JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO.

-, 2018
ª 2018 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

PUBLISHED BY ELSEVIER

Maximal Pre-Excitation Based Algorithm

for Localization of Manifest Accessory
Pathways in Adults
Thomas Pambrun, MD,a,b Rim El Bouazzaoui, MD,a Nicolas Combes, MD,a Stéphane Combes, MD,a Pedro Sousa, MD,c
Mathieu Le Bloa, MD,a Grégoire Massoullié, MD,b Ghassen Cheniti, MD,b Ruairidh Martin, MD,b Xavier Pillois, PHD,b
Josselin Duchateau, MD,b Frédéric Sacher, MD, PHD,b Mélèze Hocini, MD,b Pierre Jaïs, MD,b Nicolas Derval, MD,b
Agustín Bortone, MD,d Serge Boveda, MD,a Arnaud Denis, MD,b Michel Haïssaguerre, MD,b Jean-Paul Albenque, MDa

ABSTRACT

OBJECTIVES The study evaluated a new algorithm relying on maximal pre-excitation.

BACKGROUND Prior knowledge of accessory pathway (AP) location facilitates an individual ablation strategy. Delta-
wave analysis on a 12-lead electrocardiogram is recognized as crucial for predicting ablation site, but can be ambiguous at
basal state.

METHODS An algorithm based on maximal pre-excitation, as induced by atrial pacing during an electrophysiological
study, was initially developed in 132 patients with a single manifest AP. The maximally pre-excited QRS features included
the global polarity in lead V1 (step 1), inferior leads (step 2), and leads V3 or I (step 3), as well as the morphology in lead II
(step 4). Three investigators prospectively tested the new algorithm in 207 consecutive patients by comparing its efficacy
to a control algorithm relying on basal pre-excitation.

RESULTS The accuracy, defined as the percent of patients with an exact prediction of AP location, was significantly
greater with the new algorithm (90% vs. 63%; p < 0.001). The reproducibility, defined as the level of agreement be-
tween investigators in determining AP location, was excellent (k > 0.75; p < 0.05) with the new algorithm and fair
(0.40 < k < 0.75; p < 0.05) with the control algorithm.

CONCLUSIONS An algorithm based on maximal pre-excitation allows accurate and reproducible localization of mani-
fest APs. When ablation is indicated, the analysis of maximal pre-excitation is a sensible approach for giving a head start
in endocardial mapping. (J Am Coll Cardiol EP 2018;-:-–-) © 2018 by the American College of Cardiology Foundation.

C atheter ablation of the accessory pathway

(AP) can be indicated in patients with pre-
excitation (1). Pre-ablation AP localization
inferred from the 12-lead surface electrocardiogram
(ECG) facilitates a tailored ablation strategy, whose
specific risk is properly assessed (2–11). As they are
based on pre-excitation in the basal state, available
algorithms can lead to complex and ambiguous
analysis, mostly due to the variable influence of the
AP over QRS morphology (12–15). The electrophysio-
logical study can maximize this influence by maneu-
vers such as rapid atrial pacing.
To date, studies considering maximally pre-excited
ECG have not proposed a stepwise algorithm (6–11).
Therefore, the present study aimed to develop an
algorithm based on maximal pre-excitation and

From the aDépartement de Rythmologie, Clinique Pasteur, Toulouse, France; bHôpital Cardiologique du Haut-Lévêque, CHU
Bordeaux, L’Institut de RYthmologie et modélisation Cardiaque (LIRYC), Université Bordeaux, Bordeaux, France; cCardiology
Department, University Hospital of Coimbra, Coimbra, Portugal; and the dService de Cardiologie, Hôpital Privé Les Franciscaines,
Nîmes, France. This study received financial support from the French Government as part of the “Investments of the Future”
program managed by the National Research Agency (ANR), Grant reference ANR-10-IAHU-04. Dr. Albenque has served as a
consultant for Abbott, Biosense Webster, Inc., and ACT. All other authors have reported that they have no relationships relevant to
the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ in-
stitutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit
the JACC: Clinical Electrophysiology author instructions page.

Manuscript received February 5, 2018; revised manuscript received March 6, 2018, accepted March 29, 2018.

ISSN 2405-500X/$36.00 https://doi.org/10.1016/j.jacep.2018.03.018

2 Pambrun et al.
Maximal Pre-Excitation and Accessory Pathway Location
ABBREVIATIONS compare its efficacy to a widely used algo-
AND ACRONYMS rithm relying on basal pre-excitation.
AP = accessory pathway
METHODS
CS = coronary sinus
DCS = deep coronary sinus
GLOBAL POPULATION. The study cohort
ECG = electrocardiogram consisted of 207 consecutive patients over 15
LL = left lateral years of age, referred to our institutions for
LPL = left posterolateral assessment of a manifest AP. Patients were
LPS = left paraseptal included when they fulfilled the following
NH = nodo-Hisian criteria: 1) indication for first ablation of a
RA = right anterior manifest AP (1); 2) absence of structural heart
RL = right lateral
disease; 3) absence of multiple APs, as
assessed by electrophysiological study; and
RP = right posterior
4) absence of Mahaïm fibers.
RPS = right paraseptal
The local institutional ethics committees
on human research approved the study protocol.
Written and informed consent was obtained from all
patients.
PRE-EXCITATION MAXIMIZATION AND ABLATION
PROCEDURE. Maximal pre-excitation was systemat-
ically provoked during an electrophysiological study,
even when previously documented during sponta-
neous pre-excited AF. All procedures were performed
in sinus rhythm after withdrawal of all antiar-
rhythmic drugs, with 2 diagnostic catheters percuta-
neously inserted through the right femoral vein.
Atrial pacing aimed at determining the AP antero-
grade refractory period to risk stratify
life-threatening arrhythmic events and to induce
AP-related tachycardia. This maneuver provided the
maximal pre-excitation upon which the AP location
was predicted by the new algorithm.
When ablation was indicated, 1 diagnostic catheter
was moved into the coronary sinus (CS), serving as an
anatomical landmark of the CS ostium and the mitral
annulus, while the other diagnostic catheter was
placed at the His bundle. The ablation catheter posi-
tion was ascertained in the left anterior oblique view,
so that the tip of the His-bundle catheter could
identify both the annuli frontal plane by pointing
toward the operator and the CS ostium located
directly below, at the same sagittal plane. During
ablation, the ventricular insertion site was systemat-
ically targeted, as previously described (16). Vascular
access (femoral vein or artery), additional material
(sheath for catheter stability), and energy source
(standard radiofrequency, irrigated radiofrequency,
or cryothermy) varied with AP location (Figure 1).
ALGORITHM DEVELOPMENT. The terminology of the
9 anatomical locations of the AP was defined as fol-
lows: 1) right anterior (RA), from 10 to 1 o’clock; 2)
right lateral (RL), from 8 to 10 o’clock; 3) right
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018
- 2018:-–-
posterior (RP), below the CS ostium and extending to
8 o’clock; 4) right paraseptal (RPS) (formerly known
as posteroseptal), near the CS ostium, including its
terminal portion; 5) nodo-Hisian (NH), above the CS
ostium and extending to 1 o’clock, therefore
including the atrioventricular node (midseptal area)
and His bundle (anteroseptal area); 6) deep CS (DCS),
more than 1 cm within the CS, including the middle
cardiac vein; 7) left paraseptal (LPS) (formerly known
as left posteroseptal), from 6 to 8 o’clock; 8) left
posterolateral (LPL), from 4 to 6 o’clock; and 9) left
lateral (LL), from 0 to 4 o’clock (Figure 1).
Maximal pre-excitation, obtained by rapid atrial
pacing for assessment of AP anterograde effective
refractory period, was recorded on the surface ECG at
a speed of 25 mm/s and band-pass filter of 0.1 to 50
Hz. The QRS polarity was defined as positive or
negative, depending on whether the QRS morphology
was mainly above or under the baseline. Of impor-
tance, in the particular case of isoelectric QRS
complex (QR or RS pattern), the polarity of the initial
deflection (negative for QR pattern and positive for
RS pattern) determined the classification. Hence, 9
morphologies were observed and categorized as: 1)
negative for the QS-smooth, QS-notched, QrS, rS, or
QR pattern; and 2) positive for the RS, rsr’, Rs, or R
pattern (Figure 2A). The QRS ratio between 2 leads
compared the amplitude of their positive component,
the greatest being defined as the one displaying the
for
highest amplitude from the onset to the peak of the R-
wave (Figure 2B).
The new algorithm was empirically developed in
132 initial patients by correlating QRS polarity
during maximal pre-excitation and the exact
anatomical location of the AP, defined as the site
where energy application led to abolition of the
pre-excitation.
ALGORITHM ASSESSMENT. The new algorithm was
then prospectively tested in 207 consecutive patients.
The exact ablation site of the AP being determined by
the operator, 3 investigators used the new algorithm
to decipher the AP location on the maximally
pre-excited ECG. All 3 were blinded to the ablation
procedures and to each other’s conclusions.
The results of the new algorithm, based on
maximal pre-excitation, were compared with those
obtained with an algorithm relying on basal pre-
excitation (5). The choice of Arruda’s algorithm was
supported by a similar classification of the AP
anatomical location, hence simplifying the compari-
son between both algorithms (Table 1).
STATISTICAL ANALYSIS. Categorical data, expressed
as percentage, were compared using the chi-square
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018 Pambrun et al. 3
- 2018:-–- Maximal Pre-Excitation and Accessory Pathway Location

F I G U R E 1 Accessory Pathway’s Distribution

Schematic representation of the atrioventricular annuli, the atrioventricular node (AVN) and the His bundle (HIS), as viewed in the left anterior oblique view. Small black
numbers (132 initial patients for the development of the new algorithm) and bold blue numbers (207 consecutive patients for the prospective assessment of the new
algorithm) illustrate accessory pathway distribution across the 9 locations: right anterior (RA), right lateral (RL), right posterior (RP), right paraseptal (RPS), nodo-Hisian
(NH), deep coronary sinus (DCS), left paraseptal (LPS), left posterolateral (LPL), and left lateral (LL). Preferential ablation approaches performed in our institutions are
described, according to accessory pathway location.

test. The accuracy of the algorithm was defined as
the percent of patients with an exact prediction of the
successful ablation site. The reproducibility of the
algorithm was defined as the level of agreement be-
tween investigators in determining AP location.
Kappa values below 0.40, from 0.40 to 0.75, and over
0.75 were considered to indicate poor, fair, and
excellent agreement, respectively, and p values
<0.05 were considered statistically significant.
RESULTS
ALGORITHM DESCRIPTION. The new algorithm,
depicted in Figures 3 and 4, is articulated around 4
steps.
S t e p 1 ( l e a d V 1 p o l a r i t y ) . Indicates a right-sided AP
(including NH AP) when negative and a left-sided AP
when positive.
S t e p 2 ( i n f e r i o r l e a d s p o l a r i t y ) . When none are
positive, it indicates RP or RPS locations (posterior
pair) for right-sided APs and DCS, LPS, or LPL loca-
tions for left-sided APs. When 1 or 2 are positive, it
indicates RL or NH locations (intermediate pair) for
right-sided APs, and directly identifies LPL location
for left-sided APs. When all 3 are positive, it indicates
RA or NH locations (anterior pair) for right-sided APs,
and directly identifies LL location for left-sided APs.
S t e p 3 ( l e a d V 3 p o l a r i t y ) . Only examined for the
final identification of right-sided APs. A negative and
a positive polarity distinguish RP from RPS location,
respectively, for a posterior pair; RL from NH location
for an intermediate pair; and RA from NH location for
an anterior pair.
S t e p 3 ( V 1 / I r a t i o ) . Only examined in case of pos-
terior left-sided APs. A ratio <1 indicates DCS or LPS
location. A ratio $1 directly identifies LPL location.
4 Pambrun et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018
Maximal Pre-Excitation and Accessory Pathway Location - 2018:-–-

F I G U R E 2 Morphologies of Maximally Pre-Excited QRS Complex

Example of QRS morphologies typically observed during maximal pre-excitation. (A) Classification into negative or positive polarity. (B) QRS ratio between
lead V1 and lead I.

S t e p 4 ( l e a d I I m o r p h o l o g y ) . Only examined to values of the 2 algorithms for each location are

discriminate DCS from LPS APs. A notched QS in- summarized in Table 2. Regarding the accuracy, AP
dicates DCS location. locations were correctly identified by the 3 in-
ALGORITHM EFFICACY. The distribution of suc- vestigators in 90% (558 of 621) of patients with the
cessful AP ablation sites is shown in Figure 1. The new algorithm and 63% (389 of 621) of patients with
averaged sensitivity, specificity, and predictive Arruda’s algorithm (p < 0.001). Accuracy proved
consistently higher with the new algorithm,
T A B L E 1 Correlation Between Algorithm Nomenclature
regardless of which investigator it was (for investi-
gator 1: 92.3% vs. 67.1%, p < 0.001; for investigator
Anatomical Location of the AP
2: 88.9% vs. 63.3%, p < 0.001; for investigator 3:
Study algorithm RA RL RP RPS NH DCS LPS LPL LL
88.4% vs. 57.5%, p < 0.001). Regarding the repro-
Arruda algorithm RA RL RP PSTA MSTA MCV PSMA LP LL
RAL RPL AS VA LP LPL LAL ducibility, agreement between investigators was
excellent (k > 0.75) with the new algorithm and fair
Of note, because no Delta-wave characteristics could reliably differentiate these contiguous lo-
(0.40 < k < 0.75) with Arruda’s algorithm (Table 3).
cations, the following pairs of regions were grouped together in the original article by Arruda
et al. (5): right anterior (RA) and right anterolateral (RAL), right posterior (RP) and right A representative maximally pre-excited 12-lead ECG
posterolateral (RPL), middle cardiac vein (MCV) and venous anomaly (VA), left posterior (LP) and
left posterolateral (LPL), left lateral (LL) and left anterolateral (LAL).
of the 9 predefined AP locations is illustrated in
AP ¼ accessory pathway; AS ¼ anteroseptal; DCS ¼ deep coronary sinus; LPS ¼ left paraseptal; Figure 5. Detailed information about the prediction
MSTA ¼ mid-septal tricuspid annulus; NH ¼ nodo-Hisian; PSMA ¼ posteroseptal mitral annulus;
PSTA ¼ posteroseptal tricuspid annulus; RL ¼ right lateral; RPS ¼ right paraseptal.
of each investigator for the 2 algorithms is provided
in the supplementary data (Online Tables 4 to 9).
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018 Pambrun et al. 5
- 2018:-–- Maximal Pre-Excitation and Accessory Pathway Location

F I G U R E 3 New Stepwise Algorithm Depicted as an Anatomical Scheme

Accessory pathway (AP) locations are green when right sided and red when left sided. AP locations are typed in white in the absence of positive polarity in the inferior
leads, in gray when positive polarity is observed in 1 or 2 inferior leads, and in black when all 3 inferior leads display a positive polarity. Of note, LPL APs can have 0, 1, or
2 inferior leads with positive polarity, whereas NH APs can have 1, 2, or 3 inferior leads with positive polarity. Right-sided APs are framed orange or yellow when the V3
lead is negative or positive, respectively. Left posterior APs are framed blue when V1/I ratio is <1 or purple when V1/I ratio is $1. Abbreviations as in Figure 1.

DISCUSSION pre-excitation are mostly found when QRS complex

The present study reports a stepwise algorithm
exclusively based on induced maximal
excitation. This approach demonstrates accurate and
reproducible localization of manifest APs, while being
easily performed during a standardized electrophys-
iological study.
PRE-EXCITATION MAXIMIZATION: AN ACCURATE AND
REPRODUCIBLE APPROACH. Studies comparing avail-
able algorithms systematically show variable efficacy,
which may be explained by ambiguous interpretation
of the Delta wave in the basal state (12–15).
Disagreement between algorithms relying on basal
duration drops below 120 ms, due to competing acti-
vation from the His-Purkinje system, resulting in a
pre- restriction of the AP influence over QRS morphology
(15). Therefore, it has been proposed to circumvent
this pitfall by focusing on the initial 20 to 40 ms of the
Delta wave. Nevertheless, whether pre-excitation is
minimal or maximal, the final part of the QRS com-
plex results either from fast ubiquitous activation
through the His-Purkinje system or slow focal acti-
vation originating from the annulus. As a result,
beyond the Delta wave, which reflects ventricular
depolarization at the vicinity of the AP, the rest of the
maximally pre-excited QRS complex also contains
additional information related to AP location, while
6 Pambrun et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018
Maximal Pre-Excitation and Accessory Pathway Location - 2018:-–-

F I G U R E 4 New Stepwise Algorithm Depicted as a Decision Tree

Abbreviations as in Figure 1. Color coding is as described in Figure 3.

the minimally pre-excited QRS complex does not
(Figure 6A). Furthermore, confining the analysis to
the first 20 ms of the Delta wave raises 2 practical
concerns. First, the Delta-wave onset at 1 specific
lead, measured from the earliest onset simulta-
neously identified in any other leads, can be equiv-
ocal, depending on the lead subjectively selected as
the reference. Second, this indirect method requires
classification of Delta-wave polarity as positive,
negative, or isoelectric, which, given the short win-
dow of analysis, can be difficult to differentiate when
the Delta-wave slope is not steep (Figure 6B). Hence,
the analysis of the entire QRS polarity after pre-
excitation maximization by rapid atrial pacing gives

T A B L E 2 Algorithm Performance for Each of the 9 Locations

Sensitivity (%) PPV (%) Specificity (%) NPV (%)

Ablation Study Arruda’s Study Arruda’s Study Arruda’s Study Arruda’s

Site Algorithm Algorithm Algorithm Algorithm Algorithm Algorithm Algorithm Algorithm

RA 91 29 88 46 99 97 99 94
RL 100 33 85 61 100 100 100 99
RP 96 11 87 33 99 98 99 93
RPS 83 65 97 64 99 92 97 92
NH 78 59 76 26 98 92 99 98
DCS 67 6 71 2 99 93 99 97
LPS 74 67 77 78 97 98 97 96
LPL 92 58 86 65 98 95 99 94
LL 100 89 98 85 99 92 100 95

NPV ¼ negative predictive value; PPV ¼ positive predictive value; other abbreviations as in Table 1.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018
- 2018:-–-
more detailed information and less room for inter-
pretation, as shown by the more accurate and repro-
ducible AP’s localization found in the present study.
MAXIMAL PRE-EXCITATION: NEGLECTED DATA.
Recommendations for catheter ablation of a manifest
AP usually refer to situations where maximal pre-
excitation is collected de facto, either spontaneously
(documentation of a pre-excited AF) or during an
electrophysiological study (1). In the latter case,
maximal pre-excitation is obtained during assess-
ment of AP anterograde refractory period, either for
induction of AP-related tachycardia in symptomatic
patients or for risk stratification in asymptomatic
patients. Interestingly, the collection of this previ-
ously neglected data does not preclude any ablation
strategies, as no decision has been taken at this stage
regarding the choice of vascular access, additional
material, or energy source. It could be argued that,
once catheters are in place, a careful analysis of
intracardiac electrograms is mandatory for ablation.
But in daily practice, rather than randomly starting
endocardial mapping at an undetermined area of the
annulus, a previous analysis of pre-excitation
provides an efficient head start by giving a clear
idea of the area to primarily target. From this respect,
although basal pre-excitation remains useful during
outpatient consultation to inform about the specific
risks related to AP location, maximal pre-excitation
proves more effective in the EP laboratory to guide
endocardial mapping. Of note, maximal pre-
excitation can be noninvasively obtained with aden-
osine infusion. But in contrast to an electrophysio-
logical study, adenosine is contraindicated in patients
with severe or poorly controlled airway disease and
does not provide any indication for AP ablation.
RELEVANCE OF DISTINCT MAXIMALLY PRE-EXCITED
QRS FEATURES FOR AP LOCALIZATION. Previous
studies have analyzed the relationship between
maximal pre-excitation pattern and AP location.
However, the vast majority of studies were restricted
to septal APs (6–10). To our knowledge, only 1 study
considered the entire annulus perimeter, but did not
propose an algorithm (11). By providing a global
stepwise algorithm, the current study may help to
rule out potential similarities of the QRS pattern be-
tween adjacent locations. Subsequently, we consider
the distinct QRS features identified, with reference to
previously published studies, where possible.
S t e p 1 . Lead V 1 polarity, either negative or positive,
appears essential to distinguish right from left APs,
respectively. Accordingly, Fananapazir et al. (11) re-
ported that >75% of APs display a negative lead V 1
polarity when right sided and a positive lead V1
Pambrun et al. 7
Maximal Pre-Excitation and Accessory Pathway Location
T A B L E 3 Investigators’ Agreement
Investigator 1 vs. Investigator 1 vs. Investigator 2 vs.
Investigator 3 Investigator 2 Investigator 3
Study Arruda’s Study Arruda’s Study Arruda’s
Algorithm Algorithm Algorithm Algorithm Algorithm Algorithm
Kappa value 0.941 0.612 0.923 0.686 0.917 0.600
p value 0.018 0.038 0.020 0.036 0.021 0.038
polarity when left sided. Although this seminal work
considered the paraseptal region as a single entity,
further studies described a significant shift from
negative polarity, when APs are located in the CS
ostium vicinity (RPS), to positive polarity, when APs
are located in the middle cardiac vein (DCS) or at the
left endocardial surface (LPS) (6,7). Of note, these
findings are consistent with our current results.
However, due to the muscular thickness and
anatomical complexity of the inferior pyramidal
space, paraseptal APs, either with right or left ven-
tricular exit site, can sometimes require an ablation
from both sides and therefore decrease the algo-
rithm’s performance.
S t e p 2 . Inferior leads polarity contributes to APs
localization, most notably proving highly predictive
for RA and LL APs when all 3 are positive. Fitzpatrick
et al. (3) first proposed inferior leads analysis, based
on a sum of Delta-wave polarities at basal state
(ascertained during the first 40 ms and defined as
negative, isoelectric, or positive), but confined it to
septal APs. Fananapazir et al. (11) extended this
approach to all AP locations during maximal pre-
excitation and similarly found that all 3 inferior
leads were positive in most RA and LL APs.
S t e p 3 . Lead V3 polarity, either negative or positive,
seems useful for identification of right ventricular
free wall and septal APs, respectively. As the right
ventricular free wall is located forward of the septum,
its subsequent rearward depolarization may account
for negativity in lead V 3. Consistent with this princi-
ple, studies describing the morphology of maximally
pre-excited QRS complex in septal APs report that a
vast majority of midseptal and parahisian APs display
positivity in lead V 3 (8,9). Of note, as no data on lead
V3 polarity are available for RPS APs during maximal
pre-excitation (6,7), the current study is the first to
point out its relevance in this location.
The V 1 /I ratio ensures a reliable classification in
case of posterior location of left-sided APs. Although
defining 2 different planes, the right-facing lead V1 is
positioned opposite to the left-facing lead I. There-
fore, the 2 leads behave inversely when the origin of
ventricular depolarization shifts from the septal to
the lateral part of the mitral annulus, leading to an
8 Pambrun et al.
Maximal Pre-Excitation and Accessory Pathway Location
F I G U R E 5 Representative Panel of Maximally Pre-Excited 12-Lead ECG Observed in the 9 Locations
The key leads (V1, II, III, aVF, V3, and I) are colored according to the algorithm’s steps, as described in Figure 3. For the DCS example, a typical notch is encircled in lead II.
Abbreviations as in Figure 1.
amplitude increase for lead V 1 and a decrease for lead
I. In our experience, equalization of the 2 leads is
found at 6 o’clock.
S t e p 4 . Lead II notch is more frequent in DCS than in
LPS APs. The former are associated with muscular
strands arising from the CS myocardial coat and epi-
cardially connecting the left atrium with the left
ventricle, further away from the mitral annulus (17).
These sleeve-like extensions, interspersed with adi-
pose and fibrous tissue, can display various orienta-
tions (18). By favoring local
abnormalities, these conditions may lead to inhomo-
geneous ventricular activation, which is recognized
as the underlying mechanism of QRS notching (19). Of
note, despite maximal pre-excitation analysis, local-
ization of epicardial APs still displays the lowest
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018
- 2018:-–-
performance in terms of sensitivity and positive
predictive value.
STUDY LIMITATIONS. The new algorithm displays
some limitations. First, as previously observed, DCS
APs were predominantly sited in the middle cardiac
vein rather than the diverticulum (1 case in the cur-
rent study) (17). Therefore, no firm conclusion can be
drawn regarding the maximally pre-excited QRS
pattern of the latter location. Second, because it is not
designed or tested for multiple APs, its applicability
conduction to this rare condition remains unknown. Finally, as
the oblique course of some AP can lead to substantial
discrepancies between atrial and ventricular insertion
site (20), the latter only, on which maximal pre-
excitation pattern exclusively depends, is targeted
by the new algorithm.
JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018 Pambrun et al. 9
- 2018:-–- Maximal Pre-Excitation and Accessory Pathway Location

F I G U R E 6 Benefit of Using Maximal Pre-Excitation Addressed by 2 Representative Examples

(A) The 12-lead electrocardiogram on the left side of the panel shows basal pre-excitation in sinus rhythm. Following the successive steps of Arruda’s algorithm, the key leads
are expanded to ensure a more detailed analysis of the first 40 ms after the earliest Delta wave’s onset (red point). The 12-lead electrocardiogram on the right side of the
panel shows maximal pre-excitation during atrial pacing. Following the successive steps of our algorithm, the key leads are highlighted with the color code depicted in
Figure 3. (A) Gain in accuracy: the polarity of a minimally pre-excited Delta wave does not prejudge the polarity of a maximally pre-excited QRS complex. In this example, after
the first step, both algorithms conclude to a left-sided accessory pathway (AP). However, as of the second step, we observe a complete reversal of polarity between a positive
Delta wave in aVF (sinus rhythm) and a negative QRS complex in inferior leads (atrial pacing). Of note, as our algorithm is based on the global polarity of the entire QRS
complex during maximal pre-excitation, the polarity of the inferior leads must be considered negative, even if the initial part displays a slightly positive deflexion. Hence, the
polarity of a small Delta wave erroneously predicts an LL AP, while the polarity of entire QRS complex exactly identifies an LPL AP. (B) Gain in reproducibility: a polarity gives
less room for interpretation when defined by the entire QRS morphology rather than the Delta wave’s slope. In this example, both algorithms exactly identify a right
paraseptal (RPS) AP. However, when the Delta wave’s slope varies, the window of analysis is shortened to the first 20 ms, favoring ambiguous interpretation due to the
degree of the Delta wave’s slope or the reference lead chosen to determine the Delta wave’s onset. In step 3, although the global Delta wave’s polarity is positive, the first 20
ms show an isoelectric slope, correcting the false diagnosis of right lateral location. In step 4, although the global Delta wave’s polarity is negative, the first 20 ms show an
isoelectric slope, resulting in a diagnosis of paraseptal location without discriminating between the right and left sides. By contrast, the analysis of the entire QRS polarity
during maximal pre-excitation straightforwardly predicts a RPS location.

CONCLUSIONS pre-excitation can effectively guide endocardial

mapping in daily practice.
The present study demonstrates that the analysis of
maximal pre-excitation allows accurate and repro- ADDRESS FOR CORRESPONDENCE: Dr. Thomas
ducible localization of manifest APs for a vast ma- Pambrun, Hôpital Cardiologique du Haut-Lévêque,
jority of patients manifest APs. Since de facto CHU Bordeaux, Avenue de Magellan, 33604 Pessac
collected when ablation is indicated, maximal Cedex, France. E-mail: thomaspambrun@aol.com.
10 Pambrun et al. JACC: CLINICAL ELECTROPHYSIOLOGY VOL. -, NO. -, 2018
Maximal Pre-Excitation and Accessory Pathway Location - 2018:-–-

PERSPECTIVES

COMPETENCY IN MEDICAL KNOWLEDGE: Maximal maximal pre-excitation is systematically available when

pre-excitation allows more accurate and reproducible
localization of the AP than basal pre-excitation does.
There are 2 main reasons for this. Since exclusively
reflecting ventricular depolarization originating from the
AP, maximally pre-excited QRS obtained during rapid
atrial pacing contains more information relative to its
location. ECG analysis gives less room for interpretation
when applied to the entire QRS rather than confined to
the delta-wave’s onset. Since AP anterograde refractory
period is obtained for induction of AP-related tachycardia
and risk stratification for life-threatening arrhythmias,
ablation is indicated. Therefore, the use of a localization
algorithm based on maximal pre-excitation appears as a
rational approach to start endocardial mapping at the
adequate area.
TRANSLATIONAL OUTLOOK: High-density mapping
during maximal pre-excitation may offer further insight
into the variation of ventricular activation underlying the
modification of QRS morphology observed from one AP
location to another.

REFERENCES

1. Page RL, Joglar JA, Caldwell MA, et al. 2015
ACC/AHA/HRS guideline for the management of
adult patients with supraventricular tachycardia:
executive summary: a report of the American
College of Cardiology/American Heart Association
task force on clinical practice guidelines and the
Heart Rhythm Society. J Am Coll Cardiol 2016;67:
1575–623.
2. Milstein S, Sharma AD, Guiraudon GM, Klein GJ.
An algorithm for the electrocardiographic
localization of accessory pathways in the Wolff-
Parkinson-White syndrome. Pacing Clin Electro-
physiol 1987;10:555–63.
3. Fitzpatrick AP, Gonzales RP, Lesh MD,
Modin GW, Lee RJ, Scheinman MM. New algorithm
for the localization of accessory pathways. J Am
Coll Cardiol 1994;23:107–16.
4. d’Avila A, Brugada J, Skeberis V, Andries E,
Sosa E, Brugada P. A fast and reliable algorithm to
localize accessory pathways based on the polarity
of the QRS complex on the surface ECG during
sinus rhythm. Pacing Clin Electrophysiol 1995;18:
1615–27.
5. Arruda MS, McClelland JH, Wang X, et al.
Development and validation of an ECG algorithm
for identifying accessory pathway ablation site in
Wolff-Parkinson-White syndrome. J Cardiovasc
Electrophysiol 1998;9:2–12.
6. Takahashi A, Shah DC, Jaïs P, Hocini M,
Clementy J, Haïssaguerre M. Specific electrocar-
diographic features of manifest coronary vein
posteroseptal accessory pathways. J Cardiovasc
Electrophysiol 1998;9:1015–25.
7. Kobza R, Hindricks G, Tanner H, et al. Paraseptal
accessory pathway in Wolff-Parkinson-White syn-
drome: ablation from the right, from the left or
within the coronary sinus/middle cardiac vein?
J Interv Card Electrophysiol 2005;12:55–60.
8. Epstein AE, Kirklin JK, Holman WL, Plumb VJ,
Kay GN. Intermediate septal accessory pathways:
electrocardiographic characteristics, electrophysi-
ologic observations and their surgical implications.
J Am Coll Cardiol 1991;17:1570–8.
9. Haïssaguerre M, Marcus F, Poquet F, Gencel L,
Le Métayer P, Clémenty J. Electrocardiographic
characteristics and catheter ablation of parahissian
accessory pathways. Circulation 1994;90:1124–8.
10. Haïssaguerre M, Montserrat P, Warin JF,
Donzeau JP, Le Metayer P, Massiere JP. Catheter
ablation of left posteroseptal accessory pathways
and of long RP’ tachycardias with a right endo-
cardial approach. Eur Heart J 1991;12:845–59.
11. Fananapazir L, German LD, Gallagher JJ,
Lowe JE, Prystowsky EN. Importance of preexcited
QRS morphology during induced atrial fibrillation
to the diagnosis and localization of multiple
accessory pathways. Circulation 1990;81:578–85.
12. Basiouny T, De Chillou C, Fareh S, et al. Accu-
racy and limitations of published algorithms using
the twelve-lead electrocardiogram to localize
overt atrioventricular accessory pathways.
J Cardiovasc Electrophysiol 1999;10:1340–9.
13. Maden O, Balci KG, Selcuk MT, et al. Compar-
ison of the accuracy of 3 algorithms in predicting
accessory pathways among adult Wolff-
Parkinson-White syndrome patients. J Interv Card
Electrophysiol 2015;44:213–9.
14. Alvarez Lόpez M, García Alberola A,
Tercedor L, et al. Comparative study of 3 algo-
rithms to localize the accessory pathway in Wolff-
Parkinson-White syndrome. Rev Esp Cardiol 1996;
49:589–97.
15. Katsouras CS, Greakas GF, Goudevenos JA,
et al. Localization of accessory pathways by the
electrocardiogram: which is the degree of accor-
dance of 3 algorithms in use? Pacing Clin Elec-
trophysiol 2004;27:189–93.
16. Haïssaguerre M, Dartigues JF, Warin JF, Le
Metayer P, Salamon R. Electrogram patterns pre-
dictive of successful catheter ablation of accessory
pathways. Value of unipolar recording mode. Cir-
culation 1991;84:188–202.
17. Sun Y, Arruda M, Otomo K, et al. Coronary
sinus-ventricular accessory connections producing
posteroseptal and left posterior accessory path-
ways: incidence and electrophysiological identifi-
cation. Circulation 2002;106:1362–7.
18. von Lüdinghausen M, Ohmachi N, Boot C.
Myocardial coverage of the coronary sinus and
related veins. Clin Anat 1992;5:1–15.
19. Flowers NC, Horan LG, Thomas JR,
Tolleson WJ. The anatomic basis for high-
frequency components in the electrocardiogram.
Circulation 1969;39:531–9.
20. Otomo K, Gonzalez MD, Beckman KJ, et al.
Reversing the direction of paced ventricular and
atrial wavefronts reveals an oblique course in
accessory AV pathways and improves localization
for catheter ablation. Circulation 2001;104:550–6.
KEY WORDS accessory pathway,
algorithm, catheter ablation, maximal pre-
excitation
A PPE NDI X For supplemental tables, please
see the online version of this paper.