Professional Documents
Culture Documents
Introduction
Reperfusion-induced ventricular arrhythmias (VAs) have interested researchers
since their first description in 1935.[1] Preclinical models of reperfusion arrhythmias
explored patho- and electrophysiological mechanisms and assessed anti-
arrhythmic drug efficacy,[2] generating concern about possible initiation of
malignant arrhythmias with intracoronary thrombolytic therapy in patients with ST-
elevation myocardial infarction (STEMI).[3] Clinical experience, however, proved
reassuring.[3,4]Widespread use of intravenous thrombolytics saw reperfusion VAs
[particularly accelerated idioventricular rhythms (AIVRs)] regarded optimistically
as a non-invasive electric biomarker of successful epicardial reperfusion.[5] More
recently, it has been suggested that the presence of reperfusion-induced VAs
following primary percutaneous coronary intervention (PCI) reflects an adverse
cellular response to epicardial recanalization,[6] although this is controversial.[7,8]
Study Population
The data used for this retrospective analysis were obtained from subjects
participating in the CAldaret ST-Elevation Myocardial Infarction (CASTEMI) study,
which has been previously reported.[15] Briefly, CASTEMI was a multicentre,
randomized, double-blind trial that evaluated adjunctive MCC-135 (caldaret) in
subjects undergoing primary PCI for acute STEMI, defined as >20 min of chest
pain with onset <6 h from presentation, with >10 mm ST-segment elevation
summed over the presenting 12-lead ECG. No drug effect was detected; thus, for
this analysis, data for all CASTEMI patients with anterior MI and final TIMI grade
3 flow after primary PCI were pooled. Data were collected by an independent
clinical research organization and subjected to quality control and validation
procedures.
(Enlarge Image)
Figure 1.
Three-dimensional graphic output from 12-lead digital ECG monitor (NEMON 180+, NorthEast
Monitoring, Inc.), showing ST-segment level (mm) (y-axis) for all 12 leads (z-axis) with respect to
time (x-axis) in a study subject with anterior STEMI treated with primary PCI presented as abrupt
ST-segment resolution anterior myocardial wall leads V2-V6; V4 is the peak ST lead. The
substantial data gap in the graphic at the time of ST-elevation resolution was caused by a sudden
increase ('bursts') of VAs from which ST-segment levels were excluded for graphic three-
dimensional continuous ST-segment recovery visualization.
Statistical Methods
Outlier Detection Methodology. This method uses all VA counts obtained over
the course of the Holter recording for a patient and automatically separates
outliers (VA burst counts), if any, from background VA counts. Background counts
are considered to arise from a Poisson distribution with mean . A VA count is
considered an outlier when it is statistically significantly large enough that it likely
does not arise from background VA activity. To establish which VA counts are
significantly large, we considered type I error =0.01 with a conservative
Bonferroni adjustment for multiplicity of comparisons by dividing by 288 [the
number of 5 min intervals in a 24 h Holter recording (288=24*60/5)]. Thus, a VA
count above cutoff C() equal to the (1 - /288) quantile of the background count
distribution with mean is considered an outlier. This method is based on the
forward search and is an iterative process, because it is unknown a priori which
VA counts form the background VA activity.[22] To form the initial background
collection of VA counts, we chose a low mean =1 corresponding to a cutoff
equal to seven VAs/5 min interval. Ventricular arrhythmia counts below this initial
cutoff level were considered as initial 'background,' which was required to contain
at least 10% of the recorded VA counts; rarely, few next-smallest VA counts >7
were added to the initial 'background' collection to satisfy this requirement. Based
on the current set of background counts, a new background average was
computed and the next (higher) cutoff point C() was determined as before,
based on this new higher mean . If any VA count currently outside the
background collection was below the new cutoff level, the smallest such count
was added to the existing background collection and the next average with the
next cutoff C() was computed. The iteration was repeated until the background
could not be enlarged, either because all recorded VA counts were already in the
background (i.e. no outliers, and thus, no bursts detected), or VA counts not
currently in the background were all above the current cutoff C() and thus were
identified as outliers forming the initial VA bursts. For patients with initial bursts, a
statistical smoothing of counts over the duration of Holter recording was
performed, and single-burst counts or groups of time-contiguous burst counts with
maximum smoothed value below the last cutoff C() were eliminated. The
resulting collection of burst counts, if any remained, formed the subject's final VA
bursts.
(Enlarge Image)
Figure 2.
Concomitantly acquired (A) coronary angiography assessments of pre- and post-primary PCI
TIMI flow grades in a study subject with a total occlusion in the proximal left anterior descending
(LAD) artery; (B) continuous digital 12-lead ECG monitoring for ST-segment recovery analysis;
and (C) complete beat-to-beat Holter monitoring for quantitative rhythm analysis to identify VA
bursts in excess of established patient-specific background rates of VAs using statistical outlier
detection methodology. Cx, circumflex artery.
(Enlarge Image)
Figure 3.
Distribution of time interval between onset of reperfusion to first VA burst in all subjects with VA
bursts.
Characteristics of Reperfusion Ventricular Arrhythmia Bursts
The characteristics of VA bursts during or subsequent to reperfusion (reperfusion
VA bursts) are summarized in Table 1 . The median time from onset of
reperfusion to first VA burst was 4 min (IQR: 0-43.3). VA bursts comprised a
median total of 1290 VPCs (with a median of 97 VPCs per 5 min interval) and
persisted for a total median duration of 105 min. Runs of three or more VPCs
appeared to be more common in VA bursts than were isolated VPCs or couplets;
in background VAs, however, isolated VPCs were more common (Figure 4).
(Enlarge Image)
Figure 4.
Contribution of isolated VPCs (yellow bars), VPC couplets (blue bars), or runs of three or more
VPCs (red bars) to all background VAs (lower grey bars) or VA bursts (upper grey bars).
Figure 5.
(A) Temporal correlation between concomitantly acquired anatomic onset of reperfusion (defined
as angiographic documentation of re-establishment of blood flow in the IRA), and physiological
onset of reperfusion (defined as time of transition from peak ST level immediately preceding
50% stable ST-segment recovery). (B) Delay from time of onset of anatomic reperfusion,
defined as angiographic documentation of re-establishment of flow in the IRA, to onset of
physiological reperfusion, defined as time of transition from peak ST level immediately preceding
50% stable ST-segment recovery in subjects with reperfusion VA bursts vs. subjects without
bursts.
Discussion
Our analysis shows that with simultaneous angiographic TIMI flow grading,
continuous high-fidelity 12-lead ST-segment recovery, and beat-to-beat Holter
rhythm analysis, reperfusion-induced VAs can be defined more precisely in
subjects with anterior STEMI. Our data suggest that VA bursts in excess of
subject-specific background VA rates are temporally associated with angiographic
TIMI 3 flow restoration (anatomic reperfusion) and stable ST-segment recovery
(physiological reperfusion). We therefore consider reperfusion-induced VAs in
human subjects to be defined by the presence of these reperfusion VA bursts.
This definition provides a uniquely quantitative characterization of VAs specifically
related to both epicardial recanalization and myocardial reperfusion. Unlike
previous definitions, reperfusion VA bursts are statistically defined as outliers
relative to a quantified, patient-specific background VA density and can thus
include any of the many ventricular rhythms, from isolated VPCs to sustained fast
VTs, that may comprise reperfusion arrhythmias. This definition may also be
especially relevant to modern primary PCI therapy for STEMI, as both
angiographic and ECG monitoring signals are technologies already routinely used
in managing these patients.
From the earliest days of thrombolytics to the modern era of primary PCI for
STEMI, reperfusion arrhythmias have sustained a wide range of mechanistic and
prognostic interests.[3,5,6,8-14,23-29] A major barrier to advancing understanding in this
area, or even to achieving a meaningful synthesis of available reports, is the
present heterogeneity of definitions used for reperfusion as well as for reporting
and timing of VAs, or the correlation of the two. Observational reports of
arrhythmias in large cohorts receiving thrombolytics almost certainly include
arrhythmias from patients who never reperfuse, as recanalization rates with
thrombolytics exceeding 75% of the patients treated have never been reported. [30-
35] In addition, sporadic capture of arrhythmias in some reports is likely subject to
Our Holter data show that virtually all subjects experienced VAs in the course of
infarction; however, this was not necessarily related to reperfusion per se. This is
consistent with the report of Miller et al.,[13]who analysed Holter data in patients
with simultaneous angiographic observations associated with delivery of
intracoronary thrombolytics, and it was found that quantification of VAs, both
overall and by specific category, could not distinguish patients who reperfused
from those who did not. Ventricular arrhythmia bursts may provide a conceptual
framework for identifying unique events useful for specifically characterizing
reperfusion-related electrophysiological signals above the background activity
level.
Although our objective in performing this analysis of subjects with anterior STEMI
from the CASTEMI cohort was not to establish mechanistic or clinical outcome
correlations to reperfusion VA bursts, nonetheless, the observed correlation of
such bursts with higher ST peak deviation and with the 'right-shift' of the time
delay from anatomic to physiological reperfusion peak is noteworthy. Worsening
of ST-segment elevation following angiographic reperfusion has been previously
reported as a marker of adverse cellular response to reperfusion. [24,25,42] The
greater tendency for patients with reperfusion VA bursts to show delayed peaking
of ST injury current following angiographic reperfusion compared with patients
without such bursts suggests a mechanistic effect consonant with the findings of
Engelen et al.; i.e. that reperfusion followed by additional ST-segment elevation
and reperfusion VA bursts may indicate patients with cytotoxic reperfusion
responses who are less likely to recover mechanical function of the infarct zone.
Limitations
Our study has a number of limitations, chief among which is the fact that our
analysis is intended to provide observational details and a novel definition for
reperfusion VA bursts. As such, our findings, although suggestive, cannot fully
elucidate the mechanistic or prognostic significance of the observed
phenomenon. Further, our study is retrospective, albeit prospectively planned
across blinded core laboratory and data coordinating centre angiographic, ST-
segment recovery, Holter arrhythmia, and clinical descriptor datasets. Finally, our
findings are drawn from a modest cohort of subjects with anterior STEMI
undergoing primary PCI. Important subgroup analyses are limited by cohort size,
logistical consideration, and by the limitations inherent in the data collected for the
study that formed the basis of our retrospective analysis; thus, the generalizability
of our observations to non-anterior MIs or to patients treated with thrombolytic
therapy and other medication must be regarded with caution.
Mechanistic questions remain as well. Although heart rate variability data were
not available from the eECG Core laboratory in these studies, the relationship
between heart rate behaviour and reperfusion VA bursts is interesting for future
research. Even with continuous ECG monitoring and early angiography
associated with primary PCI in the CASTEMI study, it is conceivable that some
patients had transient or 'cyclic' changes in IRA patency prior to enrollment and
monitoring. Whether or not such events might be protective, or might
'precondition' the infarct zone substrate, cannot be determined from our
data.Continue Reading
Conclusions
In the current era of primary PCI for STEMI characterized by final TIMI 3 flow
success rates of >80%, short- and long-term implications of reperfusion VAs
cannot be imputed from existing literature, primarily due to limitations in existing
definitions. By combining concomitant ECG monitoring and angiographic
evidence of epicardial recanalization in conjunction with a patient-specific
statistical definition of VA bursts outside of background ventricular ectopy rates,
we have developed a novel definition of reperfusion arrhythmias that is more
quantitative, comprehensive, and specific than any hitherto described. Further
testing in larger patient cohorts is necessary in order to assess the degree to
which this definition enables further insight into the mechanistic and prognostic
implications of VAs in the modern era of STEMI therapy.
Turkey
Received 17 December 2012; Accepted 16 January 2013
Academic Editors: C. Spaulding and X. L. Tang
Copyright 2013 Ersan Tatli et al. This is an open access article distributed under
the Creative Commons Attribution License, which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is
properly cited.
Abstract
Objective. The most important step in the treatment of ST elevation myocardial
infarction is to sustain myocardial blood supply as soon as possible. The two main
treatment methods used today to provide myocardial reperfusion are
thrombolytic therapy and percutaneous coronary intervention. In our study,
reperfusion arrhythmias were investigated as if they are indicators of coronary
artery patency or ongoing ischemia after revascularization. Methods. 151 patients
with a diagnosis of acute ST elevation myocardial infarction were investigated. 54
patients underwent primary percutaneous coronary intervention and 97 patients
were treated with thrombolytic therapy. The frequency of reperfusion arrythmias
following revascularization procedures in the first 48 hours after admission was
examined. The relation between reperfusion arrhythmias, ST segment regression,
coronary artery patency, and infarct related artery documented by angiography
were analyzed. Results. There was no statistically significant difference between
the two groups in the frequency of reperfusion arrhythmias (). Although
angiographic vessel patency was higher in patients undergoing percutaneous
coronary intervention, there was no significant difference between the patency
rates of each group with and without reperfusion arrythmias. Conclusion. Our
study suggests that recorded arrhythmias following different revascularization
procedures in acute ST elevation myocardial infarction may not always indicate
vessel patency and reperfusion. Ongoing vascular occlusion and ischemia may lead
to various arrhythmias which may not be distinguished from reperfusion
arrhythmias.
1. Introduction
The term, reperfusion arrhythmias, was used in the first studies of thrombolytic
therapy guided revascularization in acute myocardial infarction. To confirm the
presence of vessel patency and successful reperfusion of the myocardiumfollowing
thrombolytic therapy, electrocardiographic data in addition to clinical and
laboratory measures are used by the clinicians. These include normalization or
more than 50% regression of ST elevation, T-wave inversion, and any other
arrythmia observed in electrocardiography. The most frequently observed
arrhytmias that are defined as reperfusion arrhytmias are ventricular premature
contractions, sustained or nonsustained episodes of ventricular tachycardia,
accelerated idioventricular rhythm, atrial fibrillation, and ventricular fibrillation.
These arrhythmias are thought to be indicators of successful reperfusion.
However, in some studies it was mentioned that these arrhytmias may be due to
ongoing myocardial cell damage and ischemia [1].
In acute myocardial infarction, ventricular tachycardia (VT) and ventricular
fibrillation (VF) may occur due to either complete occlusion or reperfusion.
Rhythm disorders associated with coronary occlusion are defined as ischemic
arrhytmias, whereas arrhythmias occurring as the result of increased myocardial
perfusion are called reperfusion arrhythmias [2]. Accelerated idioventricular
rhythm may be a marker of early reperfusion and continuing arterial patency. In a
previous study, the presence of accelerated idioventricular rhytm combined with
normalization of ST segments was demonstrated to indicate successful reperfusion
in patients treated with thrombolytics and there was no requirement for
emergency coronary angiography and rescue percutaneous coronary intervention
(PCI) in this group of patients [3].
The aim of our study is to observe the arrhythmias after application of different
revascularization treatment modalities (thrombolysis or primary) and investigate
whether these are indicators of arterial patency or ischemia due to ongoing
vascular occlusion by reviewing the arrhythmias recorded in the first 48 hours. In
addition, reperfusion arrhythmias observed after thrombolysis and percutaneous
revascularization were compared in this study.
3. Results
151 patients with STEMI were included in the study. To 54 (35.8%) cases primary
PCI was performed, and 97 (64.2%) cases were given thrombolytic drug therapy.
The age of patients ranged between 26 and 88 years. Mean age was years. 118
patients (78.1%) were male and 33 (21.9%) were female (Table 1). There was not
any statistically significant difference between primary PCI and thrombolytic
treatment groups in terms of age, height, weight, and body mass index. And also
mean levels of demographic data were not significantly different between the men
and women (). There was no statistically significant difference between the two
groups in hypertension, diabetes mellitus, smoking, history of coronary artery
disease, preinfarction angina, and previous history of MI (). Systolic blood
pressures, inhospital cardiovascular events, ejection fractions, onset of chest pain,
and time to implementation of treatment (pain-reperfusion period) did not show
a significant difference between the two groups (). The time from onset of pain to
treatment time in the primary PCI and thrombolytic groups was,
respectively, hours and hours (). Tables 1 and 2 show demographic and clinical
characteristics of both groups.
4. Discussion
In our study, the incidence of reperfusion arrhythmias detected in the first 48
hours did not differ significantly between the two treatment groups. 83.3% of
patients in primary PCI group and 88.7% of the thrombolytic group, had at least
one reperfusion arrhythmia. When the rates of different arrhythmia findings were
examined separately, the ratio of sustained VT, nonsustained VT, VF, frequently
VEA, and AV block was similar between the two groups. However AIVR and AF
ratio were higher in the thrombolytic group.
The incidence of the development of AF in acute MI is about 510% and it is known
that AF development in acute MI is usually due to impaired left ventricular function
or poor reperfusion [5]. Celik et al. reported that in patients who underwent PTCA,
p dispersion was reduced, showing that successful reperfusion may reduce the
likelihood of development of AF [6]. On the contrary, the incidence of AF could
increase in patients with poor reperfusion. The left ventricular functions in both
groups of patients were normal in this study. Therefore, the cause of AF was
thought to be residuel ischemia or poor reperfusion instead of impaired left
ventricular functions. Unlike thrombolytic group, we observed significantly lower
rates of AF in PTCA group. Six et al. investigated the predictive value of ventricular
arryhtmias as an indicator of angiographic arterial patency after thrombolytic
therapy. Among these, AIVR was the most sensitive and specific arrythmia in cases
of successful reperfusion [7]. In contrast, Bonnemeier et al. demonstrated that only
19 of 125 patients who were successfully treated with primary PCI exhibited AIVR
which indicated a poor relationship between TIMI 2 or 3 flow and reperfusion
arrhythmia and declared that AIVR may not be used as a reperfusion criteria [8].
Terkelsen et al. mentioned that although AIVR does not meet criteria for
reperfusion, it may be an indicator of more extensive myocardial damage and
delayed microvascular reperfusion in a study of 503 patients who were treated by
primary PTCA [9]. In another study, Gibson et al. [10] exhibited the development
of VT and VF in 3491 patients with STEMI after thrombolytic therapy, and reported
that cases who developed VT or VF had TIMI 02 flow.
The relationship between arterial patency rates demonstrated via coronary
angiography (TIMI flow grades). The analysis of TIMI flows allows better
understanding of the relationship between arterial patency rates and reperfusion
arryhtmias.
In the primary PCI group, TIMI 3 patency was obtained in 83.3% of cases. There
was no statistically significant relation between the presence of reperfusion
arryhtmia and TIMI flow grades in primary PCI group although reperfusion
arryhtmia was observed with a ratio of 82.2% in these patients with TIMI 3 flow.
42.3% and 38.1% of the patients who had received thrombolytic therapy had TIMI
3 and TIMI 2 flow, respectively. The incidence of reperfusion arryhthmia observed
in the group of patients with TIMI 1 flow was significantly lower in thrombolytic
therapy arm. On the other hand, there was no relationship between the presence
of reperfusion arryhthmia and higher TIMI flow grades (TIMI 2 and TIMI 3) in the
same group.
Wehrens et al. [11] compared the electrocardiographic changes following
reperfusion in AMI patients to whom thrombolytic therapy, primary PTCA, or
rescue PTCA were performed. The researchers demonstrated that
electrocardiographic changes as a noninvasive tool associated with reperfusion
did not provide sufficient information to clinicians to distinguish TIMI 2 and 3
flows from each other after thrombolytic therapy. This result is compatible with
our results. Gressin et al., investigated the relationship between ventricular
arrhytmias and arterial patency by recording ventricular arrhytmias in the first 24
hours and examining the angiographic views in a group of patients with acute
myocardial infarction treated with thrombolytic therapy [12]. On the basis of
ventricular arrhythmias detected; arryhtmia rates, in patients with TIMI 2 and 3
flow grades and in those without arterial patency (TIMI 0 or 1 flow grades), was
found to be similar. These results suggest that some arrythmias may be due to
different reasons such as ongoing ischemia or metabolic abnormalities. However,
in another study, rates of sustained VT or AIVR occurring in the first 6 hours were
found significantly higher in patients with arterial patency and these arrhytmias
were defined as noninvasive indicators of early coronary reperfusion [13]. But it
must be emphasized that the results and assessment of this study were limited to
the detection of arrythmias in the first 6 hours, which is different from previous
studies.
Engelen et al. demonstrated that at least 97% of patients developed one ventricular
reperfusion arrhythmia in a study group of 62 patients with acute anterior
myocardial infarction who underwent primary PCI. Nonsustained VT, AIVR, and
frequent VEA were the most frequently arrhythmias (similar to our findings) and
ventricular arrhytmias were correlated with reduced and worsening left
ventricular functions after the acute phase of myocardial infarction among this
group of patients treated via primary PCI and concluded that reperfusion
arrhythmias are noninvasive indicators of myocardial cell damage [14].
4.1. Study Limitations
The main limitation of the present study resides in its retrospective design. Larger
scale studies may provide additional information for clinical applications.
5. Conclusion
In conclusion, reperfusion arrhytmia frequency is not significantly different
between patients treated by primary PCI and thrombolytics although PCI is known
to provide better arterial patency. AIVR, which is so far known as a reperfusion
arrhytmia, may indicate myocardial damage and ongoing ischemia rather than
vascular patency. Further larger prospective studies are required to validate this
hypothesis.
References
1. G. Heper, M. E. Korkmaz, and A. Kilic, Reperfusion arrhythmias: are they only a
marker of epicardial reperfusion or continuing myocardial ischemia after acute
myocardial infarction? Angiology, vol. 58, no. 6, pp. 663670, 2008. View at
Publisher View at Google Scholar View at Scopus
2. G. Francis and J. Alpert, Modern Coronary Care, Little, Brown and Company, Boston,
Mass, USA, 1995.
3. J. A. Chiladakis, N. Vlachos, N. Patsouras, A. Mazarakis, and A. S. Manolis,
Usefulness of reperfusion ventricular arrhythmias in non-invasive prediction of
early reperfusion and sustained coronary artery patency in acute myocardial
infarction, Journal of Thrombosis and Thrombolysis, vol. 12, no. 3, pp. 231236,
2001. View at Publisher View at Google Scholar View at Scopus
4. V. Aytekin, Update on ACC/ESC criteria for acute ST-elevation myocardial
infarction, Anadolu Kardiyoloji Dergisi, vol. 7, supplement 1, pp. 1415,
2007. View at Google Scholar View at Scopus
5. The thrombolysis in myocardial infarction (TIMI) trial. Phase I findings, The New
England Journal of Medicine, vol. 312, no. 14, pp. 932936, 1985. View at Scopus
6. T. Celik, A. Iyisoy, H. Kursaklioglu et al., Effects of primary percutaneous coronary
intervention on P wave dispersion, Annals of Noninvasive Electrocardiology, vol.
10, no. 3, pp. 342347, 2005. View at Publisher View at Google Scholar View at
Scopus
7. A. J. Six, J. H. Louwerenburg, J. H. Kingma, E. O. Robles de Medina, and N. M. Van
Hemel, Predictive value of ventricular arrhythmias for patency of the infarct-
related coronary artery after thrombolytic therapy, British Heart Journal, vol. 66,
no. 2, pp. 143146, 1991. View at Google Scholar View at Scopus
8. H. Bonnemeier, J. Ortak, U. K. H. Wiegand et al., Accelerated idioventricular rhythm
in the post-thrombolytic era: incidence, prognostic implications, and modulating
mechanisms after direct percutaneous coronary intervention, Annals of
Noninvasive Electrocardiology, vol. 10, no. 2, pp. 179187, 2005. View at
Publisher View at Google Scholar View at Scopus
9. C. J. Terkelsen, J. T. Srensen, A. K. Kaltoft et al., Prevalence and significance of
accelerated idioventricular rhythm in patients with ST-elevation myocardial
infarction treated with primary percutaneous coronary intervention, American
Journal of Cardiology, vol. 104, no. 12, pp. 16411646, 2009. View at
Publisher View at Google Scholar View at Scopus
10. C. M. Gibson, Y. B. Pride, J. L. Buros et al., TIMI study group Association of impaired
thrombolysis in myocardial infarction myocardial perfusion grade with ventricular
tachycardia and ventricular fibrillation following fibrinolytic therapy for ST-
segment elevation myocardial infarction, Journal of the American College of
Cardiology, vol. 51, no. 5, pp. 546551, 2008. View at Publisher View at Google
Scholar View at Scopus
11. X. H. T. Wehrens, P. A. Doevendans, T. J. Ophuis, and H. J. Wellens, A comparison of
electrocardiographic changes during reperfusion of acute myocardial infarction by
thrombolysis or percutaneous transluminal coronary angioplasty, American
Heart Journal, vol. 139, no. 3, pp. 430436, 2000. View at Google Scholar View at
Scopus
12. V. Gressin, Y. Louvard, M. Pezzano, and H. Lardoux, Holter monitoring of
ventricular arrhythmias in the first 24 hours of myocardial infarction after
intravenous thrombolytic therapy, Archives des Maladies du Coeur et des
Vaisseaux, vol. 85, no. 12, pp. 17811788, 1992. View at Google Scholar View at
Scopus
13. J. A. Chiladakis, N. Vlachos, N. Patsouras, A. Mazarakis, and A. S. Manolis,
Usefulness of reperfusion ventricular arrhythmias in non-invasive prediction of
early reperfusion and sustained coronary artery patency in acute myocardial
infarction, Journal of Thrombosis and Thrombolysis, vol. 12, no. 3, pp. 231236,
2001. View at Publisher View at Google Scholar View at Scopus
14. D. J. Engelen, V. Gressin, M. W. Krucoff et al., Usefulness of frequent arrhythmias
after epicardial recanalization in anterior wall acute myocardial infarction as a
marker of cellular injury leading to poor recovery of left ventricular
function, American Journal of Cardiology, vol. 92, no. 10, pp. 11431149,
2003. View at Publisher View at Google Scholar View at Scopus
Last literature review version 19.3: Fri Sep 30 00:00:00 GMT 2011 | This topic last
updated: Wed Oct 05 00:00:00 GMT 2005 (More)
Sustained ventricular arrhythmias are less common in patients with an acute non-ST
elevation MI or unstable angina, as illustrated in a pooled analysis of four major trials of
over 25,000 such patients [2]. The overall incidence of VT or VF was 2.1 percent, lower
than the 10.2 percent incidence in GUSTO-1 in STEMI [3]. VT occurred in 0.8 percent, VF
in 1 percent, and VT and VF in 0.3 percent. The median time to arrhythmia was 78
hours.
Sustained VT and VF in the setting of myocardial infarction result from the complex
interaction of multiple factors, including myocardial ischemia, necrosis, reperfusion,
healing, and scar formation. These events produce the mechanisms that initiate
arrhythmias and the substrate for arrhythmia perpetuation. Arrhythmia pathogenesis
varies at different stages in this process. For ventricular arrhythmias occurring more than
48 to 72 hours after an acute myocardial infarction (MI), scar formation is of primary
importance.
Ventricular arrhythmias during acute MI are typically classified based upon their time of
onset. This scheme is useful clinically because VT and VF occurring early (24 to 48
hours) have been thought to be epiphenomena of the MI and do not require long-term
therapy because they are not associated with a worse prognosis after hospital discharge.
In contrast, VT or VF occurring later is generally thought to reflect the development of
arrhythmic substrate, and require chronic therapy because of an increased risk of sudden
cardiac death. (See "Clinical features and treatment of ventricular arrhythmias during
acute myocardial infarction".)
This topic will review the mechanisms of arrhythmogenesis during acute MI, with
particular attention to serious ventricular arrhythmias occurring in the first 48 to 72
hours. The clinical features and management of these arrhythmias are discussed
separately. (See "Clinical features and treatment of ventricular arrhythmias during acute
myocardial infarction".)
The proper milieu for the development and maintenance of ventricular arrhythmias is
ultimately based upon the rapid and profound effects that acute myocardial ischemia has
on the electrophysiologic characteristics of the myocyte. Changes in the resting
membrane potential and in the inward and outward ionic fluxes during the action
potential lead to alterations in conduction, refractoriness, and automaticity of cardiac
muscle cells, all of which contribute to the occurrence of ventricular arrhythmias [7].
The net increase in potassium leakage out of the cells leads to a rise in the local
extracellular potassium concentration from 4 to 15 meq/L within 15 minutes of coronary
occlusion [12]. Potassium accumulation occurs in two phases, since blood flow is absent
[11,13,14]:
The first phase is reversible with reperfusion, provided it occurs within 15 minutes
of occlusion.
The second phase is irreversible and represents permanent cellular damage.
Animal experiments have shown that the Na-K-ATPase pump is only mildly depressed in
the early stages of ischemia (approximately 10 minutes) [7]. It has therefore been
suggested that initial potassium loss is secondary to increased permeability of the cell
membrane to anions, such as inorganic phosphate and lactate, or to the development of
an intracellular acidosis [12,15].
Isolated normal myocardium and Purkinje fibers exposed to LPGs show electrical
changes that are similar to those seen in ischemic myocardium [16-18].
Small amounts of LPGs augment the effect of hypoxia and elevated potassium,
effects that are intensified by acidosis [19].
High concentrations of LPGs cause disruption of the cell membrane, contributing
to massive calcium entry and cell death [20,21].
The metabolic changes and acidosis that accompany occlusion and ischemia can be
altered by multiple brief occlusions, as may occur in the patient with unstable angina or
an acute myocardial infarction with ongoing episodes of ischemia. Studies using
repetitive occlusions lasting less than five minutes have demonstrated a reduced
incidence of VF; this phenomenon has been termed "ischemic preconditioning" [22].
(See "Clinical implications of ischemic preconditioning".)
However, longer periods of occlusion may not produce the same results. One animal
study, for example, evaluated the effect of repetitive 15 minute coronary occlusions
followed by 45 minute reperfusions on the incidence of occlusion and release arrhythmias
[23]. Repetitive occlusions were associated with progressively higher ST segment
elevation, more frequent ST segment alternans, and a greater incidence of VF,
suggesting increasing metabolic derangement. However, there was also faster
normalization of the ECG and a reduced incidence of ventricular fibrillation during
repetitive reperfusions, suggesting a progressively faster resolution of the metabolic
derangements.
Changes in the action potential Acute ischemia significantly reduces the amplitude,
upstroke velocity, and duration of the action potential. The changes vary based upon the
area of involvement.
The central ischemic zone has the largest alteration, consisting of a decrease in
impulse velocity and an increase in the refractory period.
The areas of normal myocardium surrounding the main area of infarction have an
increase in impulse conduction velocity (due to catecholamines) and a shortened
refractory period.
The ischemic tissue at the border zone of the infarct manifests electrophysiologic
changes that are intermediate between those found in the other two regions. This
inhomogeneity is partly caused by diffusion of potassium that has leaked out of
the cells in the ischemic areas toward more normal myocardium. In animals,
ischemia at the border zone of a one-month old infarction is more arrhythmogenic
that ischemia at a distance from the infarction zone [24].
Experimental studies suggest that phase 1a and 1b arrhythmias arise via different
mechanisms.
The possible origin for this series of events is a focal mechanism such as triggered
activity secondary to afterdepolarizations and abnormal automaticity. The latter changes
result from spontaneous impulse formation based upon phase 4 depolarization in cells
with reduced resting membrane potential [32]. The occurrence of a focal mechanism is
supported by a canine study using three dimensional activation mapping, which found
that 60 percent of ventricular tachycardias in the early ischemic period were of focal
Purkinje origin [33]. (See "Enhanced cardiac automaticity".)
Phase 1b arrhythmia The mechanism that gives rise to phase 1b arrhythmias has
not been fully elucidated, but they may reflect abnormal automaticity, induced by left
ventricular wall stress [34] or the increased catecholamine release that occurs after 15 to
20 minutes of ischemia [35-37].
These observations suggests that sustained VT within the first few hours after MI may be
the result of transient arrhythmogenic mechanisms, possibly resulting from ischemia, and
its predictive value for sudden death is low. Consistent with this hypothesis is a report in
which sustained monomorphic VT could not be induced in seven patients who had
sustained monomorphic VT within 24 hours of the acute insult [38].
VPBs
Nonsustained VT (three to four beats)
Accelerated idioventricular rhythm, which occurs in up to 50 percent of patients
It is generally believed that delayed arrhythmias arise from surviving Purkinje
cells located within the subendocardium that display abnormal automaticity [7,41].
These cells have an increased sensitivity to catecholamines, and delayed arrhythmias can
be provoked by stimulation of the sympathetic nervous system.
Chronic phase arrhythmias During the chronic, healing phase of an acute MI (more
than 72 hours), the typical patient with sustained monomorphic VT has had a large, often
complicated infarct with a left ventricular ejection fraction (LVEF) 30 percent [43,44].
The VT is often quite rapid in this setting, with a roughly equal distribution of heart rates
above 220 beats/min, between 180 to 220 beats/min, and below 180 beats/min [43].
(See "Sustained monomorphic ventricular tachycardia in patients with a prior myocardial
infarction: Treatment".)
New ischemia can be superimposed upon this substrate with the appearance of a variety
of ventricular arrhythmias including nonsustained, sustained monomorphic, and
polymorphic VT. Thus, revascularization and antiischemic therapy may be part of the
antiarrhythmic regimen. Autonomic imbalance, electrolyte abnormalities, and the
proarrhythmic effects of antiarrhythmic drugs can also contribute to the appearance of
sustained monomorphic VT or other ventricular arrhythmias.
In a series of 53 patients with spontaneous VT (44 with remote MI), the heart rate
rose and measures of heart rate variability indicated a rise in sympathetic activity
before the onset of the arrhythmia [45]. (See "Heart rate variability: Technical
aspects".)
One report evaluated the role of VPBs in 52 patients, 40 with CHD, who had an
implantable cardioverter-defibrillator (ICD) [46]. Two-thirds of VT episodes were
provoked by VPBs, which were most often multiple and had a morphology that
was different from the VT. In another study using detailed electroanatomic
mapping, 29 patients with ischemic cardiomyopathy and repeated episodes of VF
were evaluated [47]. In all subjects, VF was initiated by monomorphic VPBs
originating in the border zone of the infarct scar. (See "Clinical significance and
treatment of ventricular premature beats".)
Inducibility and reentry Sustained monomorphic VT can be initiated during
electrophysiology study in approximately 90 to 95 percent of these patients
[43,48,49]. The arrhythmia usually remains inducible after coronary artery bypass
graft surgery [49]. Patients who have noninducible sustained monomorphic VT
have a decreased probability of spontaneous VT, suggesting that the substrates
for inducible and spontaneous VT may evolve in parallel over time [50].
These findings support the hypothesis that late sustained monomorphic VT is typically
reentrant in origin. Experimental evidence in both the subacute and chronic phases after
MI suggest that scar with the interdigitation of viable myocardium may provide the
substrate for reentry [51,52]. Late potentials and slow conduction can occur in areas with
normal action potentials; thus, slow conduction may be due to disruption of the gap
junctional connections between cells by the collagen of the scar, resulting in
nonuniformity of anisotropic conduction [53].
Reentry may occur around the area of infarction with a single circuit or two circuits. In a
canine model of VT in healing MI, observed reentry circuits included [54]:
Figure of eight
Functional with long lines of functional block
Short lines of functional block with slow conduction transverse to fiber orientation
(anisotropic reentry)
Leading circle reentry
Support for the role of the sympathetic nervous system comes from a study of 49
patients who had electrical storm, defined as multiple recurrent episodes of VF,
associated with an acute MI [59]. Compared to patients receiving antiarrhythmic therapy,
those treated with sympathetic blockade (beta blockers or stellate ganglionic blockade)
had a lower one week mortality (22 versus 82 percent) and overall mortality at one year
(33 versus 95 percent).
The longer term important of the role of sympathetic activity comes from the clear
evidence of reduced mortality, including arrhythmic death, in patients treated with beta
blockers (figure 1) [60]. (See "Beta blockers in the management of acute coronary
syndrome".)
Indirect clinical evidence supporting a role for reduced parasympathetic tone are the
observations that decreased baroreflex sensitivity and impaired heart rate variability
(which reflect reduced vagal activity) are associated with increased inducibility of
sustained monomorphic VT and with sudden death during the first 6 to 12 months after a
myocardial infarction [61-63]. However, extrapolation of these findings to the acute
phase of MI is at present not justified. (See "Incidence of and risk stratification for
sudden cardiac death after acute myocardial infarction".)
Sympathetic efferent fibers en route to the left ventricle cross at the atrioventricular
groove, travel within the superficial subepicardium in a base-to-apex direction, and
penetrate the myocardium to innervate the endocardium. A transmural MI can damage
the sympathetic fibers and result in functional denervation of the myocardium apical to
the area of the infarction. The following observations illustrate the range of findings:
Animal and human studies have shown that, after an infarction, regions of the
ventricular myocardium below the infarcted zone have no uptake of iodine-123-
metaiodobenzylguanidine (MIBG), a radiolabeled guanethidine analog that is
actively taken up by sympathetic nerve terminals, even though this tissue remains
viable, as indicated by persistent thallium uptake [65-67].
Denervated regions show denervation hypersensitivity to infused norepinephrine
or isoproterenol, as manifested by a greater shortening in refractoriness
compared to normal myocardial tissue or easier inducibility of sustained VT during
sympathetic stimulation [68,69]. The resulting heterogeneity in refractoriness
induced by circulating catecholamines is a potential mechanism for
arrhythmogenicity after infarction that can be minimized by beta blockers [68]. In
comparison, regional denervation may not affect the inducibility of sustained VT at
electrophysiology study [67].
Reinnervation of the periinfarction area, but not the infarcted tissue, and the loss
of supersensitivity occur within 12 months after the infarction [70].
The profibrillatory effect of the thrombus in the above report was not detectable after
ischemic preconditioning [72]. This may account, in part, for the observation that a
sudden obstruction of the coronary artery without preexisting coronary artery narrowing
is associated with a higher incidence of ventricular fibrillation compared to an obstruction
occurring in association with preexisting stenoses. (See "Clinical implications of ischemic
preconditioning".)
The calcium paradox - The calcium paradox occurs when calcium is restored to
the extracellular space after having been depleted by ischemia. This ensuing
accumulation of calcium in the cell causes damage to the respiratory chain and
decreases ATP production, leading to depletion of mitochondrial energy [86].
Severe cell damage ensues with muscle contracture and cell death.
The oxygen paradox - The oxygen paradox is closely linked to the calcium
paradox since oxygen mediates the uptake of calcium by the mitochondria [87].
Recycling of calcium occurs when ATP is restored during reperfusion, with the
subsequent possibility of developing calcium-dependent arrhythmias [88,89].
Oxygen also leads to cell damage during reperfusion by the formation of oxygen
free radicals [90].
The production of oxygen free radicals - The role of oxygen free radicals in the
genesis of reperfusion arrhythmias is uncertain. Free radical-induced membrane
damage occurs 5 to 10 minutes after reperfusion [91], and reduced levels of free
radical scavengers have been observed within three hours of angioplasty in
patients with an infarction [92]. Reperfusion arrhythmias, however, usually occur
within seconds after abrupt restoration of blood flow. The observation that free
radicals reduce the occurrence of reperfusion ventricular fibrillation is also against
their playing a primary role in reperfusion arrhythmias [93].
Angiotensin II - A role for angiotensin II in reperfusion arrhythmias was suggested
by studies in angiotensin II type 1a receptor knockout mice which, compared to
wild-type mice, showed less reperfusion arrhythmia despite a similar infarct size.
Furthermore, administration with a selective angiotensin II type 1 receptor before
ischemia blocked reperfusion arrhythmias in the wild-type mice [94].
(See "Actions of angiotensin II on the heart".)
At present, there are no prophylactic therapeutic maneuvers that have been proven to
reduce the incidence of these events, but there is interest in the possible efficacy of
calcium channel blocking agents and free radical scavengers, such as superoxide
dismutase and adenosine. As an example, dipyridamole, an inhibitor of cellular uptake of
adenosine that may reduce intracellular calcium, was evaluated in a study of 61 patients
undergoing primary angioplasty for an acute anterior wall myocardial infarction [78].
Pretreatment with dipyridamole prevented all episodes of AIVR and VT, which occurred in
18 and 8 percent, respectively, of patients not receiving pretreatment. In addition,
dipyridamole terminated arrhythmia in all nine patients who developed a sustained
ventricular tachyarrhythmia after reperfusion
Clinical features and treatment of ventricular arrhythmias during acute
myocardial infarction
Authors
Philip J Podrid, MD
Leonard I Ganz, MD, FHRS, FACC
Section Editor
James Hoekstra, MD
Deputy Editor
Brian C Downey, MD, FACC
Disclosures
Last literature review version 19.3: Fri Sep 30 00:00:00 GMT 2011 | This topic last
updated: Mon Feb 07 00:00:00 GMT 2011 (More)
Early series underestimated the incidence of unstable VT and VF, because they did not
include patients whose MI resulted in out-of-hospital sudden cardiac death (SCD) [9].
The incidence of ventricular arrhythmias appears to be higher with ST elevation (Q wave)
than non-ST elevation (non-Q wave) infarctions (STEMI and NSTEMI).
VT is less common than VF in the periinfarction period. This is probably due to the fact
that VT usually requires a reentrant circuit and a fixed arrhythmic substrate, such as the
scar from a healed MI. When VT occurs during an acute MI, it may reflect triggered
activity, enhanced automaticity, or a prior infarction. (See "Pathogenesis of ventricular
tachycardia and ventricular fibrillation during acute myocardial infarction".)
Fibrinolytic therapy Among patients with acute MI in the fibrinolytic era, the
incidence of VF has ranged from 3.7 to 6.7 percent in large studies [12-17].
The largest experience in patients (40,895) with acute STEMI treated with fibrinolytic
therapy comes from the GUSTO-1 trial [13]. The overall incidence of sustained VT or VF
was 10.2 percent (3.5 percent developed VT, 4.1 percent VF, and 2.7 percent both VT
and VF). Approximately 80 to 85 percent of these arrhythmias occurred in the first 48
hours.
A similar frequency of VF was noted in a report from the GISSI-2 database of 9720
patients with a first MI treated with a fibrinolytic agent [12].
Two problems with fibrinolytic trials bring into question the validity of their reported
arrhythmia incidence. First, by nature of their patient inclusion and consent issues,
patients who died of SCD in the prehospital arena were excluded from the trial
denominator. Secondly, fibrinolysis and reperfusion may lead to reperfusion arrhythmias,
falsely increasing the incidence of MI associated arrhythmias. Given the reduced
incidence of VT and VF compared to older studies, it seems likely that reperfusion does
not necessarily increase the overall incidence of VF or sustained VT. Ventricular
premature beats (VPBs) and accelerated idioventricular rhythm (AIVR) have been
thought to be a marker of reperfusion following fibrinolysis [18]. (See "Pathogenesis of
ventricular tachycardia and ventricular fibrillation during acute myocardial infarction",
section on 'Reperfusion arrhythmias'.)
Primary PCI Among 5745 STEMI patients with planned percutaneous coronary
intervention (PCI) enrolled in the APEX AMI trial, VT or VF occurred in 329 (5.7 percent)
[19]. The arrhythmia was noted before catheterization in 25, during catheterization in
180, and after catheterization but before discharge in 170 patients. Most events (282)
occurred within the first 48 hours. (See "Primary percutaneous coronary intervention in
acute ST elevation myocardial infarction: Determinants of outcome".)
Among patients presenting with a first ST elevation MI, 330 survivors of primary
VF (ie, VF in the absence of heart failure or shock) were compared to 372 controls
[20]. A family history of SCD was more common among those who experienced
primary VF (odds ratio 2.7, 95% confidence interval 1.8 to 4.0, compared to
controls).
In a Finnish series of consecutive SCD victims who underwent autopsy (which is
mandatory for unexplained SCD in Finland), 138 were identified who had an acute
coronary event, but no history of prior MI [21]. These cases were compared to
two control groups; 254 consecutive patients surviving an acute MI, and 470
healthy subjects. A history of SCD in first-degree relatives was ascertained in each
group. The incidence of SCD was significantly higher in the relatives of SCD
victims than in those of the acute MI survivors or the control subjects (5.6, 3.3,
and 1.6 percent, respectively).
Although these observations are intriguing, this study design is subject to the possibility
of recall bias, which could lead to an overestimation of the familial risk.
In contrast, other studies have found that an increased frequency of VPBs alone may be
associated with an increased mortality risk [26-28]. As an example, the European
Infarction Study Group noted a 4 percent two-year mortality in post-MI patients with
more than 10 VPBs per hour, a value significantly higher than in that seen in those with
no or fewer VPBs [27]. The two-year mortality was much greater in patients with
ventricular couplets (22 and 10 percent if more than ten and one to nine couplets per
day, respectively).
There are also conflicting data as to whether the effect of VPBs upon sudden death in
post-MI patients is independent of left ventricular dysfunction. It is possible, for example,
that VPBs may only signify the presence of a severe myocardial process and may not be
an independent provoker of a terminal electrical event [29]. However, although a
relationship exists between the frequency of ventricular arrhythmias and the degree of
left ventricular dysfunction, the results of several large well-designed studies have found
that an increased frequency of complex VPBs is an important risk factor independent of
the degree of myocardial damage [24,30,31].
A review from the GISSI-2 trial evaluated 8676 patients with STEMI treated with
thrombolytic therapy undergoing 24-hour Holter monitoring before discharge: 64 percent
had at least one VPB per hour and 20 percent had more than 10 VPBs per hour [32]. At
six months, the mortality rate was 5.5 percent in patients with more than 10 VPBs per
hour compared to 2.0 and 2.7 percent in patients without ventricular arrhythmia or with
1 to 10 VPBs per hour, respectively. After adjusting for risk factors, more than 10 VPBs
per hour remained a significant predictor for both total (relative risk 1.62, 95% CI 1.16-
2.26) and sudden mortality (RR 2.24, 95% CI 1.22-4.08).
The findings were different in the Canadian Assessment of Myocardial Infarction (CAMI)
study, which included 3178 patients with either an STEMI or a non-ST elevation
myocardial infarction (NSTEMI) [35]. Although cardiac mortality increased progressively
with increasing frequency of VPBs, VPBs alone had no independent predictive value in
multivariate analysis.
Primary PCI era Very limited data have been published assessing VPBs following
primary percutaneous coronary intervention (PCI). One single-center study of 44
consecutive STEMI patients treated with primary PCI measured the frequency of VPBs in
the first 24 hours post-PCI [36]. Compared to historical controls treated with
thrombolysis, no significant difference was seen in the frequency of VPBs in the initial 24
hours post-PCI, although there was a trend toward fewer VPBs in the patients treated
with PCI.
Treatment of VPBs Suppression of VPBs with antiarrhythmic drugs does not improve
either short- or long-term outcomes, and, with some drugs, may actually increase
mortality [37,38]. Thus, treatment for the suppression of asymptomatic VPBs
is not recommended [39]. (See "Prophylaxis against ventricular arrhythmias following
myocardial infarction".)
AIVR occurs in up to 50 percent of patients with acute MI. Some studies have suggested
an association with reperfusion following fibrinolytic therapy [18]. However, AIVR is
neither a sensitive nor very specific marker for successful reperfusion.
(See "Pathogenesis of ventricular tachycardia and ventricular fibrillation during acute
myocardial infarction", section on 'Reperfusion arrhythmias'.)
Sustained monomorphic VT (SMVT) in the periinfarction period (ie, within the first 48
hours after the infarct) occurs in approximately 2 to 3 percent of patients with an ST
elevation MI (STEMI) [13,43] and less than 1 percent with a non-STEMI or unstable
angina [14]. SMVT is associated with larger MI size [43]. The factors responsible for
SMVT differ in the very early (30 minutes) and later phases (6 to 48 hours) of the early
post-MI period. (See "Pathogenesis of ventricular tachycardia and ventricular fibrillation
during acute myocardial infarction".)
The clinical presentation may include palpitations, worsening ischemic symptoms due to
the elevated heart rate, and hemodynamic compromise or collapse. Patients with faster
tachycardias and worse left ventricular function are less likely to tolerate the arrhythmia.
In the early stage of an evolving infarction (ie, the first 6 to as many as 48 hours),
monomorphic VT may result from transient arrhythmogenic phenomena in ischemic and
infarcting tissue, such as abnormal automaticity, triggered activity, and reentrant circuits
created by heterogeneous conduction and repolarization.
SMVT in any other setting is considered a marker of permanent arrhythmic substrate and
an increased long-term risk of arrhythmia recurrence and SCD. Because of the
physiologic and mechanistic link between SMVT and permanent substrate, it is not clear
that SMVT at any point, even early after an MI, should be attributed to transient
phenomena. Furthermore, SMVT in the setting of an acute MI may reflect permanent
substrate from a prior infarction. As a result, the prognostic significance of SMVT in the
early period after an MI is unclear. This issue is discussed in greater detail below.
(See 'Prognosis after early VT' below.)
Some patients (8 of 434 consecutive patients in one series) do develop typical pause-
dependent torsades de pointes in association with progressive QT interval prolongation
[45]. This pattern of events occurs during the healing phase of an infarct, usually days 3
to 11, and does not suggest recurrent ischemia. (See "Acquired long QT syndrome".)
Due to the relatively small numbers of patients and the combined analysis of patients
with VT and VF or nonsustained VT and sustained VT, these studies did not adequately
address the long-term prognostic significance of SMVT in the early post-MI setting. As
noted above, there are physiologic reasons to suggest that SMVT has different
implications than VF or nonsustained VT.
Somewhat different findings were noted in the analysis of over 40,000 patients in
GUSTO-I, illustrated by the following results [13] (figure 3):
The 2010 AHA guidelines give a weak recommendation regarding the use of
intravenous lidocaine for the treatment of recurrent sustained VT, stating it should only
be considered for use if amiodarone is not available [50]. However, lidocaine is a
relatively safe agent that has no adverse hemodynamic effects and a rapid onset of
action, and therefore may be useful in some patients. If lidocaine is used, it should be
administered by intravenous push in a dose of 0.5 to 0.75 mg/kg; this dose is repeated
every 5 to 10 minutes as needed. At the same time, a continuous intravenous lidocaine
infusion of 1 to 4 mg/min is begun. The maximum total dose is 3 mg/kg over one hour.
The majority of episodes of VF occur within the first 48 to 72 hours after the onset of
symptoms [12-14]. It is presumably a manifestation of ischemia and is associated with
lack of perfusion via the infarct-related artery [46,53].
Primary and nonprimary VF Ventricular fibrillation (VF) can be further subclassified
as primary or nonprimary:
Primary VF refers to VF that occurs early (usually <48 hours post myocardial
infarction), and is not associated with recurrent ischemia or heart failure. This
category is defined to identify patients who experience VF despite a relatively
uncomplicated myocardial infarction (ie, as a primary electrical event).
Nonprimary VF refers to all other episodes. VF is more common in patients with
myocardial infarctions that are complicated by heart failure or recurrent ischemia.
VF Predictors There are no clinical features that reliably identify the individual patient
who is likely to develop ventricular fibrillation (VF) in the periinfarction period. However,
retrospective analyses of several clinical trials and a large meta-analysis of both clinical
trials and observational series have identified some factors that are associated with an
increased risk of early VF [12,14,54,55]. These include:
In the APEX AMI trial of STEMI patients with planned PCI, multivariate analysis identified
the following significant predictors of early (prior to the end of cardiac catheterization) VF
or sustained VT [19]:
Pre-PCI Thrombolysis in MI (TIMI) flow grade 0 (HR 2.94, 95% CI 1.93 to 4.47)
Inferior infarction (HR 2.16, 95% CI 1.58 to 2.93)
Total baseline ST segment deviation (HR 1.39, 95% CI 1.19 to 1.63)
Killip Class (table 1) greater than I (HR 1.88, 95% CI 1.29 to 2.76).
In this same study, the following predictors of late (after the end of catheterization but
before discharge) VF or sustained VT were found [19]:
A report from the GISSI-2 trial evaluated the prognostic significance of ventricular
arrhythmias in 9720 patients with a first ST elevation MI (STEMI) who were treated with
fibrinolytic therapy [12]. Arrhythmias were characterized as primary VF (occurring in the
absence of heart failure or shock), nonprimary VF (all other VF), and VT. Primary VF was
further characterized as early (within 6 hours), or late (between 6 and 48 hours). This
distinction is in contrast to most other reports, in which the term "late VF" usually refers
to events that occur more than 48 to 72 hours after the MI. The following findings were
noted (figure 5):
Both early and late primary VF (ie, all events occurring within the first 48 hours)
were associated with a significant increase in in-hospital mortality.
The mortality rate from discharge to six months was not affected by primary VF.
Mortality from discharge to six months was markedly increased in those with
nonprimary VF.
Similar findings were noted in the GUSTO-I trial cited above, which included over 40,000
patients with STEMI who were treated with fibrinolytic therapy [13]. The 1229 patients
with early (first 48 hours) VF alone (ie, without VT) had an increased rate of in-hospital
mortality (19.8 versus 4.2 percent in those without sustained ventricular arrhythmia);
however, among those who survived 30 days, the one year mortality was the same as
that in patients without ventricular arrhythmias (2.7 percent in both groups) (figure 3).
Data are more limited in patients with a non-STEMI. In the pooled analysis noted above
of patients with non-ST elevation MI or unstable angina, VF was a significant predictor of
increased mortality at both 30 days and six months (adjusted hazard ratio 23 and 15,
respectively) [14]. The increase in risk was largely due to more deaths in the first 30
days.
Additional information on the prognosis of patients with early VF comes from the APEX
trial. (See 'Prognosis after early VT' above.)
Acute therapy Ventricular fibrillation (VF) is almost universally lethal if not treated,
with very few spontaneous reversions to sinus rhythm. Defibrillation is the definitive
therapy for VF. If available, a biphasic waveform defibrillator is preferable since the
success rate for defibrillation is higher than with monophasic waveforms. (See "Basic
principles and technique of cardioversion and defibrillation", section on 'Monophasic
versus biphasic waveforms'.)
The 2010 AHA guidelines for adult advanced cardiac life support (ACLS) recommended
that, for biphasic defibrillators, the initial shock should be at 120 to 200 joules, with
subsequent shocks at the highest available biphasic energy level (200 joules for most
devices) [50]. For monophasic defibrillators, nonescalating shocks beginning at 360
joules should be used. (See "Advanced cardiac life support (ACLS) in adults".)
Following successful reversion to sinus rhythm, patients should be treated with
intravenous amiodarone for 24 to 48 hours. The dosing schedule is similar to that used
for sustained VT (see 'Sustained ventricular tachycardia' above). At present, an ICD is
not generally appropriate for VF occurring in the first 24 to 48 hours of an acute MI.
The approach to VF that persists after three rapid shocks delivered in a row is discussed
in detail separately. (See "Supportive data for advanced cardiac life support in adults
with sudden cardiac arrest".)
The outcomes in patients who are resuscitated from a cardiac arrest in the setting of an
ST elevation MI who undergo emergent primary PCI were illustrated in a multicenter
series of 186 patients [59]:
The 2006 ACC/AHA/ESC guidelines for the management of ventricular arrhythmias made
the following recommendations [60]:
Late VT "Late" VT (ie, VT that occurs after the first 24 to 48 hours) is most often
related to healing of the infarct and may reflect the development of an arrhythmogenic
substrate that promotes the development of VT (eg, scar tissue). The associated fibrosis
leads to areas of conduction block with interdigitation of viable myocardium; the ensuing
slowing of conduction at the border of the infarct can lead to stable reentry circuits and
VT [61]. (See "Pathogenesis of ventricular tachycardia and ventricular fibrillation during
acute myocardial infarction", section on 'Chronic phase arrhythmias'.)
Late VT is a predictor of a worse prognosis [13,17,62]. The following findings were noted
in the patients with late ventricular arrhythmia in the GUSTO-I trial [13]:
The in-hospital mortality was significantly higher in patients who had late
sustained VT alone (37.5 percent) or VT and VF (57.0 percent) compared to those
without sustained ventricular arrhythmia (4.2 percent).
The one-year mortality of 30 day survivors was significantly higher among those
who had late VT alone (24.7 percent) compared to those without sustained
ventricular arrhythmia (2.7 percent).
A similar adverse effect was noted in the GISSI-3 trial in which late VT occurred in 1
percent of patients between 48 hours and six weeks [62]. Late VT was associated with a
more complicated in-hospital and six-week posthospital course, a higher incidence of
pump failure, atrial arrhythmias, asystole, atrioventricular block, and VF, and a higher
mortality at six weeks (35 versus 5 percent without VT, hazard ratio 6.13).
(See "Sustained monomorphic ventricular tachycardia in patients with a prior myocardial
infarction: Treatment".)
Late VF Patients who develop sustained VF more than 48 hours after their MI ar1e
also thought to be at risk for recurrent ventricular tachyarrhythmias and to have a worse
prognosis [13,17,62]. Such patients typically receive an ICD as well as revascularization
and optimal medical therapy [39]. (See "Implantable cardioverter-defibrillators: Clinical
trials of secondary prevention of sudden cardiac death".)
SUMMARY
REFERENCES
Authors
Richard C Becker, MD
Carey Kimmelstiel, MD
Section Editor
Freek Verheugt, MD, FACC, FESC
Deputy Editor
Gordon M Saperia, MD, FACC
Disclosures
Last literature review version 19.3: Fri Sep 30 00:00:00 GMT 2011 | This topic last
updated: Fri Oct 08 00:00:00 GMT 2010 (More)
Prompt and sustained coronary arterial patency has repeatedly been found to limit infarct
size. Among patients with an acute ST elevation (Q wave) MI, over 90 percent have
complete occlusion of the culprit artery, patency can be achieved by fibrinolysis, primary
percutaneous coronary intervention (PCI), or a combination of both modalities. In
addition, there are many factors (hemodynamic, anatomic, cellular) that influence
coronary artery patency regardless of the technique used (figure 1). (See "Characteristics
of fibrinolytic (thrombolytic) agents and clinical trials in acute ST elevation myocardial
infarction" and "Primary percutaneous coronary intervention in acute ST elevation
myocardial infarction: Determinants of outcome" and "Primary percutaneous coronary
intervention versus fibrinolysis in acute ST elevation myocardial infarction: Clinical
trials".)
TIMI FLOW GRADE The degree of perfusion in the infarct-related artery (IRA) is
typically described by the TIMI flow grade:
OPEN VESSEL HYPOTHESIS In the early 1940s, the possible relationship between a
patent IRA, myocardial damage, and mortality was first advanced in the open vessel
hypothesis. This theory holds that early reperfusion results in myocardial salvage,
preserves ventricular performance, and is ultimately responsible for improved patient
survival.
The foundation for the open vessel hypothesis came from the demonstration in dogs that
the extent of infarction was affected by the duration of coronary arterial occlusion. As a
result of this and subsequent investigations, the clinical importance of a patent, as
opposed to an occluded infarct-related coronary artery, is now widely accepted.
Compared to those with a patent artery and 50 percent residual stenosis, those with
mild residual stenosis were younger, had fewer prior infarctions, and fewer eccentric or
ulcerated lesions, but also had a greater thrombus burden. A milder residual stenosis was
associated with a lower incidence of in-hospital death, myocardial infarction, or heart
failure (2.8 versus 7 percent).
ST segment resolution and IRA ST segment resolution also correlates with IRA
patency and TIMI flow grade (figure 3). This was seen in the TIMI 14 trial of 444 patients
with interpretable ECGs who underwent angiography at 90 minutes [18]. Seventy-nine
percent of those with complete (70 percent) ST segment resolution had TIMI grade 3
flow and their 30-day mortality was 1 percent. In contrast, patients with partial or no ST
segment resolution were less likely to have TIMI grade 3 flow (50 and 44 percent,
respectively) and had a higher 30-day mortality (4.2 and 5.9 percent, respectively)
(figure 4). Although patients without ST segment resolution may still have a patent
coronary artery, they are nevertheless at increased risk for mortality, which is likely due
to extensive microvascular and tissue injury. (See "Fibrinolytic (thrombolytic) agents in
acute ST elevation myocardial infarction: Markers of efficacy".)
Blood flow in nonculprit artery Although reperfusion trials have focused on blood
flow in the culprit artery, flow in the nonculprit arteries is also of importance. It has been
assumed that flow in these arteries is normal, but data are limited. One report measured
the corrected TIMI frame count (ie, the number of frames required for contrast to reach a
standardized distal landmark) in 1817 nonculprit arteries of patients entered into the
TIMI 4, 10A, 10B, and 14 fibrinolytic trials [20]. At 90 minutes, blood flow in the
nonculprit vessels was 45 percent slower than normal flow in the absence of a myocardial
infarction. Patients with reduced blood flow in nonculprit arteries had a higher incidence
of regional wall motion abnormalities within the distribution of the nonculprit vessel.
Culprit and nonculprit blood flows were directly related at baseline, and relief of the
residual stenosis in the culprit artery with angioplasty restored its flow to a level equal to
the nonculprit artery; however, after angioplasty, blood flow in both culprit and
nonculprit arteries remained 45 percent below normal.
As described below, the attainment of a patent IRA after fibrinolysis is associated with
lower rates of morbidity and mortality. Outcomes are also improved with spontaneous
reperfusion [22,23].
In the PAMI trials, for example, 16 percent of patients had TIMI 3 (normal) flow before
PCI (ie, spontaneous reperfusion) [22]. The patients with TIMI 3 flow had the following
significant benefits: greater clinical and angiographic evidence of myocardial salvage; a
lesser likelihood of presenting in or developing heart failure; and a lower rate of early
and late mortality (six month mortality: 0.5 versus 2.8 and 4.4 percent with TIMI grade
2 and 0/1 flow, respectively).
Non-ST elevation MI Spontaneous early reperfusion is much more common in
patients with non-ST elevation acute coronary syndromes. Coronary arteriography,
performed during the acute period in these disorders, demonstrates that the IRA is not
occluded in 60 to 85 percent of cases [24-27]. The nonoccluding thrombi that are present
are primarily grayish-white (ie, platelet-rich), and therefore less likely to respond to
fibrinolytic therapy [28], in contrast to being almost always reddish (ie, fibrin-rich) in
patients with ST elevation MI [29]. In addition, microvascular perfusion is often reduced
in patients with non-ST elevation acute coronary syndromes; as a result, the ongoing
mechanism of ischemia is more likely embolization than epicardial vessel occlusion [27].
(See "Fibrinolytic (thrombolytic) agents in unstable angina and acute non-ST elevation
myocardial infarction".)
IRA PATENCY AND OUTCOME AFTER FIBRINOLYSIS Patency of the IRA after
fibrinolytic therapy appears to have beneficial effects on both the short and long term
outcome of patients with acute MI. This is most likely to occur when therapy is given
within two to three hours of symptom onset (figure 5). A meta-analysis from the
Fibrinolytic Therapy Trialists' (FTT) Collaborative Group found that the absolute mortality
benefit from fibrinolytic therapy at five weeks was 3 percent for those presenting within
six hours from symptom onset, 2 percent for those presenting within 7 to 12 hours, and
a nonsignificant 1 percent for those presenting within 13 to 18 hours [30].
(See "Fibrinolytic (thrombolytic) agents in acute ST elevation myocardial infarction:
Therapeutic use", section on 'Time to therapy (door-to-needle time)'.)
This important observation has subsequently been confirmed by data from a number of
other trials (figure 6) [32-36]. GUSTO 1, for example, tested the hypothesis that early
and sustained patency of an infarct-related vessel is associated with improved survival
[34-36]. Forty-one thousand patients in over 1,000 hospitals in 15 countries with
evolving myocardial infarction presenting within six hours of symptom onset were
randomized to streptokinase (1.5 million units over 60 minutes) plus either subcutaneous
heparin (12,500 units BID) or intravenous heparin (5,000 unit bolus followed by 1,000
units per hour), or to accelerated alteplase (recombinant tissue-type plasminogen
activator or tPA) with intravenous heparin (accelerated refers to the rate at which
alteplase is given). (See "Fibrinolytic (thrombolytic) agents in acute ST elevation
myocardial infarction: Therapeutic use".) A fourth arm consisted of a combination of
streptokinase and alteplase.
Alteplase was associated with a 1 percent absolute reduction in mortality (6.3 versus 7.3
percent) compared to streptokinase. The mortality benefit for alteplase was greatest in
patients less than 75 years old and in those with anterior wall infarctions. The time of
fibrinolysis was also important, independent of the fibrinolytic agent used (alteplase or
streptokinase) [37]. Early fibrinolysis was associated with a lower 30 day mortality rate
(less than two hours 5.5 percent; greater than four hours 9.0 percent).
When data from 11,228 patients enrolled in GUSTO 1 were analyzed, 45 percent of
patients overall had an open IRA, defined as the presence of TIMI grade 3 flow [38].
Those with an open artery had a lower 30 day mortality (1.5 versus 6.3 percent for a
closed artery) even after adjusting for baseline clinical predictors of outcome including
ejection fraction. At one year, the mortality rate among those surviving after the first 30
days was still lower for those with an open vessel (3.3 versus 8.8 percent); however,
after adjusting for clinical variables, patency was no longer a significant predictor and the
only predictor of mortality was the left ventricular ejection fraction (LVEF) (figure 7).
Patency also is associated with lower morbidity. This was illustrated in a study of 327
patients who received fibrinolysis [39]. Those with coronary artery reperfusion based
upon clinical markers (eg, relief of pain, 50 percent reduction in the sum of ST segment
elevation, and abrupt initial release of creatine kinase levels >2 fold over normal) not
only had a lower 30-day mortality (3.8 versus 16.2 percent for those without evidence of
reperfusion), but also had a lower incidence of heart failure (10.5 versus 23.1 percent)
and less often progressed to cardiogenic shock (0.5 versus 12.8 percent).
TIMI frame count The clear correlation between coronary flow and patient survival
has stimulated enthusiasm for the development of more quantitative and precise
measures than the semi-quantitative TIMI flow scale. It is hoped that increasingly
accurate measures of coronary arterial flow velocity will minimize the limitation resulting
from interobserver variability seen with the present system. The TIMI frame count, which
is the number of cine frames required for dye to first reach standardized distal coronary
landmarks, has shown great promise, particularly in defining low-risk and high-risk
patient subsets [48].
One study measured the TIMI frame count in 1248 patients treated with fibrinolytic
agents in the TIMI 4, 10, and 10B trials [49]. The following findings were noted (figure
10):
Patients who died in the hospital had a higher frame count (slower flow)
compared to survivors (70 versus 50); a higher frame count was also observed in
those who died by 30 to 42 days. In a multivariate model that excluded the TIMI
flow grade, there was a 0.7 percent increase in absolute mortality for every 10-
frame rise.
The frame count further stratified patients with TIMI grade 3 flow. Although a
frame count 40 is associated with TIMI grade 3 flow, those with a frame count
20 had a 7.9 percent incidence of in-hospital adverse events (death, recurrent
infarction, shock, congestive heart failure or left ventricular ejection fraction 40
percent) compared to 15.5 percent for those with a frame count between 21 and
40.
Patients who had an adverse in-hospital outcome (death, recurrent infarction,
shock, heart failure, or ejection fraction 40 percent) had a slower global TIMI
frame count (ie, the average of all three coronary arteries) [20].
Another small study of 104 patients with a first MI found that a lower TIMI frame count
(faster flow) immediately after successful PCI (TIMI flow grade 3) predicted recovery of
regional LV function [50].
Although intraobserver and interobserver reproducibility of the TIMI frame count is good,
there are factors that can introduce variability and decrease the TIMI frame count,
including nitrate use, increasing heart rate, and injection of dye during the beginning of
diastole (versus the beginning of systole) [51].
Further work in this area is needed, particularly if angiographic flow grade is to be used
as a surrogate end point in the development of new fibrinolytics and adjunctive
treatment strategies.
Plaque morphology and coronary blood flow The relationship between plaque
rupture and coronary thrombosis among patients with acute coronary syndromes is well
established. Rupture involving the proximal aspect of an atheromatous plaque, compared
to mid-or distal plaque rupture is much more likely to cause occlusive thrombus, TIMI O
blood flow and ST-segment elevation MI. A larger proportion of fibro-fatty material, as
determined by intravascular ultrasound and greater overall plaque burden associated
with reduced myocardial blood flow (myocardial blush grade) and no reflow in patients
with MI and TIMI 3 coronary artery blood flow following primary PCI [53-55].
IRA PATENCY AND OUTCOME AFTER PCI There are two issues related to IRA
patency and PCI: reocclusion after PCI; and late PCI to open an occluded artery.
Reocclusion after PCI Late reocclusion after successful PCI has a negative impact on
long-term survival. This was illustrated in a review of 528 patients with a patent IRA
after a successful PCI who underwent a repeat angiogram at six months [57]. At 5.7 year
follow-up, those with reocclusion had a higher actuarial eight year total mortality rate (28
versus 10 percent) and higher cardiovascular mortality (25 versus 7 percent) compared
to those without reocclusion; the impact on long-term survival was greater in those with
an anterior wall MI (figure 11).
LATE PCI TO OPEN AN OCCLUDED ARTERY This section will address the difficult
issue of when to consider percutaneous coronary intervention (PCI) later than 24 hours
after STEMI. The discussion of when to consider PCI between 12 and 24 hours after
STEMI is found elsewhere. (See "Selecting a reperfusion strategy for acute ST elevation
myocardial infarction", section on 'Delayed presentation'.)
Fibrinolytic therapy administered more than 24 hours after the onset of acute ST
elevation MI does not improve clinical outcome (figure 5), possibly because it does not
restore artery patency. Although early PCI is also optimal, some have suggested that late
PCI to open an occluded artery might be of benefit. (See "Fibrinolytic (thrombolytic)
agents in acute ST elevation myocardial infarction: Therapeutic use", section on 'Time to
therapy (door-to-needle time)' and "Primary percutaneous coronary intervention in acute
ST elevation myocardial infarction: Determinants of outcome", section on 'Time from
hospital arrival (door-to-balloon time) on outcomes'.)
Recovery of left ventricular function decreases with later PCI [58], but late PCI may still
be beneficial at a time when fibrinolytic therapy is no longer effective [59]. This is an
important issue, since registry data suggest that 9 to 31 percent of patients with an
STEMI present more than 12 hours after the onset of symptoms [60,61]. It has been
estimated that, in the absence of reperfusion, an occluded IRA is present in 87 percent of
patients with an STEMI at four hours, 65 percent at 12 to 24 hours, and 45 percent at
one month [1,21].
A possible mechanism for benefit from late PCI is residual antegrade or collateral blood
flow, which was present in 73 percent of patients in the BRAVE-2 trial cited below [62].
The low rate of persistent flow may maintain myocardial viability and therefore infarct
size until blood flow is restored by PCI [63,64]. Mechanisms other than myocardial
salvage, such as prevention of ventricular remodeling, also may be involved [65,66].
There are two settings in which late PCI might be performed: in previously untreated
patients, which will be reviewed here; and in patients previously treated with fibrinolytic
therapy, which is a form of adjunctive or early elective PCI that will be discussed
separately. (See "Percutaneous coronary intervention after fibrinolysis for acute ST
elevation myocardial infarction", section on 'Adjunctive or early elective PCI'.)
A potential benefit of late PCI was suggested by observational data from the National
Registry of Myocardial Infarction (NRMI)-2 of 7358 patients with acute STEMI who
presented more than 12 hours after symptom onset [67]. Patients undergoing PCI (22
percent of the group) had a significantly lower incidence of recurrent ischemia or angina,
recurrent MI, and in-hospital mortality (3.4 versus 6.6 percent without PCI); the
association between PCI and lower mortality persisted after adjusting for factors
associated with a higher in-hospital mortality (odds ratio 0.67, 95% CI 0.49-0.92). With
a propensity analysis, which helps to adjust for treatment selection bias, there was still a
significant reduction in recurrent ischemia with PCI (10.8 versus 14.6 percent), and a
strong trend toward a reduction in in-hospital mortality (odds ratio 0.73, 95% CI 0.53-
1.01).
The randomized trials BRAVE-2, DECOPI, and OAT that have evaluated late PCI
have included patients at different time periods after symptom onset, ranging from more
than 12 hours to up to 28 days. Some have demonstrated an improvement in left
ventricular function with intervention, but none has demonstrated a significant benefit on
hard clinical outcomes.
The SWISSI II trial, which randomly assigned patients with both STEMI (including many
treated with fibrinolytic therapy) and NSTEMI to late PCI or medical therapy is discussed
separately. (See "Silent myocardial ischemia: Prognosis and therapy", section on
'Treatment with revascularization'.)
The BRAVE-2 trial evaluated 365 patients who presented with an acute STEMI 12 to 48
hours (median 23 hours) after the onset of symptoms [62]. The patients, who were no
longer symptomatic, were randomly assigned to an invasive strategy, which primarily
consisted of PCI with stenting, or conventional conservative therapy. The primary end
point was infarct size at 5 to 10 days as determined by Tc-99m sestamibi SPECT
imaging. There were two major findings:
Final left ventricular infarct size was significantly smaller in the late PCI group
(median 8 versus 13 percent). The size of the infarct with conservative therapy
was smaller than the expected value of 20 percent and may have reflected some
antegrade flow and some collateral flow to the infarct zone.
The secondary end point of death, recurrent MI, or stroke occurred less often with
late PCI (4.4 versus 6.6 percent). Although this difference was not significant
(relative risk 0.67, 95% CI 0.27-1.62), the study was underpowered to detect
such a modest difference in clinical outcomes.
The efficacy of primary PCI performed even later was evaluated in the DECOPI and OAT
trials. Neither showed a mortality benefit from intervention. In DECOPI, 212 patients with
a first STEMI and an occluded infarct-related artery who were seen 2 to 15 days (mean 5
days after symptom onset) underwent coronary angiography and were randomly
assigned to continued medical therapy or PCI [68]. Most patients had single vessel
disease and only 13 percent had been treated with fibrinolytic therapy.
At six months, more patients in the intervention group had a patent infarct-related artery
(82 versus 34 percent) and the LVEF was 5 percent higher (59 versus 54 percent), a
benefit seen only in those with patent infarct-related arteries. However, at a mean
follow-up of 34 months, the primary end point (a composite of cardiac death, nonfatal
MI, or ventricular tachyarrhythmia) was not significantly different in the two groups (7.3
versus 8.7 percent).
The much larger Occluded Artery Trial (OAT) included 2166 stable patients who had an
occluded IRA as determined by cardiac catheterization performed 3 to 28 days after
acute STEMI [69]. The patients were randomly assigned to optimal medical therapy alone
or with routine PCI with stenting. Important characteristics of the patients in this trial
included high-risk features (left ventricular ejection fraction [LVEF] less than 50 percent
or proximal occlusion of a major epicardial vessel) and the absence of recurrent clinical
ischemia or the demonstration of severe ischemia on stress testing. Twenty percent had
received fibrinolytic therapy.
The primary end point was a composite of death, myocardial reinfarction, or New York
Heart Association class IV heart failure (table 1). At four years, the rate of the composite
end point was not statistically different between the PCI and the medical therapy groups
(17.2 and 15.6 percent, respectively). It should be emphasized that this neutral result
was obtained despite the presence of retained viability in about 600 of patients. What
remains to be determined is whether the extent of viability could have an impact upon
outcomes, even if not shown in this study. In addition a quality of life substudy found
there was no significant difference between the two groups in measures of cardiac
physical function or psychological well-being at two years [70].
The logic for attempts to demonstrate the value of late opening of an occluded IRA in the
OAT trial was based upon the presumption that treated patients would have an
improvement in LVEF compared to patients who did not undergo PCI. TOSCA-2, an
angiographic follow-up substudy of 381 patients in OAT, was not able to demonstrate a
greater improvement in LVEF in patients with PCI than in those receiving medical therapy
[71]. At one year, there was no significant difference between the PCI and medical
therapy groups with regard to the increases seen in LVEF (48.3 to 52.5 and 48.0 to 51.2
percent, respectively).
Summary The evidence that can be applied to answer the question of whether late
PCI (after 24 hours) of the infarct related artery in stable patients improves outcomes
does not allow for a firm conclusion. We agree with recommendations made in the 2007
American College of Cardiology/American Heart Association focused update of prior
STEMI guidelines (not changed in the 2009 focused update) [73]:
The 2008 European Society guidelines on the Management of acute myocardial infarction
in patients presenting with persistent ST-segment elevation does not address this issue
and states that there is limited experience with PCI in STEMI patients presenting more
than 12 hours after onset of symptoms [74].
The optimal approach to similar patients who present between 12 and 24 hours is also
not known.
LONG-TERM VALUE OF A PATENT IRA Among the best data on the long-term
prognostic importance of IRA patency comes from the SAVE trial. In the SAVE trial of
post-MI patients with LV dysfunction, an occluded IRA was an independent predictor of
all-cause mortality (24 versus 14 percent for patients with an occluded or patent IRA,
respectively), cardiovascular mortality (23 versus 12 percent), and the composite end
point of cardiovascular mortality or morbidity (51 versus 37 percent) at an average
follow-up of 3.5 years.
The salutary effects of IRA patency in SAVE were independent of the benefits associated
with medical therapy (which included angiotensin-converting enzyme inhibition and beta
blockade). This study differed from other studies in that IRA patency was a prespecified
ancillary end point; the analysis was restricted to post-MI patients with LV dysfunction
without clinically apparent ischemia or heart failure, baseline coronary anatomy was
known in all patients, a variety of treatment strategies culminating in IRA patency were
included, and corrections for clinical variables, including LV function, were made.