- Email: [email protected]
Usefulness of Electrocardiographic Parameters for Risk Prediction in Arrhythmogenic Right Ventricular Dysplasia
Usefulness of Electrocardiographic Parameters for Risk Prediction in Arrhythmogenic Right Ventricular Dysplasia Ardan M. Saguner, MDa, Sabrina Ganahl, MDa, Samuel H. Baldinger, MDb, Andrea Kraus, PhDc, Argelia Medeiros-Domingo, MD, PhDa,b, Sebastian Nordbecka, Arhan R. Sagunera, Andreas S. Mueller-Burri, MDd, Laurent M. Haegeli, MDa, Thomas Wolber, MDa,e, Jan Steffel, MDa, Nazmi Krasniqi, MDa, Etienne Delacrétaz, MDb, Thomas F. Lüscher, MDa,e, Leonhard Held, PhDc, Corinna B. Brunckhorst, MDa, and Firat Duru, MDa,e,* The value of electrocardiographic findings predicting adverse outcome in patients with arrhythmogenic right ventricular dysplasia (ARVD) is not well known. We hypothesized that ventricular depolarization and repolarization abnormalities on the 12-lead surface electrocardiogram (ECG) predict adverse outcome in patients with ARVD. ECGs of 111 patients screened for the 2010 ARVD Task Force Criteria from 3 Swiss tertiary care centers were digitized and analyzed with a digital caliper by 2 independent observers blinded to the outcome. ECGs were compared in 2 patient groups: (1) patients with major adverse cardiovascular events (MACE: a composite of cardiac death, heart transplantation, survived sudden cardiac death, ventricular fibrillation, sustained ventricular tachycardia, or arrhythmic syncope) and (2) all remaining patients. A total of 51 patients (46%) experienced MACE during a follow-up period with median of 4.6 years (interquartile range 1.8 to 10.0). Kaplan-Meier analysis revealed reduced times to MACE for patients with repolarization abnormalities P Force Criteria (p [ 0.009), a precordial QRS P according to Task amplitude ratio ( QRS mV V1 to V3/ QRS mV V1 to V6) of £0.48 (p [ 0.019), and QRS fragmentation (p [ 0.045). In multivariable Cox regression, a precordial QRS amplitude ratio of £0.48 (hazard ratio [HR] 2.92, 95% confidence interval [CI] 1.39 to 6.15, p [ 0.005), inferior leads T-wave inversions (HR 2.44, 95% CI 1.15 to 5.18, p [ 0.020), and QRS fragmentation (HR 2.65, 95% CI 1.1 to 6.34, p [ 0.029) remained as independent predictors of MACE. In conclusion, in this multicenter, observational, long-term study, electrocardiographic findings were useful for risk stratification in patients with ARVD, with repolarization criteria, inferior leads TWI, a precordial QRS amplitude ratio of £0.48, and QRS fragmentation constituting valuable variables to predict adverse outcome. Ó 2014 Elsevier Inc. All rights reserved. (Am J Cardiol 2014;113:1728e1734) Arrhythmogenic right ventricular dysplasia (ARVD) is a genetically determined cardiomyopathy characterized by fibrofatty replacement of the right ventricle (RV), which may eventually lead to ventricular arrhythmias, heart failure, and sudden cardiac death (SCD).1,2 Risk stratification in ARVD still has the potential for improvement.3 Particularly, noninvasive and easily obtainable tools for risk stratification are needed. Twelve-lead electrocardiography constitutes
such a readily available clinical tool and plays an important role in ARVD diagnosis.4 However, its role for risk stratification and therapeutic decision making in ARVD remains controversial.5e8 The purpose of the present study was to identify electrocardiographic depolarization or repolarization abnormalities that predict an adverse outcome in our large cohort of patients with ARVD. Methods
a Department of Cardiology, University Heart Center Zurich, Zurich, Switzerland; bDepartment of Cardiology, University Hospital Bern, Bern, Switzerland; cDivision of Biostatistics, Institute for Social and Preventive Medicine and eCenter for Integrative Human Physiology, University of Zurich, Zurich, Switzerland; and dDepartment of Cardiology, Triemli Hospital Zurich, Switzerland. Manuscript received December 5, 2013; revised manuscript received and accepted February 12, 2014. Drs. A.M. Saguner and S. Ganahl contributed equally to this article and are shared first authors. This work and the Zurich ARVC Program are supported by a grant from the Georg and Bertha Schwyzer-Winiker Foundation, Zurich, Switzerland. See page 1733 for disclosure information. *Corresponding author: Tel: (þ41) 44 2552099; fax: (þ41) 44 2554401. E-mail address: fi[email protected] (F. Duru).
0002-9149/14/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2014.02.031
The study population included 111 patients from 3 tertiary care centers in Switzerland with a possible, borderline, or definite diagnosis of ARVD (possible diagnosis: 2 minor or 1 major criterion; borderline: 1 major and 1 minor or 3 minor criteria; and definite: 2 major, 1 major plus 2 minor, or 4 minor criteria according to the 2010 revised Task Force Criteria [TFC]),4 who had a 12-lead surface electrocardiogram (ECG) recorded from February 1987 to March 2013. Ninety-eight patients were index patients (probands) and 13 were family members. Clinical information regarding demographics and symptoms were obtained from hospital records at the time of the electrocardiography. This observational study was approved by the local institutional ethical committees of the participating centers. Electrocardiographic www.ajconline.org
Cardiomyopathy/ECG Predicts Outcome in ARVD
1729
Table 1 Definitions of electrocardiographic parameters and measurements used in this study Variable Epsilon (e) wave TAD TWI Inferior leads TWI Major repolarization criteria Minor repolarization criteria Precordial QRS amplitude ratio Parietal block Fragmented QRS Peripheral low voltage QRS dispersion QT dispersion Precordial TpeakeTend interval Precordial TpeakeTend dispersion Complete right bundle branch block
Incomplete right bundle branch block R or r0 in V1/V2
Definition Distinct waves of small amplitude that occupy the ST segment in the right precordial leads and are distinct from the QRS complex Longest value in V1 through V3 from the nadir of the S wave to the end of all depolarization Any T-wave negativity Inverted T waves in 2 of 3 inferior leads Definition according to 2010 TFC Definition according P to 2010 TFC P QRS mV V1V3/ QRS mV V1eV6 QRSd in lead V1 through V3 that exceeds the QRSd in lead V6 by >25 ms Additional deflections/notches at the beginning of the QRS, on top of the R wave, or in the nadir of the S wave in either 1 right precordial lead or in >1 lead including all remaining leads QRS amplitude 5 mm (0.5 mV) in each peripheral lead Difference between the longest and shortest QRS durations in all leads Difference between the longest and shortest uncorrected QT intervals in all leads Difference of the interval of the Q wave to the end of the T wave (QTend) and the interval of the Q wave to the peak of the T wave (QTpeak) measured in the precordial leads, representing a surrogate marker of the transmural dispersion of repolarization Difference between the longest and shortest precordial TpeakeTend interval QRSd 120 ms and A1: R0 or r0 in V1 or V2 A2: S duration > R duration in I and V6 A3: S duration >40 ms in I and V6 A4: R peak time >50 ms in V1 or V2 a: A1 þ A2 b: A1 þ A3 c: A4 þ (A2 or A3) QRS <120 ms and R peak time in V1 or V2 >50 ms A positive deflection in V1/V2 after an S wave
Figure 1. (A) A representative 12-lead surface ECG (25 mm/s, 10 mm/mV) from a patient with definite ARVD and adverse outcome showing measurements for calculation of the precordial QRS amplitude ratio in the precordial leads. This patient had poor R-wave progression in the right precordial leads resulting in a P P precordial QRS amplitude ratio ( QRS mV V1 to V3/ QRS mV V1 to V6) of 0.27 (0.48). QRS voltage amplitudes are marked in blue in leads V1 to V3 and in red in leads V4 to V6. Please note that this ECG shows a major repolarization criterion (TWIs V1 to V4) and a minor depolarization criterion (TAD 55 ms) according to the 2010 revised TFC. (B) The precordial leads V1 to V3 taken from a 12-lead surface ECG (25 mm/s, 10 mm/mV) demonstrate extensive QRS fragmentation (black arrows) in this patient with definite ARVD and adverse outcome. (C) A representative 12-lead surface ECG (25 mm/s, 10 mm/mV) from a patient with definite ARVD showing inferior leads TWIs in III and aVF (black arrows on the left) and in V1 to V3 (black arrows on the right), fulfilling a major repolarization criterion according to the 2010 revised TFC.
definitions used in this study are listed in Table 1. ECGs were recorded at rest (25 mm/s, 10 mm/mV) with standard lead positions, digitized with a high-resolution scanner, and
analyzed with a digital caliper (Screen Caliper, version 4.0; Iconico, www.iconico.com, New York, New York).9 ECGs were enlarged 4 times. Electrocardiographic intervals were
1730
The American Journal of Cardiology (www.ajconline.org)
Table 2 Baseline clinical characteristics Characteristic Age (yrs) Men Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (beats/min) Body surface area (m2) Body mass index (kg/m2) LV ejection fraction <50% LV ejection fraction <40% Previous sustained VT/VF/sudden cardiac arrest Fractional area change <33% Previous syncope “Definite” ARVD Premature SCD in a first- or second-degree family member Amiodarone b Blocker Sotalol Other antiarrhythmic therapy
All Patients (n ¼ 111)
MACE (n ¼ 51)
Favorable (n ¼ 60)
p Value
43 15.4 71 (64) 120 (110e129) 76 9 65 (57e75) 1.9 (1.7e2.0) 24.5 (21.8e26.1) 15 (14) 8 (7) 52 (47) 48 (43) 35 (32) 80 (72) 11 (10) 18 (16) 48 (43) 13 (12) 9 (8)
46 14.9 33 (65) 120 (110e126) 77 9 63 (55e79) 1.9 (1.7e2.0) 24.5 (21.4e26.2) 12 (24) 7 (14) 35 (69) 35 (69) 24 (47) 44 (86) 5 (10) 11 (22) 29 (57) 7 (14) 6 (12)
40 15.3 38 (63) 120 (110e131) 76 10 65 (57e73) 1.9 (1.7e2.0) 24.3 (22.1e26.1) 3 (5) 1 (2) 17 (28) 16 (27) 11 (18) 36 (60) 6 (10) 6 (10) 18 (30) 6 (10) 3 (5)
0.03 1.00 0.12 0.4 0.39 0.3 0.76 0.005 0.023 <0.001 <0.001 0.016 0.003 1.00 0.11 0.007 0.57 0.3
Values are mean SD, median (interquartile range), or number (percentage). p Values were calculated by 2-sided unpaired Student t test for continuous variables and by Fisher’s exact test for categorical variables.
measured in 2 consecutive beats in each lead, and the mean value was used. When the difference between the 2 beats was >10 ms, then the mean of 3 beats was taken. ECGs were independently analyzed by 2 experienced blinded readers. Differences in electrocardiographic interpretation were adjudicated by a third experienced reader, and a final conclusion was made by consensus. Heart rate, PR interval, QRS duration, QRS dispersion, QT interval, QT dispersion, Tpeak to Tend duration, parietal block, and terminal activation duration (TAD) were measured according to current practice.10e15 The Bazett formula was used to correct the QT interval for heart rate. QRS fragmentation (Figure 1) and peripheral low voltage were defined as previously described.7,8,16e18 We also included a recently proposed parameter termed “precordial QRS amplitude ratio” in our analysis (Figure 1), which is the ratio of the sum of PQRS amplitudes inP V1 to V3 divided by the sum in V1 to V6 ( QRS mV V1 to V3/ QRS mV V1 to V6), with low values reflecting lower voltages in the right precordial leads compared with the lateral precordial leads.19 We also evaluated the impact of inferior leads T-wave inversions (TWIs; Figure 1).20 Repolarization abnormalities according to the 2010 TFC were defined as precordial TWI in at least V1 and V2 (minor criterion). Follow-up for outcome data was performed by chart review, including implantable cardioverter-defibrillator interrogations, Holter ECGs, clinical visits, and from telephone interviews of patients or treating physicians. No patient was lost to follow-up. The role of electrocardiographic parameters on outcome was compared in 2 patient groups: (1) patients with major adverse cardiovascular events (MACE; “adverse outcome”); composite of cardiac death (defined as any death that occurred because of SCD or terminal congestive heart failure), heart transplantation, survived SCD, ventricular fibrillation (VF), sustained ventricular tachycardia (VT), and arrhythmogenic syncope21 and (2) all remaining patients (“favorable outcome”). For Kaplan-Meier estimates and Cox regression analyses, time from electrocardiography to MACE
was the event of interest. Survived SCD, VF, sustained VT, and arrhythmogenic syncope were defined as previously reported.21,22 Continuous variables are presented as mean SD or median (with interquartile range) and were compared using a 2-sided unpaired Student t test. Categorical variables are reported as frequency (percentage) and were compared between groups by Fisher’s exact test. Cut-off points for continuous electrocardiographic variables that maximized the difference in times to MACE for patients with values lower and for those greater than the cut-off were determined by minimizing the p value (log-rank). The final p value was corrected for multiple testing. Cumulative probabilities of survival free of MACE were determined by the Kaplan-Meier method and differences in survival with the log-rank test. Baseline variables associated with MACE were identified by univariable Cox regression. Variables with a p value of <0.1 in univariable analysis were considered for the multivariable model. Patients with ARVD with a previous episode of sustained VT, VF, and/or sudden cardiac arrest have a high probability of MACE. Thus, we adjusted for this variable by performing a stratified analysis. For continuous variables, the intraclass correlation was used to quantify interobserver variability with values >0.75 representing good reliability of measurements. For nominal variables, we calculated the agreement coefficient k, with a value of 1 indicating perfect agreement. A 2-sided p value of <0.05 was considered significant. Statistical analysis was performed using R programing language (R Development Core Team, 2009) and GraphPad Prism 5 (GraphPad Software Inc., La Jolla, California). Results ECGs were available for all patients. Patient demographic data, baseline characteristics, and ECG variables are summarized in Tables 2 and 3. ARVD was classified as definite in 80 patients (72%), borderline in 15 patients (14%), and
Cardiomyopathy/ECG Predicts Outcome in ARVD Table 3 Twelve-lead electrocardiographic data Variable Heart rate (beats/min) RR interval (seconds) PQ interval (ms) QRS duration (ms) QRS dispersion (ms) QT interval (ms) QT interval corrected (ms) QT dispersion (ms) Precordial QRS ratio Precordial TpeakeTend interval (ms) Precordial TpeakeTend dispersion (ms) TAD (ms) TAD 55 ms Minor depolarization criteria Major depolarization criteria (epsilon waves) Fragmented QRS Isolated late potentials in inferior leads Major repolarization criteria Minor repolarization criteria TWIs in 3 precordial leads Inferior leads TWI TWIs in 3 precordial leads combined with inferior leads TWI Complete left bundle branch block Complete right bundle branch block Left anterior fascicular block Incomplete right bundle branch block Parietal block Premature ventricular complexes present on 12-lead ECG
All (n ¼ 111) 65 (57e75) 0.92 (0.8e1.1) 160 (144e184) 111 (100e127) 39 (28e51) 420 (392e465) 443 (425e480) 57 (46e82) 0.475 0.091 96.8 (80e107.4) 30.8 (21e46.8) 52 (43e64) 36 (32) 21 (19) 22 (20) 42 (38) 4 (4) 37 (33) 21 (19) 32 (29) 48 (43) 12 (11) 2 13 8 3 13 19
(2) (12) (7) (3) (12) (17)
Values are mean SD, median (interquartile range), or number (percentage). Definitions: see Table 1.
Table 4 Specification of major adverse cardiovascular events (MACE) during follow-up MACE
n ¼ 51
Cardiac death (SCD or congestive heart failure) Heart transplantation VF Sustained VT Arrhythmic syncope
3 4 8 34 2
(6) (8) (16) (67) (3)
Values are presented as number (percentage).
possible in 16 patients (14%). Eleven of 16 patients (69%) with a possible diagnosis were family members. Index cases with possible ARVD fulfilled 1 major (n ¼ 1) or 2 minor criteria (n ¼ 4). Two of those patients had suspected family disease, and 1 patient had RV scar identified by electroanatomic voltage mapping. After study inclusion, during a follow-up with median of 4.6 years (interquartile range 1.8 to 10), 51 patients (46%) experienced MACE (Table 4). Median time from study inclusion to MACE was 12 months (interquartile range 4.1 to 36). At last follow-up, 108 patients (97%) were alive. Three patients died from cardiac causes (2 because of terminal congestive heart failure and 1 because of
1731
incessant VT), and cardiac transplantation was performed in 4 patients (3 because of terminal congestive heart failure and 1 because of repetitive uncontrollable implantable cardioverter-defibrillator shocks). Histology of all 4 explanted hearts showed extensive fibrofatty infiltration within the RV myocardium and the posterolateral aspect of the left ventricle (LV). There was no evidence of myocarditis or superimposed inflammatory phenomena in any of those cases. Analysis of times to MACE showed that the presence of precordial repolarization abnormalities according to the 2010 TFC, a precordial QRS amplitude ratio of 0.48, and QRS fragmentation were associated with significantly reduced times to MACE (Figure 2). Detailed data for the 2 subgroups with (n ¼ 52) and without (n ¼ 59) SCD, VF, and/or sustained VT before study inclusion are visualized by the stratified KaplanMeier plots. In general, times to MACE were longer in patients without previous SCD, VF, and/or sustained VT. After stratification for this observation, TWI in 2 precordial leads and QRS fragmentation were significantly associated with shorter survival, also in the subgroup of patients considered for primary prophylactic implantable cardioverter-defibrillator implantation (Figure 2). The number of precordial leads with TWI was significantly associated with MACE (3 leads vs >3, p ¼ 0.017). On the contrary, presence of depolarization abnormalities, parietal block, peripheral low voltage, QT prolongation, prolonged QRS, QT and/or Tpeak to Tend dispersion, and Tpeak to Tend interval prolongation did not significantly alter times to MACE. Precordial repolarization criteria, a precordial QRS amplitude ratio of 0.48, TWI in >3 precordial leads, QRS fragmentation, and inferior leads TWI were identified as univariable predictors of MACE also in the Cox model (Table 5). In multivariable analysis starting with these univariable predictors, a precordial QRS amplitude ratio of 0.48 (hazard ratio [HR] 2.92, 95% confidence interval [CI] 1.39 to 6.15, p ¼ 0.005), inferior leads TWI (HR 2.44, 95% CI 1.15 to 5.18, p ¼ 0.02), and QRS fragmentation (HR 2.65, 95% CI 1.1 to 6.34, p ¼ 0.029) remained as the only independent predictors of MACE. Separate analyses excluding 5 index patients with possible ARVD yielded similar results (Table 5). In this subpopulation of 106 patients, multivariable analysis showed the following results: HR 3.45 (95% CI 1.54 to 7.74, p ¼ 0.003) for precordial QRS amplitude ratio 0.48; HR 2.65 (95% CI 1.22 to 5.74, p ¼ 0.014) for inferior leads TWI; and HR 2.31 (95% CI 0.94 to 5.67, p ¼ 0.067) for QRS fragmentation. Kaplan-Meier subanalyses for patients with definite and borderline ARVD without previous SCD, VF, and/or sustained VT also demonstrated significantly lower survival in the presence of precordial repolarization criteria and QRS fragmentation (Figure 2). In this same subpopulation, Cox regression identified the presence of QRS fragmentation (HR 4.44, 95% CI 1.52 to 12.95, p ¼ 0.006) and precordial repolarization criteria (HR 4.4, 95% CI 1.17 to 16.56, p ¼ 0.029) as univariable predictors of MACE. The low number of MACE (n ¼ 15) in this subpopulation did not allow for multivariable analysis. The estimated intraclass correlation was 0.9 (95% CI 0.74 to 0.94) for precordial QRS amplitude ratio (k of 0.94 [95% CI 0.85 to 1.0]) for the cutoff 0.48. Interobserver correlations were higher for repolarization than for depolarization abnormalities (k ¼ 0.89 [95% CI 0.8 to 0.97] for repolarization criteria and k ¼ 0.95 [95% CI 0.9 to 1.0] for
1732
The American Journal of Cardiology (www.ajconline.org)
Figure 2. (A) Kaplan-Meier analysis of times to MACE (composite of cardiac death, heart transplant, SCD, VF, sustained VT, or arrhythmic syncope; n ¼ 111) for the time between electrocardiography and end point, stratified by repolarization criteria according to the 2010 revised TFC (left diagram), a precordial QRS amplitude ratio of 0.48 (middle diagram), and QRS fragmentation (right diagram); all analyses were stratified for the presence of survived SCD, VF, and/or sustained VT before study inclusion (black graphs: patients without survived SCD, VF, and/or sustained VT before study inclusion; purple graphs: patients with survived SCD, VF, and/or sustained VT before study inclusion). (B) Kaplan-Meier analysis of times to MACE for the time between electrocardiography and end point in the subpopulation of patients (n ¼ 44) without survived SCD, VF, and/or sustained VT before study inclusion and having a definite or borderline diagnosis of ARVD; analyses stratified by repolarization criteria according to the 2010 revised TFC (left diagram) and QRS fragmentation (right diagram). strat ¼ stratified.
inferior leads TWI, compared with k ¼ 0.5 [95% CI 0.33 to 0.66] for QRS fragmentation, k ¼ 0.49 [95% CI 0.25 to 0.69] for e waves in leads V1 to V3, k ¼ 0.38 [95% CI 0.07 to 0.6] for e waves in lead V1, k ¼ 0.36 [95% CI 0.03 to 0.7] for e waves in lead V2, k ¼ 0.24 [95% CI 0.06 to 0.6] for e waves in lead V3, and k ¼ 0.57 [95% CI 0.42 to 0.69] for the presence of TAD 55 ms). Discussion This long-term observational study suggests that 12-lead surface electrocardiographic parameters are useful to predict adverse events in ARVD based on the following findings: (1) precordial and inferior leads TWI predict MACE, (2) QRS fragmentation and the precordial QRS amplitude ratio constitute additional predictors of MACE; and (3) variability in interpretation of depolarization abnormalities may limit their clinical relevance, whereas a high interobserver correlation was found for repolarization abnormalities and the precordial QRS amplitude ratio. There is a need for noninvasive tools such as the 12-lead electrocardiography for risk stratification in ARVD.4 Previous studies about the predictive role of electrocardiography in ARVD have yielded
conflicting results.5e8 Earlier studies focused on depolarization abnormalities. Although results were not consistent,23,24 most of these studies have shown that QRS and/or QT dispersion and other depolarization abnormalities may be useful for risk prediction in ARVD.5,6,25 Recent studies with larger patient numbers and less advanced ARVD have investigated the predictive role of precordial depolarization or repolarization abnormalities, and results were again not consistent.21,23,26 A plausible explanation for these diverging findings may be related to different demographics and heterogeneous study protocols. Importantly, earlier studies reporting continuous electrocardiographic variables did not use digital calipers, which may limit the validity of these data. We show that precordial and inferior leads TWI have predictive value in ARVD, whereas depolarization abnormalities and continuous electrocardiographic variables such as e waves, TAD, and QRS, QT, and/or Tpeak to Tend dispersion only have limited predictive value, although prolonged TAD and e waves are accepted delay parameters in ARVD. A main limitation of these measures is their high interobserver variability. A higher interobserver correlation may partially explain why fragmented QRS was more valuable for risk prediction than e wave. However, interobserver variability
Cardiomyopathy/ECG Predicts Outcome in ARVD
1733
Table 5 Electrocardiographic variables associated with major adverse cardiovascular events (MACE) in the whole study population (n ¼ 111, left column) and in the subpopulation excluding index patients with possible disease (n ¼ 106, right column) Variable
Repolarization criteria according to 2010 TFC Precordial QRS amplitude ratio 0.48 TWIs >3 precordial leads QRS fragmentation TWIs in 2 inferior leads Late potentials in 1 lead Epsilon waves TpeakeTend duration >100 ms Bundle branch block TpeakeTend dispersion >20 ms Peripheral low voltage Presence of premature ventricular complexes on 12-lead ECG Parietal block TAD 55 ms
Univariable Analysis (n ¼ 111)
Univariable Analysis (n ¼ 106)
HR (95% CI)
HR (95% CI)
2.4 2.31 1.88 1.76 1.73 1.63 1.61 1.47 0.7 1.36 1.29 1.2 1.07 1.07
(1.22e4.73) (1.13e4.74) (1.07e3.31) (1.01e3.06) (0.98e3.06) (0.89e3.01) (0.86e3.03) (0.83e2.61) (0.38e1.32) (0.69e2.67) (0.63e2.66) (0.61e2.35) (0.57e1.99) (0.57e1.99)
p Value 0.011 0.022 0.029 0.047 0.058 0.11 0.14 0.19 0.27 0.38 0.48 0.6 0.83 0.83
2.49 2.7 1.87 1.69 1.79 1.75 1.71 1.4 0.71 1.48 1.33 1.24 1.05 1.1
(81.23e5.04) (1.24e5.86) (1.05e3.34) (0.95e3.03) (1.00e3.21) (0.94e3.25) (0.9e3.23) (0.77e2.55) (0.31e1.6) (0.71e3.1) (0.64e2.75) (0.63e2.46) (0.58e1.9) (0.58e2.08)
p Value 0.012 0.012 0.034 0.075 0.05 0.076 0.1 0.27 0.41 0.3 0.44 0.53 0.88 0.78
p Values were calculated by univariable Cox regression, stratified for the presence of sudden cardiac arrest, VF, and/or sustained VT before study inclusion.
does not explain the incremental prognostic value of fragmented QRS compared with TAD, as k values were similar for both. We believe that the incremental prognostic value of fragmented QRS is also explained by its greater prevalence, and that fragmented QRS—determined in all 12 electrocardiographic leads—visualizes conduction delay within more aspects of the myocardium, as it is not limited to only leads V1 to V3. Furthermore, fragmented QRS visualizes conduction delay during all phases of ventricular depolarization, and not only at later phases. Our study underlines the clinical utility of precordial TWI to predict MACE in ARVD. This study confirms that an incremental relation between precordial TWI and MACE exists, with >3 inverted precordial T waves conferring an almost twofold increased hazard compared with 3 precordial TWIs. An important and novel finding is the independent prognostic value of inferior leads TWIs, which have been linked to LV involvement, and thus may explain their predictive utility.20 In multivariable analysis, inferior leads TWI remained an independent predictor of MACE. These findings agree with the important predictive role of repolarization abnormalities in heart failure and myocardial infarction.27 These reliable and easily recognized abnormalities seem to constitute key elements of arrhythmic risk in patients with ARVD, probably because they are associated with severe RV disease as well as LV involvement.5,28 Another novel finding of our study is the predictive role of a reduced precordial QRS amplitude ratio, which remained the second independent predictor of MACE in multivariable analysis. This ratio at a recently proposed cutoff of 0.48 has been shown to predict VT recurrence after successful ablation of VT in patients with ARVD with electrical storm.19 Importantly, this ratio is easy to measure and reliable. The third independent predictor of MACE was the presence of QRS fragmentation.7,8 Importantly, besides repolarization criteria, this variable was the only significant predictor of MACE in our lower risk subpopulation of patients with definite and borderline ARVD without previous SCD, VF, and/or sustained VT. Previous work has suggested
that myocardial scar may be reflected by electrocardiographic QRS fragmentation and thus confers a poor prognosis in coronary artery disease.29 Two recent studies with smaller patient numbers have demonstrated this association in patients with ARVD. Yet, the predictive value of other electrocardiographic variables was not assessed in those studies, and the multivariable model by Canpolat et al8 was probably overfitted.7,8 Of note, interobserver variability for QRS fragmentation was rather high in our cohort, which may limit the clinical utility of this variable. This study is limited by its observational nature inherent to registries. Thus, inclusion of patients with ARVD at different stages of the disease was inevitable, although we made an effort to include patients early after first medical contact. We included patients with possible ARVD, although their phenotype may not be typical. Yet, those patients form an integral part of our routine clinical practice, and a considerable number of patients with early ARVD may be missed if only patients with advanced phenotype are considered. In general, we studied a rather high-risk cohort of patients given the high MACE rates reflecting a referral bias and indicating the need for confirmation of our results in lower risk populations, for example, in family members who generally present with a less severe phenotype.30 Therefore, our results mainly obtained in index patients may not be transferable to family members. As our ARVD program is emerging, we currently do not have sufficient genetic data to provide, and because of low patient numbers, our data do not allow subanalyses of different ethnicities. Thus, we may have missed important genotypephenotype correlations that could be associated with our findings. As such, our suggested cutoffs for the precordial QRS amplitude ratio may not be applicable to patients with desmoplakin and phospholamban mutations, who often present with LV involvement. Disclosures The authors have no conflicts of interest to disclose.
1734
The American Journal of Cardiology (www.ajconline.org)
1. Roberts WC, Ko JM, Kuiper JJ, Hall SA, Meyer DM. Some previously neglected examples of arrhythmogenic right ventricular dysplasia/cardiomyopathy and frequency of its various reported manifestations. Am J Cardiol 2010;106:268e274. 2. Kim C, Wong J, Wen J, Wang S, Wang C, Spiering S, Kan NG, Forcales S, Puri PL, Leone TC, Marine JE, Calkins H, Kelly DP, Judge DP, Chen HS. Studying arrhythmogenic right ventricular dysplasia with patient-specific iPSCs. Nature 2013;494:105e110. 3. Schuler PK, Haegeli LM, Saguner AM, Wolber T, Tanner FC, Jenni R, Corti N, Luscher TF, Brunckhorst C, Duru F. Predictors of appropriate ICD therapy in patients with arrhythmogenic right ventricular cardiomyopathy: long term experience of a tertiary care center. PLoS One 2012;7:e39584. 4. Marcus FI, McKenna WJ, Sherrill D, Basso C, Bauce B, Bluemke DA, Calkins H, Corrado D, Cox MG, Daubert JP, Fontaine G, Gear K, Hauer R, Nava A, Picard MH, Protonotarios N, Saffitz JE, Sanborn DM, Steinberg JS, Tandri H, Thiene G, Towbin JA, Tsatsopoulou A, Wichter T, Zareba W. Diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia: proposed modification of the Task Force Criteria. Eur Heart J 2010;31:806e814. 5. Peters S, Peters H, Thierfelder L. Risk stratification of sudden cardiac death and malignant ventricular arrhythmias in right ventricular dysplasia-cardiomyopathy. Int J Cardiol 1999;71:243e250. 6. Turrini P, Corrado D, Basso C, Nava A, Bauce B, Thiene G. Dispersion of ventricular depolarization-repolarization: a noninvasive marker for risk stratification in arrhythmogenic right ventricular cardiomyopathy. Circulation 2001;103:3075e3080. 7. Peters S, Truemmel M, Koehler B. Prognostic value of QRS fragmentation in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia. J Cardiovasc Med (Hagerstown) 2012;13:295e298. 8. Canpolat U, Kabakci G, Aytemir K, Dural M, Sahiner L, Yorgun H, Sunman H, Baris Kaya E, Tokgozoglu L, Oto A. Fragmented QRS complex predicts the arrhythmic events in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia. J Cardiovasc Electrophysiol 2013;24:1260e1266. 9. Roten L, Derval N, Sacher F, Pascale P, Scherr D, Komatsu Y, Ramoul K, Daly M, Denis A, Shah AJ, Hocini M, Jais P, Haissaguerre M. Heterogeneous response of J-wave syndromes to beta-adrenergic stimulation. Heart Rhythm 2012;9:1970e1976. 10. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mazur NA, Meijler FL, Mogensen L, Moret P. Criteria for intraventricular conduction disturbances and pre-excitation. World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261e1275. 11. Fontaine G, Fontaliran F, Hebert JL, Chemla D, Zenati O, Lecarpentier Y, Frank R. Arrhythmogenic right ventricular dysplasia. Annu Rev Med 1999;50:17e35. 12. Antzelevitch C, Yan GX, Shimizu W. Transmural dispersion of repolarization and arrhythmogenicity: the Brugada syndrome versus the long QT syndrome. J Electrocardiol 1999;32(Suppl):158e165. 13. Surawicz B, Childers R, Deal BJ, Gettes LS, Bailey JJ, Gorgels A, Hancock EW, Josephson M, Kligfield P, Kors JA, Macfarlane P, Mason JW, Mirvis DM, Okin P, Pahlm O, Rautaharju PM, van Herpen G, Wagner GS, Wellens H; American Heart Association E, Arrhythmias Committee CoCC, American College of Cardiology F, Heart Rhythm S. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part III: intraventricular conduction disturbances: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol 2009;53:976e981. 14. Rautaharju PM, Surawicz B, Gettes LS, Bailey JJ, Childers R, Deal BJ, Gorgels A, Hancock EW, Josephson M, Kligfield P, Kors JA, Macfarlane P, Mason JW, Mirvis DM, Okin P, Pahlm O, van Herpen G, Wagner GS, Wellens H; American Heart Association E, Arrhythmias Committee CoCC, American College of Cardiology F, Heart Rhythm S. AHA/ ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society: endorsed by the International Society for Computerized Electrocardiology. Circulation 2009;119:e241ee250.
15. Cox MG, Nelen MR, Wilde AA, Wiesfeld AC, van der Smagt JJ, Loh P, Cramer MJ, Doevendans PA, van Tintelen JP, de Bakker JM, Hauer RN. Activation delay and VT parameters in arrhythmogenic right ventricular dysplasia/cardiomyopathy: toward improvement of diagnostic ECG criteria. J Cardiovasc Electrophysiol 2008;19:775e781. 16. Rapezzi C, Merlini G, Quarta CC, Riva L, Longhi S, Leone O, Salvi F, Ciliberti P, Pastorelli F, Biagini E, Coccolo F, Cooke RM, BacchiReggiani L, Sangiorgi D, Ferlini A, Cavo M, Zamagni E, Fonte ML, Palladini G, Salinaro F, Musca F, Obici L, Branzi A, Perlini S. Systemic cardiac amyloidoses: disease profiles and clinical courses of the 3 main types. Circulation 2009;120:1203e1212. 17. Mussinelli R, Salinaro F, Alogna A, Boldrini M, Raimondi A, Musca F, Palladini G, Merlini G, Perlini S. Diagnostic and prognostic value of low QRS voltages in cardiac AL amyloidosis. Ann Noninvasive Electrocardiol 2013;18:271e280. 18. Peters S, Trummel M, Koehler B. QRS fragmentation in standard ECG as a diagnostic marker of arrhythmogenic right ventricular dysplasiacardiomyopathy. Heart Rhythm 2008;5:1417e1421. 19. Muessigbrodt A, Dinov L, Fiedler L, Sommer P, Richter S, Breithardt O, Bollmann A, Husser D, Hindricks G, Arya A. Precordial QRS amplitude ratio predicts long term outcome after catheter ablation of electrical storm due to ventricular tachycardias in patients with arrhythmogenic right ventricular cardiomyopathy [abstract]. Europace 2013;15(Suppl 2):ii82eii103.22. 20. Jain R, Dalal D, Daly A, Tichnell C, James C, Evenson A, Abraham T, Tan BY, Tandri H, Russell SD, Judge D, Calkins H. Electrocardiographic features of arrhythmogenic right ventricular dysplasia. Circulation 2009;120:477e487. 21. Corrado D, Calkins H, Link MS, Leoni L, Favale S, Bevilacqua M, Basso C, Ward D, Boriani G, Ricci R, Piccini JP, Dalal D, Santini M, Buja G, Iliceto S, Estes NA 3rd, Wichter T, McKenna WJ, Thiene G, Marcus FI. Prophylactic implantable defibrillator in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia and no prior ventricular fibrillation or sustained ventricular tachycardia. Circulation 2010;122:1144e1152. 22. Saguner AM, Medeiros-Domingo A, Schwyzer MA, On CJ, Haegeli LM, Wolber T, Hurlimann D, Steffel J, Krasniqi N, Rueger S, Held L, Luscher TF, Brunckhorst C, Duru F. Usefulness of inducible ventricular tachycardia to predict long-term adverse outcomes in arrhythmogenic right ventricular cardiomyopathy. Am J Cardiol 2013;111:250e257. 23. Hulot JS, Jouven X, Empana JP, Frank R, Fontaine G. Natural history and risk stratification of arrhythmogenic right ventricular dysplasia/ cardiomyopathy. Circulation 2004;110:1879e1884. 24. Roguin A, Bomma CS, Nasir K, Tandri H, Tichnell C, James C, Rutberg J, Crosson J, Spevak PJ, Berger RD, Halperin HR, Calkins H. Implantable cardioverter-defibrillators in patients with arrhythmogenic right ventricular dysplasia/cardiomyopathy. J Am Coll Cardiol 2004;43: 1843e1852. 25. Lemola K, Brunckhorst C, Helfenstein U, Oechslin E, Jenni R, Duru F. Predictors of adverse outcome in patients with arrhythmogenic right ventricular dysplasia/cardiomyopathy: long term experience of a tertiary care centre. Heart 2005;91:1167e1172. 26. Bhonsale A, James CA, Tichnell C, Murray B, Madhavan S, Philips B, Russell SD, Abraham T, Tandri H, Judge DP, Calkins H. Risk stratification in arrhythmogenic right ventricular dysplasia/cardiomyopathyassociated desmosomal mutation carriers. Circ Arrhythm Electrophysiol 2013;6:569e578. 27. Zabel M, Acar B, Klingenheben T, Franz MR, Hohnloser SH, Malik M. Analysis of 12-lead T-wave morphology for risk stratification after myocardial infarction. Circulation 2000;102:1252e1257. 28. Steriotis AK, Bauce B, Daliento L, Rigato I, Mazzotti E, Folino AF, Marra MP, Brugnaro L, Nava A. Electrocardiographic pattern in arrhythmogenic right ventricular cardiomyopathy. Am J Cardiol 2009;103:1302e1308. 29. Das MK, Suradi H, Maskoun W, Michael MA, Shen C, Peng J, Dandamudi G, Mahenthiran J. Fragmented wide QRS on a 12-lead ECG: a sign of myocardial scar and poor prognosis. Circ Arrhythm Electrophysiol 2008;1:258e268. 30. Saguner AM, Vecchiati A, Baldinger SH, Rüeger S, Medeiros-Domingo A, Mueller-Burri AS, Haegeli LM, Biaggi P, Manka R, Lüscher TF, Fontaine G, Delacrétaz E, Jenni R, Held L, Brunckhorst C, Duru F, Tanner FC. Different prognostic value of functional right ventricular parameters in arrhythmogenic right ventricular cardiomyopathy/ dysplasia. Circ Cardiovasc Imaging 2014;7:230e239.
such a readily available clinical tool and plays an important role in ARVD diagnosis.4 However, its role for risk stratification and therapeutic decision making in ARVD remains controversial.5e8 The purpose of the present study was to identify electrocardiographic depolarization or repolarization abnormalities that predict an adverse outcome in our large cohort of patients with ARVD. Methods
a Department of Cardiology, University Heart Center Zurich, Zurich, Switzerland; bDepartment of Cardiology, University Hospital Bern, Bern, Switzerland; cDivision of Biostatistics, Institute for Social and Preventive Medicine and eCenter for Integrative Human Physiology, University of Zurich, Zurich, Switzerland; and dDepartment of Cardiology, Triemli Hospital Zurich, Switzerland. Manuscript received December 5, 2013; revised manuscript received and accepted February 12, 2014. Drs. A.M. Saguner and S. Ganahl contributed equally to this article and are shared first authors. This work and the Zurich ARVC Program are supported by a grant from the Georg and Bertha Schwyzer-Winiker Foundation, Zurich, Switzerland. See page 1733 for disclosure information. *Corresponding author: Tel: (þ41) 44 2552099; fax: (þ41) 44 2554401. E-mail address: fi[email protected] (F. Duru).
0002-9149/14/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjcard.2014.02.031
The study population included 111 patients from 3 tertiary care centers in Switzerland with a possible, borderline, or definite diagnosis of ARVD (possible diagnosis: 2 minor or 1 major criterion; borderline: 1 major and 1 minor or 3 minor criteria; and definite: 2 major, 1 major plus 2 minor, or 4 minor criteria according to the 2010 revised Task Force Criteria [TFC]),4 who had a 12-lead surface electrocardiogram (ECG) recorded from February 1987 to March 2013. Ninety-eight patients were index patients (probands) and 13 were family members. Clinical information regarding demographics and symptoms were obtained from hospital records at the time of the electrocardiography. This observational study was approved by the local institutional ethical committees of the participating centers. Electrocardiographic www.ajconline.org
Cardiomyopathy/ECG Predicts Outcome in ARVD
1729
Table 1 Definitions of electrocardiographic parameters and measurements used in this study Variable Epsilon (e) wave TAD TWI Inferior leads TWI Major repolarization criteria Minor repolarization criteria Precordial QRS amplitude ratio Parietal block Fragmented QRS Peripheral low voltage QRS dispersion QT dispersion Precordial TpeakeTend interval Precordial TpeakeTend dispersion Complete right bundle branch block
Incomplete right bundle branch block R or r0 in V1/V2
Definition Distinct waves of small amplitude that occupy the ST segment in the right precordial leads and are distinct from the QRS complex Longest value in V1 through V3 from the nadir of the S wave to the end of all depolarization Any T-wave negativity Inverted T waves in 2 of 3 inferior leads Definition according to 2010 TFC Definition according P to 2010 TFC P QRS mV V1V3/ QRS mV V1eV6 QRSd in lead V1 through V3 that exceeds the QRSd in lead V6 by >25 ms Additional deflections/notches at the beginning of the QRS, on top of the R wave, or in the nadir of the S wave in either 1 right precordial lead or in >1 lead including all remaining leads QRS amplitude 5 mm (0.5 mV) in each peripheral lead Difference between the longest and shortest QRS durations in all leads Difference between the longest and shortest uncorrected QT intervals in all leads Difference of the interval of the Q wave to the end of the T wave (QTend) and the interval of the Q wave to the peak of the T wave (QTpeak) measured in the precordial leads, representing a surrogate marker of the transmural dispersion of repolarization Difference between the longest and shortest precordial TpeakeTend interval QRSd 120 ms and A1: R0 or r0 in V1 or V2 A2: S duration > R duration in I and V6 A3: S duration >40 ms in I and V6 A4: R peak time >50 ms in V1 or V2 a: A1 þ A2 b: A1 þ A3 c: A4 þ (A2 or A3) QRS <120 ms and R peak time in V1 or V2 >50 ms A positive deflection in V1/V2 after an S wave
Figure 1. (A) A representative 12-lead surface ECG (25 mm/s, 10 mm/mV) from a patient with definite ARVD and adverse outcome showing measurements for calculation of the precordial QRS amplitude ratio in the precordial leads. This patient had poor R-wave progression in the right precordial leads resulting in a P P precordial QRS amplitude ratio ( QRS mV V1 to V3/ QRS mV V1 to V6) of 0.27 (0.48). QRS voltage amplitudes are marked in blue in leads V1 to V3 and in red in leads V4 to V6. Please note that this ECG shows a major repolarization criterion (TWIs V1 to V4) and a minor depolarization criterion (TAD 55 ms) according to the 2010 revised TFC. (B) The precordial leads V1 to V3 taken from a 12-lead surface ECG (25 mm/s, 10 mm/mV) demonstrate extensive QRS fragmentation (black arrows) in this patient with definite ARVD and adverse outcome. (C) A representative 12-lead surface ECG (25 mm/s, 10 mm/mV) from a patient with definite ARVD showing inferior leads TWIs in III and aVF (black arrows on the left) and in V1 to V3 (black arrows on the right), fulfilling a major repolarization criterion according to the 2010 revised TFC.
definitions used in this study are listed in Table 1. ECGs were recorded at rest (25 mm/s, 10 mm/mV) with standard lead positions, digitized with a high-resolution scanner, and
analyzed with a digital caliper (Screen Caliper, version 4.0; Iconico, www.iconico.com, New York, New York).9 ECGs were enlarged 4 times. Electrocardiographic intervals were
1730
The American Journal of Cardiology (www.ajconline.org)
Table 2 Baseline clinical characteristics Characteristic Age (yrs) Men Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (beats/min) Body surface area (m2) Body mass index (kg/m2) LV ejection fraction <50% LV ejection fraction <40% Previous sustained VT/VF/sudden cardiac arrest Fractional area change <33% Previous syncope “Definite” ARVD Premature SCD in a first- or second-degree family member Amiodarone b Blocker Sotalol Other antiarrhythmic therapy
All Patients (n ¼ 111)
MACE (n ¼ 51)
Favorable (n ¼ 60)
p Value
43 15.4 71 (64) 120 (110e129) 76 9 65 (57e75) 1.9 (1.7e2.0) 24.5 (21.8e26.1) 15 (14) 8 (7) 52 (47) 48 (43) 35 (32) 80 (72) 11 (10) 18 (16) 48 (43) 13 (12) 9 (8)
46 14.9 33 (65) 120 (110e126) 77 9 63 (55e79) 1.9 (1.7e2.0) 24.5 (21.4e26.2) 12 (24) 7 (14) 35 (69) 35 (69) 24 (47) 44 (86) 5 (10) 11 (22) 29 (57) 7 (14) 6 (12)
40 15.3 38 (63) 120 (110e131) 76 10 65 (57e73) 1.9 (1.7e2.0) 24.3 (22.1e26.1) 3 (5) 1 (2) 17 (28) 16 (27) 11 (18) 36 (60) 6 (10) 6 (10) 18 (30) 6 (10) 3 (5)
0.03 1.00 0.12 0.4 0.39 0.3 0.76 0.005 0.023 <0.001 <0.001 0.016 0.003 1.00 0.11 0.007 0.57 0.3
Values are mean SD, median (interquartile range), or number (percentage). p Values were calculated by 2-sided unpaired Student t test for continuous variables and by Fisher’s exact test for categorical variables.
measured in 2 consecutive beats in each lead, and the mean value was used. When the difference between the 2 beats was >10 ms, then the mean of 3 beats was taken. ECGs were independently analyzed by 2 experienced blinded readers. Differences in electrocardiographic interpretation were adjudicated by a third experienced reader, and a final conclusion was made by consensus. Heart rate, PR interval, QRS duration, QRS dispersion, QT interval, QT dispersion, Tpeak to Tend duration, parietal block, and terminal activation duration (TAD) were measured according to current practice.10e15 The Bazett formula was used to correct the QT interval for heart rate. QRS fragmentation (Figure 1) and peripheral low voltage were defined as previously described.7,8,16e18 We also included a recently proposed parameter termed “precordial QRS amplitude ratio” in our analysis (Figure 1), which is the ratio of the sum of PQRS amplitudes inP V1 to V3 divided by the sum in V1 to V6 ( QRS mV V1 to V3/ QRS mV V1 to V6), with low values reflecting lower voltages in the right precordial leads compared with the lateral precordial leads.19 We also evaluated the impact of inferior leads T-wave inversions (TWIs; Figure 1).20 Repolarization abnormalities according to the 2010 TFC were defined as precordial TWI in at least V1 and V2 (minor criterion). Follow-up for outcome data was performed by chart review, including implantable cardioverter-defibrillator interrogations, Holter ECGs, clinical visits, and from telephone interviews of patients or treating physicians. No patient was lost to follow-up. The role of electrocardiographic parameters on outcome was compared in 2 patient groups: (1) patients with major adverse cardiovascular events (MACE; “adverse outcome”); composite of cardiac death (defined as any death that occurred because of SCD or terminal congestive heart failure), heart transplantation, survived SCD, ventricular fibrillation (VF), sustained ventricular tachycardia (VT), and arrhythmogenic syncope21 and (2) all remaining patients (“favorable outcome”). For Kaplan-Meier estimates and Cox regression analyses, time from electrocardiography to MACE
was the event of interest. Survived SCD, VF, sustained VT, and arrhythmogenic syncope were defined as previously reported.21,22 Continuous variables are presented as mean SD or median (with interquartile range) and were compared using a 2-sided unpaired Student t test. Categorical variables are reported as frequency (percentage) and were compared between groups by Fisher’s exact test. Cut-off points for continuous electrocardiographic variables that maximized the difference in times to MACE for patients with values lower and for those greater than the cut-off were determined by minimizing the p value (log-rank). The final p value was corrected for multiple testing. Cumulative probabilities of survival free of MACE were determined by the Kaplan-Meier method and differences in survival with the log-rank test. Baseline variables associated with MACE were identified by univariable Cox regression. Variables with a p value of <0.1 in univariable analysis were considered for the multivariable model. Patients with ARVD with a previous episode of sustained VT, VF, and/or sudden cardiac arrest have a high probability of MACE. Thus, we adjusted for this variable by performing a stratified analysis. For continuous variables, the intraclass correlation was used to quantify interobserver variability with values >0.75 representing good reliability of measurements. For nominal variables, we calculated the agreement coefficient k, with a value of 1 indicating perfect agreement. A 2-sided p value of <0.05 was considered significant. Statistical analysis was performed using R programing language (R Development Core Team, 2009) and GraphPad Prism 5 (GraphPad Software Inc., La Jolla, California). Results ECGs were available for all patients. Patient demographic data, baseline characteristics, and ECG variables are summarized in Tables 2 and 3. ARVD was classified as definite in 80 patients (72%), borderline in 15 patients (14%), and
Cardiomyopathy/ECG Predicts Outcome in ARVD Table 3 Twelve-lead electrocardiographic data Variable Heart rate (beats/min) RR interval (seconds) PQ interval (ms) QRS duration (ms) QRS dispersion (ms) QT interval (ms) QT interval corrected (ms) QT dispersion (ms) Precordial QRS ratio Precordial TpeakeTend interval (ms) Precordial TpeakeTend dispersion (ms) TAD (ms) TAD 55 ms Minor depolarization criteria Major depolarization criteria (epsilon waves) Fragmented QRS Isolated late potentials in inferior leads Major repolarization criteria Minor repolarization criteria TWIs in 3 precordial leads Inferior leads TWI TWIs in 3 precordial leads combined with inferior leads TWI Complete left bundle branch block Complete right bundle branch block Left anterior fascicular block Incomplete right bundle branch block Parietal block Premature ventricular complexes present on 12-lead ECG
All (n ¼ 111) 65 (57e75) 0.92 (0.8e1.1) 160 (144e184) 111 (100e127) 39 (28e51) 420 (392e465) 443 (425e480) 57 (46e82) 0.475 0.091 96.8 (80e107.4) 30.8 (21e46.8) 52 (43e64) 36 (32) 21 (19) 22 (20) 42 (38) 4 (4) 37 (33) 21 (19) 32 (29) 48 (43) 12 (11) 2 13 8 3 13 19
(2) (12) (7) (3) (12) (17)
Values are mean SD, median (interquartile range), or number (percentage). Definitions: see Table 1.
Table 4 Specification of major adverse cardiovascular events (MACE) during follow-up MACE
n ¼ 51
Cardiac death (SCD or congestive heart failure) Heart transplantation VF Sustained VT Arrhythmic syncope
3 4 8 34 2
(6) (8) (16) (67) (3)
Values are presented as number (percentage).
possible in 16 patients (14%). Eleven of 16 patients (69%) with a possible diagnosis were family members. Index cases with possible ARVD fulfilled 1 major (n ¼ 1) or 2 minor criteria (n ¼ 4). Two of those patients had suspected family disease, and 1 patient had RV scar identified by electroanatomic voltage mapping. After study inclusion, during a follow-up with median of 4.6 years (interquartile range 1.8 to 10), 51 patients (46%) experienced MACE (Table 4). Median time from study inclusion to MACE was 12 months (interquartile range 4.1 to 36). At last follow-up, 108 patients (97%) were alive. Three patients died from cardiac causes (2 because of terminal congestive heart failure and 1 because of
1731
incessant VT), and cardiac transplantation was performed in 4 patients (3 because of terminal congestive heart failure and 1 because of repetitive uncontrollable implantable cardioverter-defibrillator shocks). Histology of all 4 explanted hearts showed extensive fibrofatty infiltration within the RV myocardium and the posterolateral aspect of the left ventricle (LV). There was no evidence of myocarditis or superimposed inflammatory phenomena in any of those cases. Analysis of times to MACE showed that the presence of precordial repolarization abnormalities according to the 2010 TFC, a precordial QRS amplitude ratio of 0.48, and QRS fragmentation were associated with significantly reduced times to MACE (Figure 2). Detailed data for the 2 subgroups with (n ¼ 52) and without (n ¼ 59) SCD, VF, and/or sustained VT before study inclusion are visualized by the stratified KaplanMeier plots. In general, times to MACE were longer in patients without previous SCD, VF, and/or sustained VT. After stratification for this observation, TWI in 2 precordial leads and QRS fragmentation were significantly associated with shorter survival, also in the subgroup of patients considered for primary prophylactic implantable cardioverter-defibrillator implantation (Figure 2). The number of precordial leads with TWI was significantly associated with MACE (3 leads vs >3, p ¼ 0.017). On the contrary, presence of depolarization abnormalities, parietal block, peripheral low voltage, QT prolongation, prolonged QRS, QT and/or Tpeak to Tend dispersion, and Tpeak to Tend interval prolongation did not significantly alter times to MACE. Precordial repolarization criteria, a precordial QRS amplitude ratio of 0.48, TWI in >3 precordial leads, QRS fragmentation, and inferior leads TWI were identified as univariable predictors of MACE also in the Cox model (Table 5). In multivariable analysis starting with these univariable predictors, a precordial QRS amplitude ratio of 0.48 (hazard ratio [HR] 2.92, 95% confidence interval [CI] 1.39 to 6.15, p ¼ 0.005), inferior leads TWI (HR 2.44, 95% CI 1.15 to 5.18, p ¼ 0.02), and QRS fragmentation (HR 2.65, 95% CI 1.1 to 6.34, p ¼ 0.029) remained as the only independent predictors of MACE. Separate analyses excluding 5 index patients with possible ARVD yielded similar results (Table 5). In this subpopulation of 106 patients, multivariable analysis showed the following results: HR 3.45 (95% CI 1.54 to 7.74, p ¼ 0.003) for precordial QRS amplitude ratio 0.48; HR 2.65 (95% CI 1.22 to 5.74, p ¼ 0.014) for inferior leads TWI; and HR 2.31 (95% CI 0.94 to 5.67, p ¼ 0.067) for QRS fragmentation. Kaplan-Meier subanalyses for patients with definite and borderline ARVD without previous SCD, VF, and/or sustained VT also demonstrated significantly lower survival in the presence of precordial repolarization criteria and QRS fragmentation (Figure 2). In this same subpopulation, Cox regression identified the presence of QRS fragmentation (HR 4.44, 95% CI 1.52 to 12.95, p ¼ 0.006) and precordial repolarization criteria (HR 4.4, 95% CI 1.17 to 16.56, p ¼ 0.029) as univariable predictors of MACE. The low number of MACE (n ¼ 15) in this subpopulation did not allow for multivariable analysis. The estimated intraclass correlation was 0.9 (95% CI 0.74 to 0.94) for precordial QRS amplitude ratio (k of 0.94 [95% CI 0.85 to 1.0]) for the cutoff 0.48. Interobserver correlations were higher for repolarization than for depolarization abnormalities (k ¼ 0.89 [95% CI 0.8 to 0.97] for repolarization criteria and k ¼ 0.95 [95% CI 0.9 to 1.0] for
1732
The American Journal of Cardiology (www.ajconline.org)
Figure 2. (A) Kaplan-Meier analysis of times to MACE (composite of cardiac death, heart transplant, SCD, VF, sustained VT, or arrhythmic syncope; n ¼ 111) for the time between electrocardiography and end point, stratified by repolarization criteria according to the 2010 revised TFC (left diagram), a precordial QRS amplitude ratio of 0.48 (middle diagram), and QRS fragmentation (right diagram); all analyses were stratified for the presence of survived SCD, VF, and/or sustained VT before study inclusion (black graphs: patients without survived SCD, VF, and/or sustained VT before study inclusion; purple graphs: patients with survived SCD, VF, and/or sustained VT before study inclusion). (B) Kaplan-Meier analysis of times to MACE for the time between electrocardiography and end point in the subpopulation of patients (n ¼ 44) without survived SCD, VF, and/or sustained VT before study inclusion and having a definite or borderline diagnosis of ARVD; analyses stratified by repolarization criteria according to the 2010 revised TFC (left diagram) and QRS fragmentation (right diagram). strat ¼ stratified.
inferior leads TWI, compared with k ¼ 0.5 [95% CI 0.33 to 0.66] for QRS fragmentation, k ¼ 0.49 [95% CI 0.25 to 0.69] for e waves in leads V1 to V3, k ¼ 0.38 [95% CI 0.07 to 0.6] for e waves in lead V1, k ¼ 0.36 [95% CI 0.03 to 0.7] for e waves in lead V2, k ¼ 0.24 [95% CI 0.06 to 0.6] for e waves in lead V3, and k ¼ 0.57 [95% CI 0.42 to 0.69] for the presence of TAD 55 ms). Discussion This long-term observational study suggests that 12-lead surface electrocardiographic parameters are useful to predict adverse events in ARVD based on the following findings: (1) precordial and inferior leads TWI predict MACE, (2) QRS fragmentation and the precordial QRS amplitude ratio constitute additional predictors of MACE; and (3) variability in interpretation of depolarization abnormalities may limit their clinical relevance, whereas a high interobserver correlation was found for repolarization abnormalities and the precordial QRS amplitude ratio. There is a need for noninvasive tools such as the 12-lead electrocardiography for risk stratification in ARVD.4 Previous studies about the predictive role of electrocardiography in ARVD have yielded
conflicting results.5e8 Earlier studies focused on depolarization abnormalities. Although results were not consistent,23,24 most of these studies have shown that QRS and/or QT dispersion and other depolarization abnormalities may be useful for risk prediction in ARVD.5,6,25 Recent studies with larger patient numbers and less advanced ARVD have investigated the predictive role of precordial depolarization or repolarization abnormalities, and results were again not consistent.21,23,26 A plausible explanation for these diverging findings may be related to different demographics and heterogeneous study protocols. Importantly, earlier studies reporting continuous electrocardiographic variables did not use digital calipers, which may limit the validity of these data. We show that precordial and inferior leads TWI have predictive value in ARVD, whereas depolarization abnormalities and continuous electrocardiographic variables such as e waves, TAD, and QRS, QT, and/or Tpeak to Tend dispersion only have limited predictive value, although prolonged TAD and e waves are accepted delay parameters in ARVD. A main limitation of these measures is their high interobserver variability. A higher interobserver correlation may partially explain why fragmented QRS was more valuable for risk prediction than e wave. However, interobserver variability
Cardiomyopathy/ECG Predicts Outcome in ARVD
1733
Table 5 Electrocardiographic variables associated with major adverse cardiovascular events (MACE) in the whole study population (n ¼ 111, left column) and in the subpopulation excluding index patients with possible disease (n ¼ 106, right column) Variable
Repolarization criteria according to 2010 TFC Precordial QRS amplitude ratio 0.48 TWIs >3 precordial leads QRS fragmentation TWIs in 2 inferior leads Late potentials in 1 lead Epsilon waves TpeakeTend duration >100 ms Bundle branch block TpeakeTend dispersion >20 ms Peripheral low voltage Presence of premature ventricular complexes on 12-lead ECG Parietal block TAD 55 ms
Univariable Analysis (n ¼ 111)
Univariable Analysis (n ¼ 106)
HR (95% CI)
HR (95% CI)
2.4 2.31 1.88 1.76 1.73 1.63 1.61 1.47 0.7 1.36 1.29 1.2 1.07 1.07
(1.22e4.73) (1.13e4.74) (1.07e3.31) (1.01e3.06) (0.98e3.06) (0.89e3.01) (0.86e3.03) (0.83e2.61) (0.38e1.32) (0.69e2.67) (0.63e2.66) (0.61e2.35) (0.57e1.99) (0.57e1.99)
p Value 0.011 0.022 0.029 0.047 0.058 0.11 0.14 0.19 0.27 0.38 0.48 0.6 0.83 0.83
2.49 2.7 1.87 1.69 1.79 1.75 1.71 1.4 0.71 1.48 1.33 1.24 1.05 1.1
(81.23e5.04) (1.24e5.86) (1.05e3.34) (0.95e3.03) (1.00e3.21) (0.94e3.25) (0.9e3.23) (0.77e2.55) (0.31e1.6) (0.71e3.1) (0.64e2.75) (0.63e2.46) (0.58e1.9) (0.58e2.08)
p Value 0.012 0.012 0.034 0.075 0.05 0.076 0.1 0.27 0.41 0.3 0.44 0.53 0.88 0.78
p Values were calculated by univariable Cox regression, stratified for the presence of sudden cardiac arrest, VF, and/or sustained VT before study inclusion.
does not explain the incremental prognostic value of fragmented QRS compared with TAD, as k values were similar for both. We believe that the incremental prognostic value of fragmented QRS is also explained by its greater prevalence, and that fragmented QRS—determined in all 12 electrocardiographic leads—visualizes conduction delay within more aspects of the myocardium, as it is not limited to only leads V1 to V3. Furthermore, fragmented QRS visualizes conduction delay during all phases of ventricular depolarization, and not only at later phases. Our study underlines the clinical utility of precordial TWI to predict MACE in ARVD. This study confirms that an incremental relation between precordial TWI and MACE exists, with >3 inverted precordial T waves conferring an almost twofold increased hazard compared with 3 precordial TWIs. An important and novel finding is the independent prognostic value of inferior leads TWIs, which have been linked to LV involvement, and thus may explain their predictive utility.20 In multivariable analysis, inferior leads TWI remained an independent predictor of MACE. These findings agree with the important predictive role of repolarization abnormalities in heart failure and myocardial infarction.27 These reliable and easily recognized abnormalities seem to constitute key elements of arrhythmic risk in patients with ARVD, probably because they are associated with severe RV disease as well as LV involvement.5,28 Another novel finding of our study is the predictive role of a reduced precordial QRS amplitude ratio, which remained the second independent predictor of MACE in multivariable analysis. This ratio at a recently proposed cutoff of 0.48 has been shown to predict VT recurrence after successful ablation of VT in patients with ARVD with electrical storm.19 Importantly, this ratio is easy to measure and reliable. The third independent predictor of MACE was the presence of QRS fragmentation.7,8 Importantly, besides repolarization criteria, this variable was the only significant predictor of MACE in our lower risk subpopulation of patients with definite and borderline ARVD without previous SCD, VF, and/or sustained VT. Previous work has suggested
that myocardial scar may be reflected by electrocardiographic QRS fragmentation and thus confers a poor prognosis in coronary artery disease.29 Two recent studies with smaller patient numbers have demonstrated this association in patients with ARVD. Yet, the predictive value of other electrocardiographic variables was not assessed in those studies, and the multivariable model by Canpolat et al8 was probably overfitted.7,8 Of note, interobserver variability for QRS fragmentation was rather high in our cohort, which may limit the clinical utility of this variable. This study is limited by its observational nature inherent to registries. Thus, inclusion of patients with ARVD at different stages of the disease was inevitable, although we made an effort to include patients early after first medical contact. We included patients with possible ARVD, although their phenotype may not be typical. Yet, those patients form an integral part of our routine clinical practice, and a considerable number of patients with early ARVD may be missed if only patients with advanced phenotype are considered. In general, we studied a rather high-risk cohort of patients given the high MACE rates reflecting a referral bias and indicating the need for confirmation of our results in lower risk populations, for example, in family members who generally present with a less severe phenotype.30 Therefore, our results mainly obtained in index patients may not be transferable to family members. As our ARVD program is emerging, we currently do not have sufficient genetic data to provide, and because of low patient numbers, our data do not allow subanalyses of different ethnicities. Thus, we may have missed important genotypephenotype correlations that could be associated with our findings. As such, our suggested cutoffs for the precordial QRS amplitude ratio may not be applicable to patients with desmoplakin and phospholamban mutations, who often present with LV involvement. Disclosures The authors have no conflicts of interest to disclose.
1734
The American Journal of Cardiology (www.ajconline.org)
1. Roberts WC, Ko JM, Kuiper JJ, Hall SA, Meyer DM. Some previously neglected examples of arrhythmogenic right ventricular dysplasia/cardiomyopathy and frequency of its various reported manifestations. Am J Cardiol 2010;106:268e274. 2. Kim C, Wong J, Wen J, Wang S, Wang C, Spiering S, Kan NG, Forcales S, Puri PL, Leone TC, Marine JE, Calkins H, Kelly DP, Judge DP, Chen HS. Studying arrhythmogenic right ventricular dysplasia with patient-specific iPSCs. Nature 2013;494:105e110. 3. Schuler PK, Haegeli LM, Saguner AM, Wolber T, Tanner FC, Jenni R, Corti N, Luscher TF, Brunckhorst C, Duru F. Predictors of appropriate ICD therapy in patients with arrhythmogenic right ventricular cardiomyopathy: long term experience of a tertiary care center. PLoS One 2012;7:e39584. 4. Marcus FI, McKenna WJ, Sherrill D, Basso C, Bauce B, Bluemke DA, Calkins H, Corrado D, Cox MG, Daubert JP, Fontaine G, Gear K, Hauer R, Nava A, Picard MH, Protonotarios N, Saffitz JE, Sanborn DM, Steinberg JS, Tandri H, Thiene G, Towbin JA, Tsatsopoulou A, Wichter T, Zareba W. Diagnosis of arrhythmogenic right ventricular cardiomyopathy/dysplasia: proposed modification of the Task Force Criteria. Eur Heart J 2010;31:806e814. 5. Peters S, Peters H, Thierfelder L. Risk stratification of sudden cardiac death and malignant ventricular arrhythmias in right ventricular dysplasia-cardiomyopathy. Int J Cardiol 1999;71:243e250. 6. Turrini P, Corrado D, Basso C, Nava A, Bauce B, Thiene G. Dispersion of ventricular depolarization-repolarization: a noninvasive marker for risk stratification in arrhythmogenic right ventricular cardiomyopathy. Circulation 2001;103:3075e3080. 7. Peters S, Truemmel M, Koehler B. Prognostic value of QRS fragmentation in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia. J Cardiovasc Med (Hagerstown) 2012;13:295e298. 8. Canpolat U, Kabakci G, Aytemir K, Dural M, Sahiner L, Yorgun H, Sunman H, Baris Kaya E, Tokgozoglu L, Oto A. Fragmented QRS complex predicts the arrhythmic events in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia. J Cardiovasc Electrophysiol 2013;24:1260e1266. 9. Roten L, Derval N, Sacher F, Pascale P, Scherr D, Komatsu Y, Ramoul K, Daly M, Denis A, Shah AJ, Hocini M, Jais P, Haissaguerre M. Heterogeneous response of J-wave syndromes to beta-adrenergic stimulation. Heart Rhythm 2012;9:1970e1976. 10. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C, Krikler D, Mazur NA, Meijler FL, Mogensen L, Moret P. Criteria for intraventricular conduction disturbances and pre-excitation. World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261e1275. 11. Fontaine G, Fontaliran F, Hebert JL, Chemla D, Zenati O, Lecarpentier Y, Frank R. Arrhythmogenic right ventricular dysplasia. Annu Rev Med 1999;50:17e35. 12. Antzelevitch C, Yan GX, Shimizu W. Transmural dispersion of repolarization and arrhythmogenicity: the Brugada syndrome versus the long QT syndrome. J Electrocardiol 1999;32(Suppl):158e165. 13. Surawicz B, Childers R, Deal BJ, Gettes LS, Bailey JJ, Gorgels A, Hancock EW, Josephson M, Kligfield P, Kors JA, Macfarlane P, Mason JW, Mirvis DM, Okin P, Pahlm O, Rautaharju PM, van Herpen G, Wagner GS, Wellens H; American Heart Association E, Arrhythmias Committee CoCC, American College of Cardiology F, Heart Rhythm S. AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part III: intraventricular conduction disturbances: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. Endorsed by the International Society for Computerized Electrocardiology. J Am Coll Cardiol 2009;53:976e981. 14. Rautaharju PM, Surawicz B, Gettes LS, Bailey JJ, Childers R, Deal BJ, Gorgels A, Hancock EW, Josephson M, Kligfield P, Kors JA, Macfarlane P, Mason JW, Mirvis DM, Okin P, Pahlm O, van Herpen G, Wagner GS, Wellens H; American Heart Association E, Arrhythmias Committee CoCC, American College of Cardiology F, Heart Rhythm S. AHA/ ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society: endorsed by the International Society for Computerized Electrocardiology. Circulation 2009;119:e241ee250.
15. Cox MG, Nelen MR, Wilde AA, Wiesfeld AC, van der Smagt JJ, Loh P, Cramer MJ, Doevendans PA, van Tintelen JP, de Bakker JM, Hauer RN. Activation delay and VT parameters in arrhythmogenic right ventricular dysplasia/cardiomyopathy: toward improvement of diagnostic ECG criteria. J Cardiovasc Electrophysiol 2008;19:775e781. 16. Rapezzi C, Merlini G, Quarta CC, Riva L, Longhi S, Leone O, Salvi F, Ciliberti P, Pastorelli F, Biagini E, Coccolo F, Cooke RM, BacchiReggiani L, Sangiorgi D, Ferlini A, Cavo M, Zamagni E, Fonte ML, Palladini G, Salinaro F, Musca F, Obici L, Branzi A, Perlini S. Systemic cardiac amyloidoses: disease profiles and clinical courses of the 3 main types. Circulation 2009;120:1203e1212. 17. Mussinelli R, Salinaro F, Alogna A, Boldrini M, Raimondi A, Musca F, Palladini G, Merlini G, Perlini S. Diagnostic and prognostic value of low QRS voltages in cardiac AL amyloidosis. Ann Noninvasive Electrocardiol 2013;18:271e280. 18. Peters S, Trummel M, Koehler B. QRS fragmentation in standard ECG as a diagnostic marker of arrhythmogenic right ventricular dysplasiacardiomyopathy. Heart Rhythm 2008;5:1417e1421. 19. Muessigbrodt A, Dinov L, Fiedler L, Sommer P, Richter S, Breithardt O, Bollmann A, Husser D, Hindricks G, Arya A. Precordial QRS amplitude ratio predicts long term outcome after catheter ablation of electrical storm due to ventricular tachycardias in patients with arrhythmogenic right ventricular cardiomyopathy [abstract]. Europace 2013;15(Suppl 2):ii82eii103.22. 20. Jain R, Dalal D, Daly A, Tichnell C, James C, Evenson A, Abraham T, Tan BY, Tandri H, Russell SD, Judge D, Calkins H. Electrocardiographic features of arrhythmogenic right ventricular dysplasia. Circulation 2009;120:477e487. 21. Corrado D, Calkins H, Link MS, Leoni L, Favale S, Bevilacqua M, Basso C, Ward D, Boriani G, Ricci R, Piccini JP, Dalal D, Santini M, Buja G, Iliceto S, Estes NA 3rd, Wichter T, McKenna WJ, Thiene G, Marcus FI. Prophylactic implantable defibrillator in patients with arrhythmogenic right ventricular cardiomyopathy/dysplasia and no prior ventricular fibrillation or sustained ventricular tachycardia. Circulation 2010;122:1144e1152. 22. Saguner AM, Medeiros-Domingo A, Schwyzer MA, On CJ, Haegeli LM, Wolber T, Hurlimann D, Steffel J, Krasniqi N, Rueger S, Held L, Luscher TF, Brunckhorst C, Duru F. Usefulness of inducible ventricular tachycardia to predict long-term adverse outcomes in arrhythmogenic right ventricular cardiomyopathy. Am J Cardiol 2013;111:250e257. 23. Hulot JS, Jouven X, Empana JP, Frank R, Fontaine G. Natural history and risk stratification of arrhythmogenic right ventricular dysplasia/ cardiomyopathy. Circulation 2004;110:1879e1884. 24. Roguin A, Bomma CS, Nasir K, Tandri H, Tichnell C, James C, Rutberg J, Crosson J, Spevak PJ, Berger RD, Halperin HR, Calkins H. Implantable cardioverter-defibrillators in patients with arrhythmogenic right ventricular dysplasia/cardiomyopathy. J Am Coll Cardiol 2004;43: 1843e1852. 25. Lemola K, Brunckhorst C, Helfenstein U, Oechslin E, Jenni R, Duru F. Predictors of adverse outcome in patients with arrhythmogenic right ventricular dysplasia/cardiomyopathy: long term experience of a tertiary care centre. Heart 2005;91:1167e1172. 26. Bhonsale A, James CA, Tichnell C, Murray B, Madhavan S, Philips B, Russell SD, Abraham T, Tandri H, Judge DP, Calkins H. Risk stratification in arrhythmogenic right ventricular dysplasia/cardiomyopathyassociated desmosomal mutation carriers. Circ Arrhythm Electrophysiol 2013;6:569e578. 27. Zabel M, Acar B, Klingenheben T, Franz MR, Hohnloser SH, Malik M. Analysis of 12-lead T-wave morphology for risk stratification after myocardial infarction. Circulation 2000;102:1252e1257. 28. Steriotis AK, Bauce B, Daliento L, Rigato I, Mazzotti E, Folino AF, Marra MP, Brugnaro L, Nava A. Electrocardiographic pattern in arrhythmogenic right ventricular cardiomyopathy. Am J Cardiol 2009;103:1302e1308. 29. Das MK, Suradi H, Maskoun W, Michael MA, Shen C, Peng J, Dandamudi G, Mahenthiran J. Fragmented wide QRS on a 12-lead ECG: a sign of myocardial scar and poor prognosis. Circ Arrhythm Electrophysiol 2008;1:258e268. 30. Saguner AM, Vecchiati A, Baldinger SH, Rüeger S, Medeiros-Domingo A, Mueller-Burri AS, Haegeli LM, Biaggi P, Manka R, Lüscher TF, Fontaine G, Delacrétaz E, Jenni R, Held L, Brunckhorst C, Duru F, Tanner FC. Different prognostic value of functional right ventricular parameters in arrhythmogenic right ventricular cardiomyopathy/ dysplasia. Circ Cardiovasc Imaging 2014;7:230e239.