Slow Conducting Electroanatomic Isthmuses: An Important ...electrophysiology.onlinejacc.org/content/jcep/early/2018/03/22/j.jace… · grams (

J A C C : C L I N I C A L E L E C T R O P H Y S I O L O G Y VO L . - , N O . - , 2 0 1 8

ª 2 0 1 8 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O UN DA T I O N

P U B L I S H E D B Y E L S E V I E R

Slow Conducting ElectroanatomicIsthmuses: An Important LinkBetween QRS Duration and VTin Tetralogy of Fallot
Gijsbert F.L. Kapel, MD,a Charlotte Brouwer, MD,a Zakaria Jalal, MD,b,c Frédéric Sacher, MD, PHD,b
Jeroen Venlet, MD,a Martin J. Schalij, MD, PHD,a Jean-Benoît Thambo, MD, PHD,c Monique R.M. Jongbloed, MD, PHD,a

Nico A. Blom, MD, PHD,a Marta de Riva, MD,a Katja Zeppenfeld, MD, PHDa

ABSTRACT

ISS

Fro

Mo

dio

res

He

rep

All

ins

vis

Ma

OBJECTIVES This study sought to evaluate the influence of slow conducting anatomic isthmuses (SCAI) as dominant

ventricular tachycardia (VT) substrate on QRS duration.

BACKGROUND QRS prolongation has been associated with VT in repaired tetralogy of Fallot.

METHODS Seventy-eight repaired tetralogy of Fallot patients (age 37 � 15 years, 52 male, QRS duration 153 � 29 ms,

67 right bundle branch blocks [RBBB]) underwent programmed stimulation and electroanatomic activation mapping

during sinus rhythm. Right ventricular (RV) surface, RV activation pattern, RV activation time, conduction velocity at AI,

and remote RV sites were determined.

RESULTS Twenty-four patients were inducible for VT (VTþ); SCAI was present in 22 of 24 VTþ but only in 2 of 54

patients without inducible VT (VT�). Conduction velocity through AI was slower in VTþ patients (median of 0.3 (0.3–0.4)

vs. 0.7 (0.6–0.9) m/s; p < 0.01) but conduction velocity in the remote RV did not differ between groups. In non-RBBB,

QRS duration was similar in VTþ patients (n ¼ 6) and VT� patients (n ¼ 5), but RV activation within SCAI exceeded QRS

offset in VTþ patients (37 � 20 ms vs. –5 � 9 ms, p < 0.01). In RBBB, both QRS duration and RV activation time were

longer in VTþ patients (n ¼ 18, 17 of 18 QRS > 150 ms) compared with VT� patients (n ¼ 49, 27 of 49 QRS > 150 ms)

(173 � 22 ms vs. 156 � 20 ms; p < 0.01; 141 � 22 ms vs. 129 � 21 ms; p ¼ 0.04). In VTþ patients, QRS prolongation

>150 ms (n ¼ 17) was due to SCAI or blocked isthmus in 15 patients (88%) and 1 (6%). In contrast, in VT� patients, QRS

prolongation >150 ms (n ¼ 27) was due to enlarged RV or blocked isthmus in 10 patients (37%) and 8 (30%), but due to

SCAI in only 1 (4%). After exclusion of a severely enlarged RV, a QRS duration >150 ms was highly predictive for SCAI/

blocked AI (OR: 17; 95% CI: 3.3 to 84; p < 0.01).

CONCLUSIONS A narrow QRS interval does not exclude VT-related SCAI. In the presence of RBBB, SCAI further

prolongs QRS duration. QRS duration >150 ms is highly suspicious for SCAI or isthmus block distinguishable by elec-

troanatomic mapping. (J Am Coll Cardiol EP 2018;-:-–-) © 2018 by the American College of Cardiology Foundation.

N 2405-500X/$36.00 https://doi.org/10.1016/j.jacep.2018.02.002

m the aDepartment of Cardiology, Leiden University Medical Centre, Leiden, the Netherlands; bElectrophysiology and Heart

deling Institute, Bordeaux University Hospital, Bordeaux, France; and cDepartment of Paediatric and Adult Congenital Car-

logy, Bordeaux University Hospital, Bordeaux, France. The Department of Cardiology Leiden has received unrestricted

earch and fellowship grants from Edward Lifesciences, Boston Scientific, Medtronic and Biotronik. The Electrophysiology and

art Modeling Institute has received financial support from the French government (grant ANR-10-IAHU-04). The authors have

orted that they have no relationships relevant to the contents of this paper to disclose.

authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’

titutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information,

it the JACC: Clinical Electrophysiology author instructions page.

nuscript received November 2, 2017; revised manuscript received January 25, 2018, accepted February 1, 2018.

https://doi.org/10.1016/j.jacep.2018.02.002

http://www.electrophysiology.onlinejacc.org/content/instructions-authors

ABBR EV I A T I ON S

AND ACRONYMS

AI = anatomic isthmus

CI = confidence interval

CMR = cardiac magnetic

resonance

CV = conduction velocity

EAM = electroanatomic map

ECG = electrocardiogram

IQR = interquartile range

LV = left ventricle

MVT = monomorphic

ventricular tachycardia

OR = odds ratio

PV = pulmonary valve

PES = programmed electrical

stimulation

RBBB = right bundle branch

block

rTOF = repaired tetralogy of

Fallot

RV = right ventricle

RVAT = total right ventricular

activation time

RVOT = right ventricular

outflow tract

SAECG = signal-averaged

electrocardiography

SCAI = slow conducting

anatomic isthmus

SCD = sudden cardiac death

SR = sinus rhythm

TA = tricuspid annulus

VA = ventricular arrhythmia

VSD = ventricular septal defect

VT = ventricular tachycardia

Kapel et al. J A C C : C L I N I C A L E L E C T R O P H Y S I O L O G Y V O L . - , N O . - , 2 0 1 8

QRS Duration and Risk for VT in Tetralogy of Fallot - 2 0 1 8 :- –-

2

A s a result of earlier repair andevolving surgical techniques, themajority of patients with repaired

tetralogy of Fallot (rTOF) survive to adult-hood (1–3), but they may remain at risk ofsudden cardiac death (SCD) due to ventricu-lar arrhythmias (VA) (4,5). Monomorphicventricular tachycardia (VT) is the most com-mon arrhythmia subtype affecting >14% ofpatients and accounting for >80% of appro-priate implantable cardioverter-defibrillatortherapies (6,7). Over the last decades, effortshave been made to noninvasively identifyrTOF patients at risk for VT. Although no sin-gle risk factor has shown sufficient indepen-dent predictive value for the presence of anarrhythmic substrate in an individual pa-tient, a prolonged QRS duration >180 mshas consistently been associated with VTand SCD (8–10). In contemporary cohorts, alower QRS cutoff value to identify a patientat risk has been suggested and attributed toadvances in surgical approaches (11,12). Themodern transatrial-transpulmonary surgicalapproach with smaller transannular patchesmay not only prevent severe pulmonaryvalve regurgitation and right ventricular(RV) dilatation, but also right bundle branchdamage. Accordingly, contemporary rTOFpatients may have less RV dilatation,different RV activation patterns, and even anarrow post-operative QRS interval (13).QRS prolongation was initially thought to bethe combined effect of a surgically createdright bundle branch block (RBBB) and pro-gressive, RV dilatation with global conduc-tion delay, the latter creating a substrate for

VT (9). However, invasive electroanatomic mappingstudies have demonstrated that the dominant sub-strate for VT are slow conducting anatomic isthmuses(SCAI) confined to the right ventricular outflow tract(RVOT) (14). We hypothesize that the interplay be-tween a post-operative RBBB and conduction delayin SCAI determines the RV activation pattern andQRS prolongation. The objective of this study is toevaluate, with the use of invasive electroanatomicmapping, whether QRS prolongation is due to local-ized conduction delay across SCAI, rather than dueto global conduction delay.

METHODS

PATIENT SELECTION. The cohort consisted of 83consecutive rTOF patients with either documented

sustained VT, or considered at risk for VT and/or withindication for reoperation who underwent electro-physiological evaluation and RV electroanatomicmapping (EAM) between 2005 and 2013 at LeidenUniversity Medical Center (n ¼ 53) and BordeauxUniversity Hospital (n ¼ 30). Patients were consideredat risk for VT if $1 of the following risk factors waspresent: syncope, nonsustained VT on Holtermonitor, QRS duration $180 ms, late repair ($5years), at least moderately depressed RV or left ven-tricular (LV) function or presence of a transannularpatch (9,10,15–19). The Dutch Central Committee onHuman-Related Research permits use of anonymousdata without prior approval of an institutional reviewboard, if the data are obtained for patient care and ifthe data do not contain identifiers that could betraced back to the individual patient. All patientswere treated according to our standard clinical pro-tocol and provided informed consent (14).

BASELINE EVALUATION. Patient records werereviewed for date and type of repair, device implan-tation, and documented VA. Nonpaced 12-lead elec-trocardiograms (ECG) recorded at 25 mm/s wereassessed for QRS duration and intraventricular con-duction disturbances. In 53 of 83 patients, QRSduration was also measured using LEADS (LeidenECG Analysis and Decomposition Software) (LeidenUniversity Medical Center, Leiden, the Netherlands)(see the Online Methods) (20). Holter recordings werereviewed for nonsustained VT. RV and LV cardiacfunction was assessed by cardiac magnetic resonance(CMR) and/or echocardiography. A mildly reduced orgood RV function (RV ejection fraction >40%) with atricuspid plane systolic excursion $14 mm and LVejection fraction $40% were classified as preservedRV and LV function, respectively (21). Cardiac vol-umes were determined by CMR and indexed for bodysurface area. RV end-diastolic volume index $180 ml/m2

was considered severely enlarged (22).

EAM AND ELECTROPHYSIOLOGICAL EVALUATION.

Programmed electrical stimulation (PES) was per-formed (see the Online Methods). Sustained VT wasdefined as lasting $30 s or causing hemodynamiccompromise requiring termination. A detailed3-dimensional EAM of the RV was constructed duringsinus rhythm (SR) using a nonfluoroscopic mappingsystem (Carto3 and CartoXP, Biosense Webster,Irvine, California). The endocardial RV surface wasmeasured using dedicated tools of the Carto system.EAM RV surface was compared with CMR-derivedvolume. All mapping points were reviewed offlinefor correct annotation of the local activation time,defined as the sharp bipolar electrogram coinciding

J A C C : C L I N I C A L E L E C T R O P H Y S I O L O G Y V O L . - , N O . - , 2 0 1 8 Kapel et al.- 2 0 1 8 :- –- QRS Duration and Risk for VT in Tetralogy of Fallot

3

with the fast down-stroke of the unipolar electro-gram. All activation maps were displayed asisochronal maps and as propagation maps (OnlineVideos 1 and 2) to determine RV activation pattern.The site(s) of earliest and latest RV activation wereidentified and the total RV activation time wascalculated. Earliest onset and latest offset of the QRScomplex in any lead were annotated on Carto toassess QRS duration. QRS duration, total right ven-tricular activation time (RVAT), and intervals be-tween onset of QRS and onset of RV activation, andbetween offset of QRS and offset of RV activationwere measured using electronic calipers (sweepspeed 200 mm/s) (Figure 1).

AI CHARACTERISTICS AND ITS RELATION TO VT.

The identification of anatomic isthmuses (AI) in rTOFhas been described in detail in the OnlineMethods (23). Isthmus 1 is located between thetricuspid annulus (TA) and the RVOT patch/RV inci-sion; isthmus 2 between the RV incision and thepulmonary valve (PV). Isthmuses 3 and 4 are locatedat the infundibular septum, with isthmus 3 betweenPV and the ventricular septal defect (VSD) patch, andisthmus 4 between the VSD patch and TA (23). Theelectroanatomic characteristics of each identified AIwere assessed. The fill threshold for EAM of the AIwas set at #10 mm and care was taken to delineatethe length of the isthmus by recording the firstnormal voltage electrograms at isthmus entrance andexit sites (14). Isthmus length is the distance betweenentrance and exit sites and the conduction timethrough the isthmus is the difference between thelocal activation time at isthmus entrance and exit.The isthmus conduction velocity (CV) was calculated(isthmus length/conduction time). CV was measuredin the direction of the RV propagation wave frontduring SR. In selected patients, we have performeddifferential pacing close to the AI to differentiatebetween slow conduction and block. SCAI wasdefined as isthmus with low bipolar voltage electro-grams (<1.5 mV) and a CV <0.50 m/s (14). Based onthe first quartile of CV distribution in normal controlsubjects, a CV $0.50 m/s and #0.63 m/s was consid-ered borderline CV (14). An AI was classified as havinga pre-existing conduction block if either no excitabletissue between its anatomic boundaries could beidentified or double potentials could be recordedalong the AI and in addition the RV activationsequence during SR was consistent with propagationtoward the isthmus from both sites. Septal block wasdefined as a pre-existing conduction block at isthmus3 in case of perimembranous VSD (absence of isthmus4) or as a pre-existing conduction block at both

isthmus 3 and 4 in case of muscular VSD. Whether aninduced VT was dependent on a SCAI was determinedby pace-mapping ($11/12 pace match) and/or con-cealed entrainment and/or diastolic activation duringVT with VT termination by radiofrequency delivery(23). To compare CV within and remote from AI, theCV was calculated for 3 additional RV areas: RVseptum; RV free wall; and low RVOT free wall (see theOnline Methods).

STATISTICAL ANALYSIS. Continuous data are pre-sented as mean � SD or median (interquartile range[IQR]) according to distribution. Categorical data arereported as percentages or frequencies. The differ-ence in QRS duration measured with LEADS on 12-lead ECG and using the Carto system was assessedby paired Student t test. Patients with and without VTwere compared using IQR, Mann-Whitney U, and chi-square tests where appropriate, including subgroups.Associations among QRS duration, RV volumes onCMR, and RV area on EAM were assessed by Pearsoncorrelation. The relation between: 1) QRS durationand RVAT; and 2) QRS duration and RV activationafter QRS offset was illustrated with a scatterplot forall patients and patients with and without RBBB, dotswere marked for VT inducibility. QRS duration ofRBBB patients were illustrated with an aligned scat-terplot, according to VT profile. In addition, in thesepatients, the unadjusted odds ratio (OR) of QRSduration of >150 ms was calculated for AI propertiesusing logistic regression. SPSS (version 22.0 for Win-dows; IBM, Armonk, New York) was used. A value ofp < 0.05 was considered significant.

RESULTS

PATIENT POPULATION. Eighty-three patients un-derwent electrophysiological evaluation. Five pa-tients with an incomplete RV map or maps acquiredduring ventricular pacing were excluded. One patienthad undergone a prior VT ablation in another center.Subsequently, the study population consisted of 78patients (age 37 � 15 years, 67% male) who wererepaired at a median age of 5.0 years (IQR: 2.1 to 8.5years ). Eleven patients (14%) had documented sus-tained VT. QRS duration was 153 � 29 ms. Sixty-sevenpatients (86%) had a broad QRS interval (>120 ms), allRBBB configuration, and 11 patients had a narrow QRSinterval. Thirteen patients (17%) had QRSduration $180 ms. All but 2 patients had a preservedLV function and 47 (60%) a preserved RV function.Baseline characteristics are provided in Table 1.

EAM AND ELECTROPHYSIOLOGICAL EVALUATION. RVactivation time during SR was 130 � 24 ms. QRS

http://jacccep.acc.org/video/2018/0519_VID1.avi



FIGURE 1 Activation Time Measurements

(A) Schematic of QRS and right ventricular (RV) activation time measurements. (B) RV electroanatomic map of a patient (right bundle branch

block, QRS duration: 132 ms) in 3 views displayed as bipolar (Bi) voltage map (left) and activation map (right). Voltage and local activation time

(LAT) are color-coded. Three anatomic isthmuses are shown: isthmus 1 (between tricuspid annulus [TA] and RV incision); isthmus 2 (between RV

incision and pulmonary valve [PV]); and isthmus 3 (between PV and ventricular septal defect patch). Total RV activation time is 108 ms with

earliest RV activation at the septum after QRS onset and latest activation at the lateral TA coinciding with QRS offset as demonstrated in the

lower schematic.



4

TABLE 1 Baseline Characteristics (N ¼ 78)

Age, yrs 37 � 15

Male 52 (67)

Total repair age, yrs 5.0 (2.1–8.5)

Total repair age $5 yrs 40 (51)

Transannular patch 29/67* (43)

Time after repair, yrs 30 � 11

PVR, yes 16 (21)

PVR age, yrs 31 � 13

Syncope 5 (6)

QRS duration, ms 153 � 29

QRS duration $180 ms 13 (17)

Right bundle branch block 67 (86)

Nonsustained VT on Holter monitor 16/66* (24)

Class III antiarrhythmic drugs 2 (3)

LV function preserved 76 (97)

RV function preserved 47 (60)

RVEDV, ml 242 � 73

RVEDV index, ml/m2 136 � 39

RVEDV $180 ml/m2 8/68 (12)

RVESV, ml 138 � 54

RVESV indexed, ml/m2 78 � 30

Cardiac function preserved 46 (59)

Values aremean� SD, n (%), n/N (%), ormedian (interquartile range). *Reports available.

LV ¼ left ventricular; PVR ¼ pulmonary valve replacement; RV ¼ right ven-tricular; RVEDV ¼ right ventricular end-diastolic-volume; RVESV ¼ right ventric-ular end-systolic-volume; VT ¼ ventricular tachycardia.

TABLE 2 QRS Duration and RV Activation Times According to Presence of VT

and RBBB

VT� VTþ p Value

All patients 54 24

QRS duration, ms 151 � 25 157 � 35 0.411

RV activation time, ms 125 � 23 142 � 23 0.005

Interval: QRS onset to RV onset, ms 23 � 13 21 � 17 0.713

Interval: QRS offset to RV offset, ms �3 � 8 6 � 21 0.040

Non-RBBB 5 6

QRS duration, ms 101 � 11 108 � 11 0.429



Interval: QRS offset to RV offset, ms �5 � 9 37 � 20 0.009

RBBB 49 18

QRS duration, ms 156 � 20 173 � 22 0.004



Interval: QRS offset to RV offset, ms �3 � 9 �4 � 4 0.524

Values are n or mean � SD.

RBBB ¼ right bundle branch block; VT� ¼ not inducible for ventricular tachycardia; VTþ ¼inducible for ventricular tachycardia; other abbreviations as in Table 1.


5

duration did not differ when measured on pre-procedural 12-lead ECG or on Carto (OR: �1.8 ms;95% confidence interval [CI]: –4.0 to 0.4 ms; p ¼ 0.11).The electroanatomic RV surface was 249 � 66 cm2 andshowed a significant correlation with the RVend-diastolic volume derived from CMR (r ¼ 0.68,p < 0.001) (Online Figure 1). A RV end-diastolicvolume index of >180 ml/m2 (severely dilated RV)corresponded to a RV surface of >293 cm2 observed in15 patients (19%). For the entire population, neitherthe RV surface nor the RV volume correlated to QRSduration (r ¼ 0.14, p ¼ 0.22; r ¼ 0.07, p ¼ 0.55,respectively) (Online Figure 2).

AI CHARACTERISTICS AND ITS RELATION TO VT. Allpatients had at least 1 AI and 24 (31%) at least 1 SCAI,which was SCAI 3 in the majority (21 of 24). One of the21 patients with SCAI 3 had normal conduction throughisthmus 4, compensating for septal conduction delay,and 1 patient had only septal SCAI 4. Accordingly, 21of 24 patients with any SCAI showed slow conductionat the infundibular septum. In 9 patients, a pre-existingblock of an AI was observed, which was septal isthmus3 in all. Thirteen patients had at least 1 AI withborderline CV, which was also isthmus 3 in all.

Twenty-four patients (31%) (including all 11 pa-tients with spontaneous VT) were inducible for atotal of 34 VT (median VT cycle length: 247 ms;

IQR: 230 to 287 ms). In 22 patients (92%) with 32 VT,SCAI was the substrate for all clinical and induced VT(see the Online Results), which was a septal SCAI (3 or4) in 19 of 22 patients (86%). There was no differencein AI CV between the 11 patients with spontaneousand inducible VT (median: 0.27 m/s; IQR: 0.20 to 0.32m/s) and the 13 patients with only inducible VT(median: 0.34 m/s; IQR: 0.29 to 0.42 m/s; p ¼ 0.16). Incontrast, SCAI was found in only 2 of 54 patients (4%)without documented or induced VT. Slow conductionwas not observed remotely from the AI. The electro-anatomic CV for the RV septum, RV free wall, andRVOT free wall were 1.3 � 0.3 m/s, 1.3 � 0.3 m/s, and0.9 � 0.2 m/s, respectively, and did not differ be-tween patients with or without VT. Fourteen of the 22patients with SCAI-related VT underwent successfulablation, defined as noninducibility of any VT andtransection of the SCAI. Two patients had no VT, butdid have SCAI 3 and underwent successful ablationdefined as bidirectional block of the SCAI. All 68 pa-tients without SCAI (52 at baseline, 16 after ablation)remained free from VT during follow-up (39 � 24months). Detailed acute and long-term ablationoutcome are added to the Online Results.

QRS DURATION, RV ACTIVATION TIME AND

PATTERN ACCORDING TO VT INDUCIBILITY. For theentire population, QRS duration did not differsignificantly between patients with and without VT(157 � 35 ms vs. 151 � 25 ms, p ¼ 0.411). However, theRVAT was significantly longer in patients with VT(142 � 23 ms vs. 125 � 23 ms in non-VT patients,p ¼ 0.005) (Table 2).

FIGURE 2 Relations Between QRS Duration and RV Activation Time

(A to C) Scatterplots show the relations between QRS duration and RV activation time for all patients (A), patients without right bundle branch

block (RBBB) (B), and with RBBB (C) according to ventricular tachycardia (VT) inducibility (gray dots not inducible [VT–], black dots inducible

for VT [VTþ]). (D to F) Scatterplots show relations between QRS duration and right ventricular (RV) activation time after QRS offset for all

patients (D), patients without RBBB (E), and with RBBB (F).



6

A subgroup analysis was performed according topresence or absence of RBBB:

1. In patients with narrow QRS interval (non-RBBB),QRS duration was similar in patients with andwithout VT (108 � 11 ms and 101 � 11 ms, p ¼0.429), whereas the RVAT was longer in patientswith VT (143 � 30 ms vs. 91 � 12 ms in patientswithout VT, p ¼ 0.017). In all patients, the earliestRV activation during SR was observed simulta-neously at the distal septum and at the anterior RVfree wall followed by rapid propagation of activa-tion wave fronts toward the lateral and septalRVOT. The onset of RV activation coincided or wasclose to the QRS onset (3 � 12) in patients with andwithout VT (p ¼ 0.931). However, QRS offset andRV activation offset coincided only in thosewithout VT. In non-RBBB patients with VT, thelatest RV activation exceeded QRS offset by 37 � 20ms (Figures 2D and 2E, Online Figure 3) and waslocated at the exit site of the VT-related SCAI in allpatients. These small late activated areas with lowbipolar voltages did not contribute to the 12-lead

surface ECG as demonstrated in Figure 3. RV sizeand RV CV remote from the AI was similar betweennon-RBBB patients with and without VT (Table 3).

2. In patients with RBBB, both QRS duration andRVAT were longer in patients with VT as comparedto patients without VT (173 � 22 and 141 � 22 vs.156 � 20 and 129 � 21, p ¼ 0.004 and p ¼ 0.036,respectively). Of interest, only 6 of 18 with VT hada QRS duration $180 ms, but all but 1 had a QRSduration >150 ms. In contrast, only 27 (55%) of the49 RBBB patients without VT had QRS duration>150 ms. In all RBBB patients, earliest RV activa-tion was observed in a broad area at the mid-to-basal RV septum, but not at the RV free wall. Inpatients without VT and normal conducting AI,activation propagated in different wave fronts:from the RV septum through the septal AI and AI 1,across the inferior wall and across the RV free wall,with fusion of wave fronts typically at the basallateral TA being the latest RV activated area(Figure 4, Online Video 1). In patients with VT andseptal activation delay or delay in AI 1, the latestactivated RV area shifted toward the lateral RVOT


FIGURE 3 Patient Without RBBB, But With VT

RV electroanatomic maps obtained during sinus rhythm of a 35-year-old female, who presented with near collapse, inducible for a fast VT (VT

cycle length: 230 ms), modified posterior view. (A) Voltage map color-coded for bipolar voltage. (B) Activation map color-coded for activation

time according to bar. QRS duration was 109 ms (on/offset: �43/66 ms), the total RV activation time was 147 ms (on/offset: �41/106 ms). RV

activation exceeds QRS offset (purple area on right map, not visible on electrocardiogram) by 40 ms, the latest activated area was located in

the low voltage area between the PV and the ventricular septal defect (VSD) patch (i.e., isthmus 3). There was capture at each site across

isthmus 3 and fusion at the exit region of isthmus 3, suggesting slow conduction through isthmus 3. In addition, a slow conducting or blocked

isthmus can be furthered differentiated by differential pacing close to the isthmus. At this site (*) of isthmus 3 the critical isthmus of the

induced VT was identified. Abbreviations as in Figures 1 and 2.


7

(Figure 4, Online Video 2). The prolongation of thepathway for total RV activation resulted both inincrease in RVAT and QRS duration. Of note, thesame activation pattern and prolongation of RVactivation and QRS duration was observed in 8patients without VT and a pre-existing septalisthmus block (excluding the septal isthmus assubstrate for VT). In RBBB patients, the onset ofthe RV activation was recorded 26 � 12 ms after theonset of the QRS both in patients with and withoutVT (p ¼ 0.305). Furthermore, QRS offset and RVactivation offset coincided (�3 � 8 ms, p ¼ 0.524)in patients with and without VT (Figure 2F andOnline Figure 3). As in patients with narrow QRS,CV of the RV remote from AI and the RV endocar-dial surface did not differ between patients withand without VT (Table 3).

INTERPLAY AMONG QRS DURATION, RV SIZE, SCAI,

AND VT IN PATIENTS WITH RBBB. Of the 18 patientswith RBBB and VT, 17 had a QRS duration >150 ms,

16 had SCAI, but only 4 had severely dilated RV(Figure 5). Accordingly, in patients with RBBB and VT,SCAI was the most prevalent mechanism for QRSprolongation. The only patient with RBBB, VT, andQRS duration <150 ms had SCAI 3–related VT. How-ever, the second fast conducting septal isthmus 4prevented septal activation delay and QRSprolongation.

In contrast, of the 49 RBBB patients without VT,only 27 had QRS duration >150 ms. Of these 27 pa-tients, the majority (n ¼ 16; 59%) had a severelyenlarged RV without SCAI (n ¼ 10; 37%) and/or a pre-existing blocked septal isthmus (n ¼ 8; 30%) asmechanisms for QRS prolongation (Figure 5). Incontrast, none of the 22 patients without VT and QRSduration #150 ms had either a severely dilated RV or ablocked septal isthmus. These data suggest that incontemporary rTOF, a QRS duration between120 and150 ms is solely due to the post-operative RBBB.

The presence of RBBB with QRS duration >150 mswas strongly associated with the presence of a slow


TABLE 3 RV Size, CV, and AI According to Presence of VT and RBBB

VT� VTþ p Value

Non-RBBB 5 6

RV size, cm2 237 � 48 247 � 63 0.785

CV remote of AI

RVOT free wall, m/s 1.1 � 0.5 1.0 � 0.2 0.455

RV free wall, m/s 1.4 � 0.4 1.4 � 0.2 0.800

RV septal, m/s 1.3 � 0.4 1.3 � 0.1 0.929

Overall, m/s 1.3 � 0.3 1.2 � 0.1 0.568

AI

SCAI present, yes 0 (0) 6 (100) 0.001

SCAI, n 0 (0–0) 2 (1–3) 0.004

SCAI, isthmus 1, yes 0 (0) 3 (50) 0.064

SCAI, isthmus 2, yes 0 (0) 3 (50) 0.064

SCAI, isthmus 3, yes 0 (0) 5 (83) 0.006

SCAI, isthmus 4, yes 0 (0) 0 (0)

Longest AI conduction time, ms 17 � 11 73 � 17 <0.001

Slowest AI CV, m/s 0.8 (0.7–0.9) 0.3 (0.2–0.4) 0.004

Blocked septal AI 0 (0) 0 (0)

RBBB 49 18

RV size, cm2 254 � 64 241 � 79 0.489

CV remote of AI

RVOT free wall, m/s 0.9 � 0.2 0.9 � 0.3 0.997

RV free wall, m/s 1.3 � 0.3 1.3 � 0.2 0.639

RV septal, m/s 1.3 � 0.3 1.2 � 0.3 0.228

Overall, m/s 1.2 � 0.2 1.1 � 0.1 0.307

AI

SCAI present, yes 2 (4) 16 (89) <0.001

SCAI, n 0 (0–0) 1 (1–1) <0.001

SCAI, isthmus 1, yes 0 (0) 3 (17) 0.003

SCAI, isthmus 2, yes 0 (0) 1 (6) 0.096

SCAI, isthmus 3, yes 2 (4) 14 (78) <0.001

SCAI, isthmus 4, yes 0 (0) 2 (11) 0.018

Longest AI conduction time, ms 22 � 12 65 � 29 <0.001

Slowest AI CV, m/s 0.7 (0.6–0.9) 0.3 (0.2–0.4) <0.001

Blocked septal AI 8 (16) 1 (6) 0.252

Values are n, n (%), mean � SD, or median (interquartile range).

AI ¼ anatomic isthmus; CV ¼ conduction velocity; RVOT ¼ RV outflow tract; SCAI ¼ slowconducting anatomic isthmus; other abbreviations as in Tables 1 and 2.



8

conducting or blocked isthmus (OR: 13.8; 95% CI: 2.9to 66.2; p ¼ 0.001). After exclusion of patients withseverely enlarged RV (n ¼ 14), the OR of a QRSduration >150 ms for slow conducting or blocked AIwas 16.6 (95% CI: 3.3 to 84; p ¼ 0.001).

DISCUSSION

To the best of our knowledge, this is the first studyusing detailed 3-dimensional RV EAM to assess therelation among QRS duration, RV size, and RV acti-vation during SR and the substrate for re-entry VT inpatients with rTOF. We could demonstrate that rTOFpatients with a narrow QRS can be at risk for life-threatening VT related to SCAI. Rapid activationthrough the distal specific conduction system with

fusion of wave fronts at the infundibulum can maskslow conduction in septal AI. In these patients,the surface ECG does not contribute to substrateidentification. In contrast, in rTOF patients with apre-existing RBBB at any level, VT-related SCAIfurther prolong the pathway for total RV activationresulting in QRS duration >150 ms. A similar QRSprolongation in rTOF patients with RBBB but withoutSCAI can be observed in those with a severely dilatedRV or with pre-existing septal isthmus block. Afterexclusion of severe RV dilatation, QRS duration >150ms is highly predictive for conduction delay at theinfundibular septum as potential substrate for VT. Tofurther distinguish between conduction delay withinan AI as substrate for VT and conduction blockrequires EAM.

VA AND SCD IN rTOF. Monomorphic ventriculartachycardia (MVT) is the most commonly docu-mented arrhythmia subtype in rTOF (6,7,14).Considering the typically short cycle length, MVTmay lead to SCD even if the biventricular function ispreserved (4,5,24). EAM studies have demonstratedthat the dominant underlying mechanism of MVT ismacro-re-entry using SCAI located in the RVOT(14,23). In the present cohort, SCAI was the substratefor all clinical and induced VT in 22 of 24 patients.

QRS PROLONGATION AS RISK FACTOR FOR VT AND

SCD. A prolonged QRS duration has consistently beenassociated with both MVT and SCD in rTOF (8–10). Ina large series of 182 patients, all 13 patients with(near-missed) SCD had QRS $180 ms, which wasobserved in only 5.3% of patients without an event.Of importance, all 9 SCD survivors underwent PESand were inducible for monomorphic fast re-entryVT, suggesting that MVT had caused the event andthat QRS prolongation may predict the substrate forre-entry VT (9). PES performed in 134 unselectedrTOF patients confirmed the association between QRSduration $180 ms and MVT. A QRS duration $180 mswas 35% sensitive and 97% specific for inducibility ofany MVT and 100% sensitive and 96% specific forinducibility of the clinical VT (8). A subsequentmulticenter study demonstrated that QRSduration $180 ms was the strongest independentpredictor for VT with a remarkable hazard ratio of41.9 (95% CI: 14.7 to 119.4) (10). In more recent co-horts, the reported sensitivity of QRS duration $180ms for death or VT was much lower (28% to 53%) anda cutoff value of 170 ms has been suggested (sensi-tivity 53%, specificity 86%) (11,12). Similarly, in ourcohort, QRS duration $180 ms was observed in only 6of 24 patients (25%) with VT and in 7 of 54 (13%)without VT.

FIGURE 4 RV Activation in Patients With RBBB According to Absence/Presence of SCAI

(A1 and A2) Twelve-lead electrocardiograms (ECG) of a patient without VT (A1) and a patient with VT (A2). (B1 and B2) Electroanatomic reconstruction of

the RV (modified posterior view) displayed as color-coded voltage map (left, purple indicates normal bipolar voltage) and activation map (right, red in-

dicates early activation, purple late activation). Anatomic isthmus 3 between PV and VSD patch (gray dots) is marked with white brackets. (B1) RV

activation map demonstrates RV activation through a normal conducting anatomic isthmus 3 with fusion of the propagation wave fronts at the lateral TA.

(B2) RV activation map demonstrates prolonged RV activation due the localized conduction delay at anatomic isthmus 3 (i.e., slow conducting anatomic

isthmus [SCAI]), causing a shift in the fusion site of activation wave fronts toward the right ventricular outflow tract (RVOT). (C1, C2, and D) Schematics

illustrate the RV activation in presence of normal conducting anatomic isthmus 3 (C1) and SCAI 3 (C2) causing a shift in RV fusion site (D) and QRS pro-

longation. See online videos 1 and 2. QRSd ¼ QRS duration; SR ¼ sinus rhythm; other abbreviations as in Figures 1 and 2.


9



FIGURE 5 QRS Duration According to VT Inducibility and Underlying Mechanisms for QRS Prolongation

Aligned dot plot, median and interquartile ranges are marked with horizontal lines. *Other VT substrates; **Patient with double septal isthmus

(isthmus 3 with slow conduction and isthmus 4 with normal conduction velocity). Abbreviations as in Figures 1 and 2.



10

A modern transpulmonary approach may prevent apost-operative RBBB. As a consequence, an importantnumber of contemporary rTOF patients have a narrowQRS interval without RBBB, which may influence thepredictive value of a prolonged QRS for VT/SCD ifpatients with and without RBBB are not separated.We could only demonstrate a significant difference inQRS duration between patients with and without VTafter excluding the 11 rTOF patients (14%) with anarrow QRS interval. Of importance, the use ofsmaller transannular patches may reduce PV regur-gitation and RV dilatation. As a result, contemporarypatients, with post-operative RBBB may still have lessadditional QRS prolongation, as the delayed myocar-dial activation of a severely dilated RV is avoided.Accordingly, in contemporary rTOF patients with apost-operative RBBB, the optimal QRS cutoff for riskstratification needs to be redefined.

RV ACTIVATION IN rTOF PATIENTS WITH NARROW

QRS INTERVAL. All patients with narrow QRS inter-val showed early simultaneous RV activation at theanterior wall and the septum via distal ramificationsof the conduction system with further propagationtoward the anterolateral RVOT and the infundibularseptum (25). In the 6 patients with narrow QRS in-terval and isthmus-related VT, fusion of activation

wave fronts during SR was typically observed at theinfundibular septum. Late activation after QRS offsetwas restricted to a small area with low bipolar volt-ages, therefore not contributing to the surface QRS.Such late potentials are also often observed in pa-tients with scar-related VT of other etiologies. Ofimportance, in our series, a narrow QRS interval didnot exclude a substrate for life-threatening VT.Although not performed in this study, signal-averaged electrocardiography (SAECG) may detectlate activated infundibular regions with low voltages.Previous studies that have assessed the associationbetween SAECG and VA were either performed inheterogeneous groups of patients with congenitalheart disease or have looked at any VA, includingpremature ventricular complexes and nonsustainedVT as outcome parameter, with conflicting results(26,27). One prior study conducted in 66 rTOF pa-tients undergoing repair via a RV ventriculotomy withsubsequent RBBB could demonstrate that patientswho developed sustained VT (n ¼ 12; 18%) had alonger filtered QRS interval on SAECG compared withthose without VT (179 � 18 ms vs. 165 � 17 ms,p ¼ 0.01) (28). Of interest, both additional SAECGcriteria (the high-frequency and low-amplitude signalduration, and the root mean square of the meanvoltage in the terminal portion of filtered QRS),


11

consistent with late activation of low amplitude re-gions, did not differ between groups. The predictivevalue of SAECG for sustained VT in contemporaryrTOF patients with a narrow QRS interval has neverbeen evaluated. Of interest, 2 patients with a narrowQRS interval had a severely dilated RV, suggestingthat in patients with rapid conduction through thespecific conduction system rather than primarymyocardial activation, RV size does not significantlycontribute to QRS duration.

RV ACTIVATION IN rTOF PATIENTS WITH RBBB. Inpatients with RBBB, rapid conduction via the specificconduction system is interrupted and RV activationdepends on slower myocardial impulse propagation.RBBB after surgery can be present at 3 levels: theproximal right fascicle; the distal ramification afterinfundibular resection; or the peripheral conductionsystem secondary to the right ventriculotomy (29). Ofinterest, a vertical ventriculotomy alone can result inan RBBB pattern with QRS prolongation by an averageof 39 ms (13). Irrespective of the level of block, thebasal lateral RV and RVOT are activated late (29). Inour patients with RBBB and no additional area ofconduction delay, the earliest RV activation during SRwas always observed at the septum, and excitationwave fronts were merging at the basal lateral TA. Inpatients with septal conduction delay due to SCAI oristhmus block, the pathway of impulse propagationwas prolonged with a shift of latest activation towardthe lateral RVOT. This latest area had preservedvoltage and contributed to prolongation of both theRVAT and QRS duration. This is in contrast to otherforms of structural heart disease with scar-related VT,where slow conduction within the scar results in latepotentials of low amplitude inscribing after QRSoffset.

QRS DURATION: GLOBAL VERSUS LOCALIZED

DELAY. In our series, all RBBB patients had normalCV remote from the anatomically defined isthmuses(i.e., normal global CV). Despite normal global CV,severe RV dilation can also prolong the pathway formyocardial impulse propagation, leading to an in-crease in QRS duration. Fourteen of 44 RBBB patientswith QRS >150 ms had a severely enlarged RV. In 8 of14, QRS prolongation was solely due to longer RVactivation time of the dilated RV, in 6 of 14 patients,SCAI or isthmus block further contributed to QRSprolongation. These data support the findings thatlocal rather global myocardial conduction delay orblock contributes to QRS prolongation in the majorityof patients.

The localized conduction delay or isthmusblock resulted in a shift of latest activation toward

the RVOT. The observed shift in latest activationprovides a potential explanation for prior findingson RV dyssynchrony showing that QRS durationwas strongly associated with the delay in RVOTshortening, which was more pronounced if QRSduration was $155 ms. The delay increased to morethan the upper 95% in all patients with a QRSduration >165 ms (30).

THE LINK BETWEEN QRS DURATION AND VT IN

rTOF. According to our data, in the absence of aseverely enlarged RV, QRS duration >150 ms is highlysuspicious for the presence of SCAI as potential sub-strate for VT. This finding provides an important linkbetween prolonged QRS interval and VT substrate inrTOF patients with RBBB. In only 1 patient of ourcohort, the VT substrate, that is, the SCAI, did notinfluence QRS duration because of the presence of asecond and normal conducting septal isthmus. A QRSduration >150 ms can also be observed in patientswith pre-existing isthmus block. Information onisthmus block can be sometimes derived from theoperation records, for example, description of an ab-sent infundibular septum. If not conclusive, EAM isrequired to distinguish slow conduction as substratefor VT from a pre-existing block. It would be inter-esting to assess in future studies, if noninvasivemethods such as ECG imaging are able to identifypresence and properties of septal and nonseptal AIwith sufficient spatial resolution, to overcome theneed for invasive EAM.

STUDY LIMITATIONS. This study is limited by thecross-sectional design. In the current study, rTOFpatients who were considered at risk for VT or whopresented with spontaneous VT underwent PES withRV mapping. The results can therefore not be appliedto the general rTOF population. In 5 patients withRBBB, we could not identify the mechanism of QRSduration >150 ms. As electroanatomic CV wasmeasured only at the endocardium and not at theepicardium, additional epicardial and transmuralconduction delay cannot be excluded. Electro-anatomically determined CV can only be a surrogatefor the complex 3-dimensional activation. However,the suggested cutoff value of <0.5 m/s is supportedby the following: 1) the range of RV electroanatomi-cally determined CV (range 0.52 to 1.89 m/s) found ina control group of patients with structural normalhearts (14); 2) a CV remote from the AI or in AI notrelated to VT of >0.5 m/s in our population; and 3) theconsistent finding that all VT-related AI had anelectroanatomically determined CV of <0.5 m/s.Differentiating between conduction block and longconduction delay can be difficult. Long conduction

PERSPECTIVES

COMPETENCY IN MEDICAL KNOWLEDGE 1: The

dominant substrate for VT in rTOF are SCAI. In rTOF

patients with RBBB, SCAI increase RV activation time

and QRS duration. In rTOF patients with a narrow QRS

interval, SCAI does not influence QRS duration.

COMPETENCY IN MEDICAL KNOWLEDGE 2: In

rTOF patients with RBBB, a QRS duration >150 ms is

highly suggestive for presence of SCAI or blocked

isthmus, in particular in patients without severe RV

dilatation. EAM can distinguish SCAI from isthmus

block and may become an important tool for VT

substrate identification.

COMPETENCY IN MEDICAL KNOWLEDGE 3: In

rTOF patients with a narrow QRS interval, a VT sub-

strate cannot be excluded.

TRANSLATIONAL OUTLOOK: Data are based on a

cross-sectional study in a selected group of rTOF

patients with documented sustained VT, considered at

risk for VT, and/or with indication for reoperation.

Further studies in the contemporary rTOF population

are desirable.



12

delay cannot be excluded if conduction time throughthe isthmus equals the conduction time required toactivate the isthmus from the other site. However,none of the patients with electroanatomic conductionblock had spontaneous VT or was inducible for VTduring long-term follow-up. In addition, a large RVwas pre-defined based on 1 RV end-diastolic volumeindex cutoff value.

CONCLUSIONS

In rTOF with pre-existing RBBB, the increased QRSduration is frequently due to localized conductiondelay across the AI rather than the global delay. TheseSCAI prolong the pathway for RV activation and QRSduration, providing an important causal link betweenthe QRS width and the substrate for VT. After exclu-sion of severe RV dilatation, a QRS duration >150 msis highly predictive for SCAI and may justify EAM todistinguish between a slow conducting and blockedAI. A narrow QRS interval in rTOF patients does notexclude SCAI and the risk for VT.

ADDRESS FOR CORRESPONDENCE: Dr. Katja Zeppenfeld,Department of Cardiology, Leiden University MedicalCentre, C5-P, PO Box 9600, 2300 RC Leiden, theNetherlands. E-mail: [email protected].

RE F E RENCE S

1. Khairy P, Ionescu-Ittu R, Mackie AS,Abrahamowicz M, Pilote L, Marelli AJ. Changingmortality in congenital heart disease. J Am CollCardiol 2010;56:1149–57.

2. Marelli AJ, Ionescu-Ittu R, Mackie AS, Guo L,Dendukuri N, Kaouache M. Lifetime prevalenceof congenital heart disease in the generalpopulation from 2000 to 2010. Circulation2014;130:749–56.

3. Moons P, Bovijn L, Budts W, Belmans A,Gewillig M. Temporal trends in survival to adult-hood among patients born with congenital heartdisease from 1970 to 1992 in Belgium. Circulation2010;122:2264–72.

4. Diller GP, Kempny A, Liodakis E, et al. Leftventricular longitudinal function predicts life-threatening ventricular arrhythmia and death inadults with repaired tetralogy of Fallot. Circula-tion 2012;125:2440–6.

5. Koyak Z, Harris L, de Groot JR, et al. Suddencardiac death in adult congenital heart disease.Circulation 2012;126:1944–54.

6. Khairy P, Harris L, Landzberg MJ, et al.Implantable cardioverter-defibrillators in tetral-ogy of Fallot. Circulation 2008;117:363–70.

7. Khairy P, Aboulhosn J, Gurvitz MZ, et al., for theAARCC Investigators. Arrhythmia burden in adultswith surgically repaired tetralogy of Fallot: a

multi-institutional study. Circulation 2010;122:868–75.

8. Balaji S, Lau YR, Case CL, Gillette PC. QRSprolongation is associated with inducible ventric-ular tachycardia after repair of tetralogy of Fallot.Am J Cardiol 1997;80:160–3.

9. Gatzoulis MA, Till JA, Somerville J,Redington AN. Mechanoelectrical interaction intetralogy of Fallot: QRS prolongation relates toright ventricular size and predicts malignant ven-tricular arrhythmias and sudden death. Circulation1995;92:231–7.

10. Gatzoulis MA, Balaji S, Webber SA, et al. Riskfactors for arrhythmia and sudden cardiac deathlate after repair of tetralogy of Fallot: a multi-centre study. Lancet 2000;356:975–81.

11. Bokma JP, Winter MM, Vehmeijer JT, et al. QRSfragmentation is superior to QRS duration in pre-dicting mortality in adults with tetralogy of Fallot.Heart 2017;103:666–71.

12. Muller J, Hager A, Diller GP, et al. Peak oxygenuptake, ventilatory efficiency and QRS-durationpredict event free survival in patients late aftersurgical repair of tetralogy of Fallot. Int J Cardiol2015;196:158–64.

13. Gelband H, Waldo AL, Kaiser GA,Bowman FO Jr., Malm JR, Hoffman BF. Etiology ofright bundle-branch block in patients undergoing

total correction of tetralogy of Fallot. Circulation1971;44:1022–33.

14. Kapel GF, Sacher F, Dekkers OM, et al.Arrhythmogenic anatomical isthmuses identifiedby electroanatomical mapping are the substratefor ventricular tachycardia in repaired Tetralogy ofFallot. Eur Heart J 2017;38:268–76.

15. Garson A Jr., Randall DC, Gillette PC, et al.Prevention of sudden death after repair of tetral-ogy of Fallot: treatment of ventricular arrhyth-mias. J Am Coll Cardiol 1985;6:221–7.

16. Ghai A, Silversides C, Harris L, Webb GD,Siu SC, Therrien J. Left ventricular dysfunction is arisk factor for sudden cardiac death in adults lateafter repair of tetralogy of Fallot. J Am Coll Car-diol 2002;40:1675–80.

17. Harrison DA, Harris L, Siu SC, et al. Sustainedventricular tachycardia in adult patients late afterrepair of tetralogy of Fallot. J Am Coll Cardiol1997;30:1368–73.

18. Jonsson H, Ivert T, Brodin LA, Jonasson R. Latesudden deaths after repair of tetralogy of Fallot:electrocardiographic findings associated with sur-vival. Scand J Thorac Cardiovasc Surg 1995;29:131–9.

19. Knauth AL, Gauvreau K, Powell AJ, et al.Ventricular size and function assessed by cardiacMRI predict major adverse clinical outcomes late

mailto:[email protected]

http://refhub.elsevier.com/S2405-500X(18)30095-1/sref1





















































































13

after tetralogy of Fallot repair. Heart 2008;94:211–6.

20. Draisma HHM, Swenne CA, van de Vooren H,et al. LEADS: an interactive research oriented ECG/VCG analysis system. Comput Cardiol 2015;32:515–8.

21. Kapel GF, Reichlin T, Wijnmaalen AP, et al. Re-entry using anatomically determined isthmuses: acurable ventricular tachycardia in repairedcongenital heart disease. Circ Arrhythm Electro-physiol 2015;8:102–9.

22. Bokma JP, Winter MM, Oosterhof T, et al.Preoperative thresholds for mid-to-late haemo-dynamic and clinical outcomes after pulmonaryvalve replacement in tetralogy of Fallot. Eur HeartJ 2016;37:829–35.

23. Zeppenfeld K, Schalij MJ, Bartelings MM, et al.Catheter ablation of ventricular tachycardia afterrepair of congenital heart disease: electroanatomicidentification of the critical right ventricularisthmus. Circulation 2007;116:2241–52.

24. Murphy JG, Gersh BJ, Mair DD, et al. Long-term outcome in patients undergoing surgicalrepair of tetralogy of Fallot. N Engl J Med 1993;329:593–9.

25. Durrer D, van Dam RT, Freud GE, Janse MJ,Meijler FL, Arzbaecher RC. Total excitation of theisolated human heart. Circulation 1970;41:899–912.

26. Perloff JK, Middlekauf HR, Child JS,Stevenson WG, Miner PD, Goldberg GD. Useful-ness of post-ventriculotomy signal averagedelectrocardiograms in congenital heart disease.Am J Cardiol 2006;98:1646–51.

27. Stelling JA, Danford DA, Kugler JD, et al. Latepotentials and inducible ventricular tachycardia insurgically repaired congenital heart disease. Cir-culation 1990;82:1690–6.

28. Russo G, Folino AF, Mazzotti E, Rebellato L,Daliento L. Comparison between QRS duration atstandard ECG and signal-averaging ECG forarrhythmic risk stratification after surgical repair

of tetralogy of Fallot. J Cardiovasc Electrophysiol2005;16:288–92.

29. Horowitz LN, Alexander JA, Edmunds LH Jr.Postoperative right bundle branch block: identifi-cation of three levels of block. Circulation 1980;62:319–28.

30. Uebing A, Gibson DG, Babu-Narayan SV, et al.Right ventricular mechanics and QRS duration inpatients with repaired tetralogy of Fallot: impli-cations of infundibular disease. Circulation 2007;116:1532–9.

KEY WORDS congenital heart disease, QRSduration, tetralogy of Fallot, ventriculartachycardia

APPENDIX For supplemental methods,results, figures, and videos, please see theonline version of this paper.






















































Documents

Slow Conducting Electroanatomic Isthmuses: An Important ...electrophysiology.onlinejacc.org/content/jcep/early/2018/03/22/j.jace… · grams (