These findings were confirmed
when patients with a systemic right ventricle
were examined separately, but they were not
found in the overall CHD group. In univentricu-
lar hearts, QRS duration decreased significantly,
and a trend for an increase in ejection fraction
was seen (P ¼ .08), despite the small sample
size. Thus, resynchronization therapy may benefit
patients who have CHD. Further, prospective
 EXERCISE INTOLERANCE IN ACHD 653

studies are needed to establish the beneficial ef- heart disease are well-established [172]; however,
fects of biventricular pacing in ACHD and to limited data exist on the effect of exercise training
identify patients who are more likely to benefit in patients who have ACHD. Systematic training
from resynchronization therapy. programs have only been introduced to small co-
horts of ACHD patients, and most studies con-
clude that the exercise training is safe and might
Medical management be beneficial, although no significant effect on ex-
ercise capacity has been demonstrated [173,174].
Neurohormonal blockade is the cornerstone of
It is generally accepted that physical exercise of
modern chronic heart failure treatment. Angio-
appropriate intensity should be encouraged in
tensin-converting enzyme inhibitors (ACEi), an-
most patients who have ACHD. Recommenda-
giotensin receptor antagonists, spironolactone,
tions for the participation of patients with CHD
and beta blockers not only improve hemodynam-
in sports according to the underlying anatomy
ics but also slow the progression of heart failure
and physiology were issued after the 36th
and improve prognosis. Unfortunately, only lim-
Bethesda Conference [175]. Counseling should,
ited data are available on the use of ACEi and
nevertheless, also include an estimation of the
beta blockers for improving exercise tolerance in
type of exercise and its potential impact on car-
the congenital population [155]. ACEi are often
diac hemodynamics [176]. Exercise testing can be
used empirically, especially in cases of right and
invaluable in aiding clinical judgment, as it re-
left ventricular volume overload [156–160]. The
produces the conditions of stress during sports
evidence for the use of beta blockers is even scant-
activities under electrocardiographic and blood
ier [161–164].
pressure monitoring. High-impact sport is best
Recently, numerous drugs have become avail-
avoided in patients with Marfan’s syndrome or
able for treating primary pulmonary hyperten-
other aortic anomalies and those on anticoagula-
sion, some of which may be applicable to ACHD
tion therapy or carrying a pacemaker. Particular
patients who have pulmonary arterial hyperten-
attention should be paid to patients at high risk
sion. Fernadez and colleagues recently demon-
of arrhythmia and sudden death, such as those
strated that epoprostenol improves functional
with arrhythmogenic right ventricular dysplasia,
capacity, systemic saturations, and pulmonary
hypertrophic obstructive cardiomyopathy and
hemodynamics in patients who have CHD and
long QT syndrome. Recommendations should be
pulmonary arterial hypertension [165]. The use of
discussed with patients, and simple preventive
epoprostenol is, however, limited by its invasive
measures such avoiding excessive dehydration
nature and associated complications such as line
should be suggested.
and systemic infections and line dislocations. Sil-
denafil, an oral phospodiasterase-5-inhibitor, has
been shown to improve functional capacity in pa- Is exercise intolerance in adult congenital heart
tients who have pulmonary arterial hypertension, disease heart failure?
including some with CHD [166]. Bosentan, an
It is clear that important similarities exist
oral dual receptor endothelin antagonist, has
between ACHD and heart failure syndrome
been demonstrated to improve exercise capacity
[177]. Current heart failure management is evi-
in patients with Eisenmenger syndrome, both in
dence-based, modeled on the results of large ran-
several open-label intention-to-treat pilot studies,
domized controlled trials with hard end points
and in a recently reported randomized placebo-
such as death. Such evidence is difficult to acquire
controlled study [167–171]. Although long-term
in a relatively small and heterogeneous population
data are lacking, Bosentan appears to be the
such as that of ACHD patients, with a relatively
most promising treatment for highly symptomatic
low overall annual mortality rate. Despite the
ACHD patients who have significant pulmonary
lack of evidence, heart failure medication such
arterial hypertension.
as ACE inhibitors is used widely in ACHD. An
effort should be made to gather data on the poten-
tial beneficial effects of established heart failure
Exercise training and recommendations
therapies through the creation of a network of
for exercise
ACHD centers and the design of registries and
The psychological and physical benefits of multi-center studies. Conversely, investigations
exercise training on patients who have acquired of the pathophysiology of exercise intolerance in
654 DIMOPOULOS
ACHD could provide interesting insights for un-
derstanding exercise limitation in acquired heart
failure.
Summary
Exercise intolerance is prevalent in the ACHD
population and affects patients from all diagnostic
groups, including those with simple lesions. Car-
diopulmonary exercise testing, with the calculation
of peak VO2 and the VE/VCO2 slope, provides a re-
liable tool for assessing the exercise capacity of
ACHD patients and for risk stratification. This
should become part of the routine clinical assess-
ment of these patients. The relief of hemodynamic
burdens and reconditioning through exercise train-
ing appear to be the management of choice. More-
over, similarities in the pathophysiology of exercise
intolerance in acquired heart failure and CHD sug-
gest that established heart failure therapies might
be beneficial to ACHD patients with exercise intol-
erance, but their safety and efficacy should be es-
tablished by multi-center randomized controlled
trials.
References
[1] Engelfriet P, Boersma E, Oechslin E, et al. The
spectrum of adult congenital heart disease in
Europe: morbidity and mortality in a 5 year
follow-up period. The Euro Heart Survey on adult
congenital heart disease. Eur Heart J 2005;26(21):
2325–33.
[2] Culbert EL, Ashburn DA, Cullen-Dean G, et al.
Quality of life of children after repair of transposi-
tion of the great arteries. Circulation 2003;108:
857–62.
[3] Kamphuis M, Ottenkamp J, Vliegen HW, et al.
Health-related quality of life and health status in
adult survivors with previously operated complex
congenital heart disease. Heart 2002;87:356–62.
[4] Moons P, Van Deyk K, Budts W, et al. Caliber of
quality-of-life assessments in congenital heart dis-
ease: a plea for more conceptual and methodologi-
cal rigor. Arch Pediatr Adolesc Med 2004;158(11):
1062–9.
[5] Hunt SA. American College of Cardiology. ACC/
AHA 2005 guideline update for the diagnosis and
management of chronic heart failure in the adult:
a report of the American College of Cardiology/
American Heart Association Task Force on Prac-
tice Guidelines (Writing Committee to Update the
2001 Guidelines for the Evaluation and Manage-
ment of Heart Failure). J Am Coll Cardiol 2005;
46(6):e1–82.
[6] Swedberg K, Cleland J, Dargie H, et al. Guidelines
for the diagnosis and treatment of chronic heart
et al
failure: executive summary (update 2005): the
Task Force for the Diagnosis and Treatment of
Chronic Heart Failure of the European Society of
Cardiology. Eur Heart J 2005;26(11):1115–40.
[7] Bolger AP, Gatzoulis MA. Towards defining heart
failure in adults with congenital heart disease. Int J
Cardiol 2004;97(Suppl 1):15–23.
[8] Kamphuis M, Zwinderman KH, Vogels T, et al. A
cardiac-specific health-related quality of life mod-
ule for young adults with congenital heart disease:
development and validation. Qual Life Res 2004;
13(4):735–45.
[9] Wasserman K, Hansen J, Sue D, Whipp B. Exercise
testing and Interpretation: an overview. In: Princi-
ples of exercise testing and interpretation. Balti-
more (MD): Lippincott Williams and Wilkins;
1999. p. 1–9.
[10] Cooper CB, Storer TW. Purpose. In: Exercise test-
ing and interpretation. Cambridge (UK): Cam-
bridge University Press; 2001. p. 1–14.
[11] Wasserman K, Hansen J, Sue D, et al. Measure-
ments during integrative cardiopulmonary exercise
testing. In: Principles of exercise testing and inter-
pretation. Baltimore (MD): Lippincott Williams
and Wilkins; 1999. p. 62–94.
[12] Cooper CB, Storer TW. Response variables. In:
Exercise testing and interpretation. Cambridge
(UK): Cambridge University Press; 2001. p. 93–148.
[13] Task Force of the Italian Working Group on Car-
diac Rehabilitation Prevention. Statement on car-
diopulmonary exercise testing in chronic heart
failure due to left ventricular dysfunction: recom-
mendations for performance and interpretation
Part I: definition of cardiopulmonary exercise test-
ing parameters for appropriate use in chronic heart
failure. Eur J Cardiovasc Prev Rehabil 2006;13(2):
150–64.
[14] Gitt AK, Wasserman K, Kilkowski C, et al. Exer-
cise anaerobic threshold and ventilatory efficiency
identify heart failure patients for high risk of early
death. Circulation 2002;106:3079–84.
[15] Francis DP, Shamim W, Davies LC, et al. Cardio-
pulmonary exercise testing for prognosis in chronic
heart failure: continuous and independent prognos-
tic value from VE/VCO2 slope and peak VO2. Eur
Heart J 2000;21(2):154–61.
[16] Dimopoulos K, Okonko DO, Diller GP, et al. Ab-
normal ventilatory response to exercise in adults
with congenital heart disease relates to cyanosis
and predicts survival. Circulation 2006;113:
2796–802.
[17] Ponikowski P, Francis DP, Piepoli MF, et al. En-
hanced ventilatory response to exercise in patients
with chronic heart failure and preserved exercise
tolerance: marker of abnormal cardiorespiratory
reflex control and predictor of poor prognosis.
Circulation 2001;103:967–72.
[18] Chua TP, Ponikowski P, Harrington D, et al. Clin-
ical correlates and prognostic significance of the
 EXERCISE INTOLERANCE IN ACHD
ventilatory response to exercise in chronic heart
failure. J Am Coll Cardiol 1997;29(7):1585–90.
[19] Tabet JY, Beauvais F, Thabut G, et al. A critical
appraisal of the prognostic value of the VE/VCO2
slope in chronic heart failure. Eur J Cardiovasc
Prev Rehabil 2003;10:267–72.
[20] Cooper CB, Storer TW. Testing methods. In: Exer-
cise testing and interpretation. Cambridge (UK):
Cambridge University Press; 2001. p. 51–92.
[21] Niedeggen A, Skobel E, Haager P, et al. Compari-
son of the 6-minute walk test with established pa-
rameters for assessment of cardiopulmonary
capacity in adults with complex congenital cardiac
disease. Cardiol Young 2005;15(4):385–90.
[22] Hoeper MM, Oudiz RJ, Peacock A, et al. End
points and clinical trial designs in pulmonary arte-
rial hypertension: clinical and regulatory perspec-
tives. J Am Coll Cardiol 2004;43:48S–55S.
[23] ATS Committee on Proficiency Standards for Clin-
ical Pulmonary Function Laboratories. ATS state-
ment: guidelines for the six-minute walk test. Am J
Respir Crit Care Med 2002;166:111–7.
[24] Oudiz RJ, Barst RJ, Hansen JE, et al. Cardiopul-
monary exercise testing and six-minute walk corre-
lations in pulmonary arterial hypertension. Am J
Cardiol 2006;97(1):123–6.
[25] Frost AE, Langleben D, Oudiz R, et al. The 6-min
walk test (6MW) as an efficacy end point in pulmo-
nary arterial hypertension clinical trials: demon-
stration of a ceiling effect. Vascul Pharmacol
2005;43(1):36–9.
[26] Wu G, Sanderson B, Bittner V. The 6-minute walk
test: how important is the learning effect? Am Heart
J 2003;146(1):129–33.
[27] Veldtman GR, Nishimoto A, Siu S, et al. The Fon-
tan procedure in adults. Heart 2001;86(3):330–5.
[28] Alphonso N, Baghai M, Sundar P, et al. Intermedi-
ate-term outcome following the Fontan operation:
a survival, functional and risk-factor analysis. Eur J
Cardiothorac Surg 2005;28(4):529–35.
[29] Daliento L, Somerville J, Presbitero P, et al. Eisen-
menger syndrome. Factors relating to deterioration
and death. Eur Heart J 1998;19:1845–55.
[30] Graham TP Jr, Bernard YD, Mellen BG, et al.
Long-term outcome in congenitally corrected
transposition of the great arteries: a multi-institu-
tional study. J Am Coll Cardiol 2000;36:255–61.
[31] Beauchesne LM, Warnes CA, Connolly HM, et al.
Outcome of the unoperated adult who presents
with congenitally corrected transposition of the
great arteries. J Am Coll Cardiol 2002;40:285–90.
[32] Piran S, Veldtman G, Siu S, et al. Heart failure and
ventricular dysfunction in patients with single or
systemic right ventricles. Circulation 2002;105(10):
1189–94.
[33] Norozi K, Buchhorn R, Kaiser C, et al. Plasma N-
terminal probrain natriuretic peptide as a marker
of right ventricular dysfunction in patients with
655
tetralogy of Fallot after surgical repair. Chest
2005;128(4):2563–70.
[34] de Ruijter FT, Weenink I, Hitchcock F, et al. Right
ventricular dysfunction and pulmonary valve re-
placement after correction of tetralogy of Fallot.
Ann Thorac Surg 2002;73:1794–800.
[35] Sietsema KE, Cooper DM, Perloff JK, et al. Dy-
namics of oxygen uptake during exercise in adults
with cyanotic congenital heart disease. Circulation
1986;73:1137–44.
[36] Clark AL, Swan JW, Laney R, et al. The role of
right and left ventricular function in the ventilatory
response to exercise in chronic heart failure. Circu-
lation 1994;89:2062–9.
[37] Fredriksen PM, Veldtman G, Hechter S, et al. Aer-
obic capacity in adults with various congenital
heart diseases. Am J Cardiol 2001;87:310–4.
[38] Hechter SJ, Webb G, Fredriksen PM, et al. Cardio-
pulmonary exercise performance in adult survivors
of the Mustard procedure. Cardiol Young 2001;11:
407–14.
[39] Fredriksen PM, Chen A, Veldtman G, et al. Exer-
cise capacity in adult patients with congenitally cor-
rected transposition of the great arteries. Heart
2001;85:191–5.
[40] Glaser S, Opitz CF, Bauer U, et al. Assessment of
symptoms and exercise capacity in cyanotic pa-
tients with congenital heart disease. Chest 2004;
125(2):368–76.
[41] Harrison DA, Liu P, Walters JE, et al. Cardiopul-
monary function in adult patients late after Fontan
repair. J Am Coll Cardiol 1995;26:1016–21.
[42] Fredriksen PM, Veldtman G, Hechter S, et al. Aer-
obic capacity in adults with various congenital
heart diseases. Am J Cardiol 2001;87:310–4.
[43] Fredriksen PM, Therrien J, Veldtman G, et al. Aer-
obic capacity in adults with tetralogy of Fallot.
Cardiol Young 2002;12:554–9.
[44] Diller GP, Dimopoulos K, Okonko D, et al. Exer-
cise intolerance in adult congenital heart disease:
comparative severity, correlates, and prognostic
implication. Circulation 2005;112(6):828–35.
[45] Clark AL, Poole-Wilson PA, Coats AJ. Exercise
limitation in chronic heart failure: central role of
the periphery. J Am Coll Cardiol 1996;28(5):
1092–102.
[46] Lipkin DP, Poole-Wilson PA. Symptoms limiting
exercise in chronic heart failure. Br Med J (Clin
Res Ed) 1986;292(6527):1030–1.
[47] Reybrouck T, Boshoff D, Vanhees L, et al. Ventila-
tory response to exercise in patients after correction
of cyanotic congenital heart disease: relation with
clinical outcome after surgery. Heart 2004;90:
215–6.
[48] Wasserman K, Stringer WW, Sun XG, et al. Circu-
latory coupling of external to muscler respiration
during exercise. In: Wasserman K, editor. Cardio-
pulmonary exercise testing and cardiovascular
656 DIMOPOULOS
health. Armonk (NY): Futura Publishing Com-
pany; 2002. p. 3–26.
[49] Takagaki M, Ishino K, Kawada M, et al. Total
right ventricular exclusion improves left ventricular
function in patients with end-stage congestive right
ventricular failure. Circulation 2003;108:II226–9.
[50] Davlouros PA, Kilner PJ, Hornung TS, et al. Right
ventricular function in adults with repaired tetral-
ogy of Fallot assessed with cardiovascular mag-
netic resonance imaging: detrimental role of right
ventricular outflow aneurysms or akinesia and ad-
verse right-to-left ventricular interaction. J Am
Coll Cardiol 2002;40:2044–52.
[51] Hurwitz RA, Caldwell RL, Girod DA, et al. Right
ventricular systolic function in adolescents and
young adults after Mustard operation for transpo-
sition of the great arteries. Am J Cardiol 1996;77:
294–7.
[52] Hornung TS, Kilner PJ, Davlouros PA, et al. Ex-
cessive right ventricular hypertrophic response in
adults with the Mustard procedure for transposi-
tion of the great arteries. Am J Cardiol 2002;90:
800–3.
[53] Lubiszewska B, Gosiewska E, Hoffman P, et al.
Myocardial perfusion and function of the systemic
right ventricle in patients after atrial switch proce-
dure for complete transposition: long-term fol-
low-up. J Am Coll Cardiol 2000;36(4):1365–70.
[54] Dodge-Khatami A, Tulevski II, Bennink GB, et al.
Comparable systemic ventricular function in
healthy adults and patients with unoperated con-
genitally corrected transposition using MRI dobut-
amine stress testing. Ann Thorac Surg 2002;73:
1759–64.
[55] Tulevski II, Zijta FM, Smeijers AS, et al. Regional
and global right ventricular dysfunction in asymp-
tomatic or minimally symptomatic patients with
congenitally corrected transposition. Cardiol
Young 2004;14(2):168–73.
[56] Hornung TS, Bernard EJ, Celermajer DS, et al.
Right ventricular dysfunction in congenitally
corrected transposition of the great arteries. Am J
Cardiol 1999;84:1116–9.
[57] Millane T, Bernard EJ, Jaeggi E, et al. Role of
ischemia and infarction in late right ventricular
dysfunction after atrial repair of transposition of
the great arteries. J Am Coll Cardiol 2000;35:
1661–8.
[58] Davlouros PA, Niwa K, Webb G, et al. The right
ventricle in congenital heart disease. Heart 2006;
92(Suppl 1):i27–38.
[59] Yoerger DM, Marcus F, Sherrill D, et al. Echocar-
diographic findings in patients meeting task force
criteria for arrhythmogenic right ventricular dys-
plasia: new insights from the multidisciplinary
study of right ventricular dysplasia. J Am Coll Car-
diol 2005;45(6):860–5.
[60] Allen BS, Winkelmann JW, Hanafy H, et al. Retro-
grade cardioplegia does not adequately perfuse the
et al
right ventricle. J Thorac Cardiovasc Surg 1995;
109(6):1116–24.
[61] Li W, Somerville J, Gibson DG, et al. Effect of
atrial flutter on exercise tolerance in patients with
grown-up congenital heart (GUCH). Am Heart J
2002;144:173–9.
[62] Katritsis D, Camm AJ. Chronotropic incompe-
tence: a proposal for definition and diagnosis. Br
Heart J 1993;70:400–2.
[63] Astrand A. Aerobic work capacity in men and
women with special reference to age. Acta Physiol
Scand 1960;49:1–92.
[64] Witte KK, Cleland JG, Clark AL. Chronic heart
failure, chronotropic incompetence, and the effects
of beta blockade. Heart 2006;92:481–6.
[65] Lauer MS, Okin PM, Larson MG, et al. Impaired
heart rate response to graded exercise. Prognostic
implications of chronotropic incompetence in the
Framingham Heart Study. Circulation 1996;93:
1520–6.
[66] Lauer MS, Francis GS, Okin PM, et al. Impaired
chronotropic response to exercise stress testing as
a predictor of mortality. JAMA 1999;281:524–9.
[67] Jouven X, Empana JP, Schwartz PJ, et al. Heart-
rate profile during exercise as a predictor of sudden
death. N Engl J Med 2005;352:1951–8.
[68] Fredriksen PM, Ingjer F, Nystad W, et al. A com-
parison of VO2 (peak) between patients with con-
genital heart disease and healthy subjects, all aged
8–17 years. Eur J Appl Physiol Occup Physiol
1999;80(5):409–16.
[69] Diller GP, Okonko DO, von Haehling S, et al.
Chronotropic incompetence in adult patients with
systemic right ventricle and univentricular circula-
tion. J Am Coll Cardiol 2005;45:325A.
[70] Diller GP, Dimopoulos K, Okonko DO, et al.
Heart rate response during exercise and recovery
is an independent predictor of death in adult con-
genital heart disease patients. Available at: http://
www.abstractsonline.com/arch/RecordView.aspx?
LookupKey¼12345&recordID¼18423. Accessed
March 18, 2006.
[71] Colucci WS, Ribeiro JR, Rocco MB, et al. Im-
paired chronotropic response to exercise in patients
with congestive heart failure. Role of postsynaptic
beta-adrenergic desensitization. Circulation 1989;
80:314–23.
[72] Fei L, Keeling PJ, Sadoul N, et al. Decreased heart
rate variability in patients with congestive heart
failure and chronotropic incompetence. Pacing
Clin Electrophysiol 1996;19:477–83.
[73] Paul T, Blaufox AD, Saul JP. Invasive electrophys-
iology and pacing. In: Gatzoulis MA, Webb GD,
Daubeney PE, editors. Diagnosis and management
of adult congenital heart disease. London: Church-
ill Livingstone; 2003. p. 115–24.
[74] O’Keefe JH Jr, Abuissa H, Jones PG, et al. Effect of
chronic right ventricular apical pacing on left ven-
tricular function. Am J Cardiol 2005;95:771–3.
 EXERCISE INTOLERANCE IN ACHD
[75] Nielsen JC, Kristensen L, Andersen HR, et al. A
randomized comparison of atrial and dual chamber
pacing in 177 consecutive patients with sick sinus
syndrome: echocardiographic and clinical out-
come. J Am Coll Cardiol 2003;42:614–23.
[76] Thambo JB, Bordachar P, Garrigue S, et al. Detri-
mental ventricular remodeling in patients with con-
genital complete heart block and chronic right
ventricular apical pacing. Circulation 2004;
110(25):3766–72.
[77] Rinfret S, Cohen DJ, Lamas GA, et al. Cost-
effectiveness of dual-chamber pacing compared
with ventricular pacing for sinus node dysfunction.
Circulation 2005;111(2):165–72.
[78] Selden R, Schaefer RA, Kennedy BJ, et al. Cor-
rected transposition of the great arteries simulating
coronary heart disease in adults. Chest 1976;69(2):
188–91.
[79] Weindling SN, Wernovsky G, Colan SD, et al.
Myocardial perfusion, function and exercise toler-
ance after the arterial switch operation. J Am Coll
Cardiol 1994;23:424–33.
[80] Hayes AM, Baker EJ, Kakadeker A, et al. Influ-
ence of anatomic correction for transposition of
the great arteries on myocardial perfusion: radio-
nuclide imaging with technetium–99m 2–methoxy
isobutyl isonitrile. J Am Coll Cardiol 1994;24:
769–77.
[81] Bengel FM, Hauser M, Duvernoy CS, et al.
Myocardial blood flow and coronary flow reserve
late after anatomical correction of transposition
of the great arteries. J Am Coll Cardiol 1998;
32(7):1955–61.
[82] Babu-Narayan SV, Kilner PJ, Li W, et al. Ventric-
ular fibrosis suggested by cardiovascular magnetic
resonance in adults with repaired tetralogy of Fal-
lot and its relationship to adverse markers of clini-
cal outcome. Circulation 2006;113(3):405–13.
[83] Babu-Narayan SV, Goktekin O, Moon JC, et al.
Late gadolinium enhancement cardiovascular mag-
netic resonance of the systemic right ventricle in
adults with previous atrial redirection surgery for
transposition of the great arteries. Circulation
2005;111(16):2091–8.
[84] Senzaki H, Masutani S, Kobayashi J, et al. Ventric-
ular afterload and ventricular work in Fontan cir-
culation: comparison with normal two-ventricle
circulation and single-ventricle circulation with
Blalock-Taussig shunts. Circulation 2002;105:
2885–92.
[85] Singh TP, Humes RA, Muzik O, et al. Myocardial
flow reserve in patients with a systemic right ventri-
cle after atrial switch repair. J Am Coll Cardiol
2001;37(8):2120–5.
[86] Hauser M, Bengel FM, Kuhn A, et al. Myocardial
perfusion and coronary flow reserve assessed by
positron emission tomography in patients after
Fontan-like operations. Pediatr Cardiol 2003;
24(4):386–92.
657
[87] Meijboom FJ. Constrictive pericarditis and restric-
tive cardiomyopathy. In: Gatzoulis MA,
Webb GD, Daubeney PE, editors. Diagnosis and
management of adult congenital heart disease.
London: Churchill Livingstone; 2003. p. 459–66.
[88] Gatzoulis MA, Munk MD, Merchant N, et al.
Isolated congenital absence of the pericardium:
clinical presentation, diagnosis, and management.
Ann Thorac Surg 2000;69(4):1209–15.
[89] Abbas AE, Appleton CP, Liu PT, et al. Congenital
absence of the pericardium: case presentation and
review of literature. Int J Cardiol 2005;98(1):21–5.
[90] Rheuban KS. Pericardial diseases. In: Allen HD,
Gutgesell HP, Clark EB, editors. Heart disease
in infants, children and adolescents. 6th edition.
Philadelphia: Lippincott Williams and Wilkins;
2001. p. 1287–97.
[91] van Son JA, Danielson GK, Callahan JA. Congen-
ital absence of the pericardium: displacement of the
heart associated with tricuspid insufficiency. Ann
Thorac Surg 1993;56(6):1405–6.
[92] Chatzis AC, Giannopoulos NM, Sarris GE. Iso-
lated congenital tricuspid insufficiency associated
with right-sided congenital partial absence of the
pericardium. J Heart Valve Dis 2004;13(5):790–1.
[93] Goetz WA, Liebold A, Vogt F, et al. Tricuspid
valve repair in a case with congenital absence of
left thoracic pericardium. Eur J Cardiothorac
Surg 2004;26(4):848–9.
[94] Fukuda N, Oki T, Iuchi A, et al. Pulmonary and
systemic venous flow patterns assessed by transeso-
phageal Doppler echocardiography in congenital
absence of the pericardium. Am J Cardiol 1995;
75(17):1286–8.
[95] Oki T, Tabata T, Yamada H, et al. Cross sectional
echocardiographic demonstration of the mecha-
nisms of abnormal interventricular septal motion
in congenital total absence of the left pericardium.
Heart 1997;77(3):247–51.
[96] Harris L, Belaji S. Arrhythmias in the adult with
congenital heart disease. In: Gatzoulis MA,
Webb GD, Daubeney PE, editors. Diagnosis and
management of adult congenital heart disease.
London: Churchill Livingstone; 2003. p. 105–13.
[97] Sietsema KE, Cooper DM, Perloff JK, et al. Con-
trol of ventilation during exercise in patients with
central venous-to-systemic arterial shunts. J Appl
Physiol 1988;64(1):234–42.
[98] Oudiz RJ, Sun XG. Abnormalities in exercise gas
exchange in primary pulmonary hypertension. In:
Wasserman K, editor: Cardiopulmonary exercise
testing and cardiovascular health. Futura, Armonk
(NY): 2002. p. 179–90.
[99] Sun XG, Hansen JE, Oudiz RJ, et al. Exercise path-
ophysiology in patients with primary pulmonary
hypertension. Circulation 2001;104(4):429–35.
[100] Oechslin E. Eisenmenger’s syndrome. In:
Gatzoulis MA, Webb GD, Daubeney PE, editors.
Diagnosis and management of adult congenital
658 DIMOPOULOS
heart disease. London: Churchill Livingstone;
2003. p. 363–77.
[101] Broberg CS, Bax BE, Okonko DO, et al. Blood vis-
cosity and its relation to iron deficiency, symptoms,
and exercise capacity in adults with cyanotic con-
genital heart disease. J Am Coll Cardiol 2006;48:
356–65.
[102] Fredriksen PM, Therrien J, Veldtman G, et al.
Lung function and aerobic capacity in adult pa-
tients following modified Fontan procedure. Heart
2001;85(3):295–9.
[103] Sulc J, Andrle V, Hruda J, et al. Pulmonary func-
tion in children with atrial septal defect before
and after heart surgery. Heart 1998;80(5):484–8.
[104] Bank A, Lee P, Kubo S. Endothelial dysfunction in
patients with heart failure: relationship to disease
severity. J Card Fail 2000;6(1):29–36.
[105] Lerman A, Zeiher AM. Endothelial function: car-
diac events. Circulation 2005;111(3):363–8.
[106] Katz SD, Hryniewicz K, Hriljac I, et al. Vascular
endothelial dysfunction and mortality risk in pa-
tients with chronic heart failure. Circulation 2005;
111:310–4.
[107] Fischer D, Rossa S, Landmesser U, et al. Endothe-
lial dysfunction in patients with chronic heart fail-
ure is independently associated with increased
incidence of hospitalization, cardiac transplanta-
tion, or death. Eur Heart J 2005;26(1):65–9.
[108] Mahle WT, Todd K, Fyfe DA. Endothelial func-
tion following the Fontan operation. Am J Cardiol
2003;91(10):1286–8.
[109] Dhillon R, Clarkson P, Donald AE, et al. Endothe-
lial dysfunction late after Kawasaki disease. Circu-
lation 1996;94(9):2103–6.
[110] Deng YB, Li TL, Xiang HJ, et al. Impaired
endothelial function in the brachial artery after
Kawasaki disease and the effects of intravenous
administration of vitamin C. Pediatr Infect Dis J
2003;22(1):34–9.
[111] Oechslin E, Kiowski W, Schindler R, et al. Systemic
endothelial dysfunction in adults with cyanotic
congenital heart disease. Circulation 2005;112(8):
1106–12.
[112] Coats AJ. The muscle hypothesis of chronic heart
failure. J Mol Cell Cardiol 1996;28(11):2255–62.
[113] Lang RE, Unger T, Ganten D, et al. Alpha atrial
natriuretic peptide concentrations in plasma of
children with congenital heart and pulmonary dis-
eases. Br Med J (Clin Res Ed) 1985;291:1241.
[114] Kikuchi K, Nishioka K, Ueda T, et al. Relationship
between plasma atrial natriuretic polypeptide con-
centration and hemodynamic measurements in
children with congenital heart diseases. J Pediatr
1987;111:335–42.
[115] Ross RD, Daniels SR, Schwartz DC, et al. Return
of plasma norepinephrine to normal after resolu-
tion of congestive heart failure in congenital heart
disease. Am J Cardiol 1987;60:1411–3.
et al
[116] Ishikawa S, Miyauchi T, Sakai S, et al. Elevated
levels of plasma endothelin-1 in young patients
with pulmonary hypertension caused by congeni-
tal heart disease are decreased after successful
surgical repair. J Thorac Cardiovasc Surg 1995;
110:271–3.
[117] Genth-Zotz S, von Haehling S, Bolger AP, et al.
Neurohormonal activation and the chronic heart
failure syndrome in adults with congenital heart
disease. Circulation 2002;106:92–9.
[118] Iivainen TE, Groundstroem KW, Lahtela JT, et al.
Serum N-terminal atrial natriuretic peptide in adult
patients late after surgical repair of atrial septal
defect. Eur J Heart Fail 2000;2:161–5.
[119] Tulevski II, Groenink M, van Der Wall EE, et al.
Increased brain and atrial natriuretic peptides in
patients with chronic right ventricular pressure
overload: correlation between plasma neurohor-
mones and right ventricular dysfunction. Heart
2001;86:27–30.
[120] Levine B, Kalman J, Mayer L, et al. Elevated circu-
lating levels of tumor necrosis factor in severe
chronic heart failure. N Engl J Med 1990;323(4):
236–41.
[121] Rauchhaus M, Doehner W, Francis DP, et al.
Plasma cytokine parameters and mortality in pa-
tients with chronic heart failure. Circulation 2000;
102(25):3060–7.
[122] Sharma R, Bolger AP, Li W, et al. Elevated circu-
lating levels of inflammatory cytokines and bacte-
rial endotoxin in adults with congenital heart
disease. Am J Cardiol 2003;92(2):188–93.
[123] Niebauer J, Volk HD, Kemp M, et al. Endotoxin
and immune activation in chronic heart failure:
a prospective cohort study. Lancet 1999;
353(9167):1838–42.
[124] Hasper D, Hummel M, Kleber FX, et al. Systemic
inflammation in patients with heart failure. Eur
Heart J 1998;19:761–5.
[125] Ashraf SS, Tian Y, Zacharrias S, et al. Effects of
cardiopulmonary bypass on neonatal and pediatric
inflammatory profiles. Eur J Cardiothorac Surg
1997;12:862–8.
[126] Seghaye M, Duchateau J, Bruniaux J, et al. Inter-
leukin-10 release related to cardiopulmonary
bypass in infants undergoing cardiac operations.
J Thorac Cardiovasc Surg 1996;111:545–53.
[127] Casey WF, Hauser GJ, Hannallah RS, et al. Circu-
lating endotoxin and tumor necrosis factor during
pediatric cardiac surgery. Crit Care Med 1992;20:
1090–6.
[128] Lequier LL, Nikaidoh H, Leonard SR, et al. Preop-
erative and postoperative endotoxemia in children
with congenital heart disease. Chest 2000;117:
1706–12.
[129] Zipes DP, Levy MN, Cobb LA, et al. Sudden car-
diac death. Neural-cardiac interactions. Circula-
tion 1987;76:I202–7.
 EXERCISE INTOLERANCE IN ACHD
[130] Schwartz PJ. The autonomic nervous system and
sudden death. Eur Heart J 1998;19(Suppl F):
F72–80.
[131] Kleiger RE, Miller JP, Bigger JT Jr, et al. De-
creased heart rate variability and its association
with increased mortality after acute myocardial in-
farction. Am J Cardiol 1987;59:256–62.
[132] La Rovere MT, Bigger JT Jr, Marcus FI, et al. Bar-
oreflex sensitivity and heart-rate variability in pre-
diction of total cardiac mortality after myocardial
infarction. ATRAMI (Autonomic Tone and Re-
flexes After Myocardial Infarction) Investigators.
Lancet 1998;351:478–84.
[133] Ponikowski P, Anker SD, Chua TP, et al. De-
pressed heart rate variability as an independent pre-
dictor of death in chronic congestive heart failure
secondary to ischemic or idiopathic dilated cardio-
myopathy. Am J Cardiol 1997;79:1645–50.
[134] McLeod KA, Hillis WS, Houston AB, et al. Re-
duced heart rate variability following repair of te-
tralogy of Fallot. Heart 1999;81:656–60.
[135] Folino AF, Russo G, Bauce B, et al. Autonomic
profile and arrhythmic risk stratification after sur-
gical repair of tetralogy of Fallot. Am Heart J
2004;148(6):985–9.
[136] Davos CH, Francis DP, Leenarts MF, et al. Global
impairment of cardiac autonomic nervous activity
late after the Fontan operation. Circulation 2003;
108(Suppl 1):II180–5.
[137] Davos CH, Davlouros PA, Wensel R, et al. Global
impairment of cardiac autonomic nervous activity
late after repair of tetralogy of Fallot. Circulation
2002;106:I69–75.
[138] Piepoli M, Clark AL, Volterrani M, et al. Contri-
bution of muscle afferents to the hemodynamic, au-
tonomic, and ventilatory responses to exercise in
patients with chronic heart failure: effects of physi-
cal training. Circulation 1996;93(5):940–52.
[139] Ponikowski P, Chua TP, Anker SD, et al. Periph-
eral chemoreceptor hypersensitivity: an ominous
sign in patients with chronic heart failure. Circula-
tion 2001;104(5):544–9.
[140] Francis DP, Willson K, Davies LC, et al. Quantita-
tive general theory of periodic breathing in chronic
heart failure and its clinical implications. Circula-
tion 2000;102:2214–21.
[141] Brassard P, Poirier P, Martin J, et al. Impact of ex-
ercise training on muscle function and ergoreflex in
Fontan patients: a pilot study. Int J Cardiol 2006;
107(1):85–94.
[142] Georgiadou P, Babu-Narayan SV, Francis DP,
et al. Periodic breathing as a feature of right heart
failure in congenital heart disease. Heart 2004;
90(9):1075–6.
[143] Mott AR, Feltes TF, McKenzie ED, et al. Im-
proved early results with the Fontan operation in
adults with functional single ventricle. Ann Thorac
Surg 2004;77(4):1334–40.
659
[144] Borowski A, Ghodsizad A, Litmathe J, et al. Severe
pulmonary regurgitation late after total repair of
tetralogy of Fallot: surgical considerations. Pediatr
Cardiol 2004;25(5):466–71.
[145] Brown JW, Ruzmetov M, Vijay P, et al. Surgical re-
pair of congenital supravalvular aortic stenosis in
children. Eur J Cardiothorac Surg 2002;21(1):50–6.
[146] Masetti P, Ussia GP, Gazzolo D, et al. Aortic pul-
monary autograft implant: medium-term follow-up
with a note on a new right ventricular pulmonary
artery conduit. J Card Surg 1998;13(3):173–6.
[147] Brochu MC, Baril JF, Dore A, et al. Improvement
in exercise capacity in asymptomatic and mildly
symptomatic adults after atrial septal defect percu-
taneous closure. Circulation 2002;106(14):1821–6.
[148] Bristow MR, Saxon LA, Boehmer J, et al. Cardiac-
resynchronization therapy with or without an im-
plantable defibrillator in advanced chronic heart
failure. N Engl J Med 2004;350:2140–50.
[149] Young JB, Abraham WT, Smith AL, et al. Com-
bined cardiac resynchronization and implantable
cardioversion defibrillation in advanced chronic
heart failure: the MIRACLE ICD trial. JAMA
2003;289:2685–94.
[150] Cazeau S, Leclercq C, Lavergne T, et al. Effects of
multi-site biventricular pacing in patients with
heart failure and intraventricular conduction delay.
N Engl J Med 2001;344:873–80.
[151] Janousek J, Vojtovic P, Hucin B, et al. Resynchro-
nization pacing is a useful adjunct to the manage-
ment of acute heart failure after surgery for
congenital heart defects. Am J Cardiol 2001;88:
145–52.
[152] Zimmerman FJ, Starr JP, Koenig PR, et al. Acute
hemodynamic benefit of multisite ventricular pac-
ing after congenital heart surgery. Ann Thorac
Surg 2003;75:1775–80.
[153] Dubin AM, Feinstein JA, Reddy VM, et al. Electri-
cal resynchronization: a novel therapy for the fail-
ing right ventricle. Circulation 2003;107:2287–9.
[154] Dubin AM, Janousek J, Rhee E, et al. Resynchro-
nization therapy in pediatric and congenital heart
disease patients: an international multicenter study.
J Am Coll Cardiol 2005;46(12):2277–83.
[155] Vonder Muhll I, Liu P, Webb G. Applying stan-
dard therapies to new targets: the use of ACE inhib-
itors and B-blockers for heart failure in adults with
congenital heart disease. Int J Cardiol 2004;
97(Suppl 1):25–33.
[156] Alehan D, Ozkutlu S. Beneficial effects of 1-year
captopril therapy in children with chronic aortic re-
gurgitation who have no symptoms. Am Heart J
1998;135:598–603.
[157] Mori Y, Nakazawa M, Tomimatsu H, et al. Long-
term effect of angiotensin-converting enzyme inhib-
itor in volume overloaded heart during growth:
a controlled pilot study. J Am Coll Cardiol 2000;
36:270–5.
660 DIMOPOULOS
[158] Kouatli AA, Garcia JA, Zellers TM, et al. Enalapril
does not enhance exercise capacity in patients after
Fontan procedure. Circulation 1997;96:1507–12.
[159] Heragu N, Mahony L. Is captopril useful in
decreasing pleural drainage in children after mod-
ified Fontan operation? Am J Cardiol 1999;84:
1109–12.
[160] Hechter SJ, Fredriksen PM, Liu P, et al. Angio-
tensin-converting enzyme inhibitors in adults
after the Mustard procedure. Am J Cardiol
2001;87:660–3.
[161] Buchhorn R, Hulpke-Wette M, Hilgers R, et al.
Propranolol treatment of congestive heart failure
in infants with congenital heart disease: The CHF-
PRO-INFANT Trial. Int J Cardiol 2001;79:167–73.
[162] Buchhorn R, Bartmus D, Siekmeyer W, et al. Beta-
blocker therapy of severe congestive heart failure in
infants with left to right shunting. Am J Cardiol
1998;81:1366–8.
[163] Buchhorn R, Ross RD, Hulpke-Wette M, et al. Ef-
fectiveness of low dose captopril versus proprano-
lol therapy in infants with severe congestive
failure due to left-to-right shunts. Int J Cardiol
2000;76:227–33.
[164] Buchhorn R, Hulpke-Wette M, Ruschewski W,
et al. Effects of therapeutic beta blockade on myo-
cardial function and cardiac remodeling in congen-
ital cardiac disease. Cardiol Young 2003;13(1):
36–43.
[165] Fernandes SM, Newburger JW, Lang P, et al. Use-
fulness of epoprostenol therapy in the severely ill
adolescent/adults with Eisenmenger physiology.
Am J Cardiol 2003;91:632–5.
[166] Galiè N, Ghofrani HA, Torbicki A, et al. Sildenafil
citrate therapy for pulmonary arterial hyperten-
sion. N Engl J Med 2005;353:2148–57.
[167] Mehta P, Simpson L, Lee E, et al. Endothelin-re-
ceptor antagonists improve exercise tolerance and
oxygen saturations in adults with the Eisenmenger
syndrome [abstract]. Circulation 2005;112:II-680.
et al
[168] Apostolopoulou SC, Manginas A, Cokkinos DV,
et al. Long-term clinical, exercise and hemody-
namic effect of the endothelin antagonist bosentan
in patients with pulmonary arterial hypertension
related to congenital heart disease [abstract].
Circulation 2005;112:II-564.
[169] Gatzoulis MA, Rogers P, Li W, et al. Safety and
tolerability of bosentan in adults with Eisenmenger
physiology. Int J Cardiol 2005;98:147–51.
[170] Schulze-Neick I, Gilbert N, Ewert R, et al. Adult
patients with congenital heart disease and pulmo-
nary arterial hypertension: first open prospective
multi-center study on bosentan therapy. Am Heart
J 2005;150:716.
[171] Galie N, Beghetti M, Gatzoulis M, et al.
BREATHE-5: Bosentan improves hemodynamics
and exercise capacity in the first randomized pla-
cebo controlled trial in Eisenmenger physiology
[abstract]. Chest 2005;128:496S.
[172] Piepoli MF, Davos C, Francis DP, et al. ExTra-
MATCH Collaborative. Exercise training meta-
analysis of trials in patients with chronic heart
failure (ExTraMATCH). BMJ 2004;328(7433):
189.
[173] Therrien J, Fredriksen P, Walker M, et al. A pilot
study of exercise training in adult patients with
repaired tetralogy of Fallot. Can J Cardiol 2003;
19:685–9.
[174] Thaulow E, Fredriksen PM. Exercise and training
in adults with congenital heart disease. Int J Cardiol
2004;97(Suppl 1):35–8.
[175] Graham TP Jr, Driscoll DJ, Gersony WM, et al.
Task Force 2: congenital heart disease. J Am Coll
Cardiol 2005;45(8):1326–33.
[176] Deanfield J, Thaulow E, Warnes C, et al. Manage-
ment of grown up congenital heart disease. Eur
Heart J 2003;24(11):1035–84.
[177] Bolger AP, Coats AJ, Gatzoulis MA. Congenital
heart disease: the original heart failure syndrome.
Eur Heart J 2003;24(10):970–6.