DEVICES
Lesion-Specific Differences for Implantable Cardioverter Defibrillator Therapies in Adults with Congenital Heart Disease DANESH K. KELLA, M.D.,* FAISAL M. MERCHANT, M.D.,* EMIR VELEDAR, PH.D.,*,†,‡ WENDY BOOK, M.D.,*,§ and MICHAEL S. LLOYD, M.D.*,§ From the *Department of Medicine, Division of Cardiology, Emory University, Atlanta, Georgia; †Rollins School of Public Health, Emory University, Atlanta, Georgia; ‡Robert Stempel College of Public Health and Social Work, Department of Biostatistics, Florida International University, Miami, Florida; and §Emory Center for Adults with Congenital Heart Disease, Atlanta, Georgia
Background: Sudden cardiac death is a major cause of late mortality in adults with congenital heart disease (ACHD). While data exist for adults with repaired Tetralogy of Fallot (TOF), little is known about those with non-TOF lesions. We examined the relative rates in implantable cardioverter defibrillator (ICD) therapy according to congenital lesion type in a large-volume adult congenital heart center. Methods: A cohort of 59 individuals (median follow up time, 3.2 years range 0–10) with ACHD and ICDs was stratified according to underlying congenital lesion and implant indication. Appropriate therapies were defined as any therapy for a physician-adjudicated ventricular arrhythmia. Rates of inappropriate and appropriate ICD therapies were analyzed according to several relevant clinical variables. Results: Thirty-three (56%) TOF, 15 (25.4%) L- or D-transposition of great arteries, and 11 (18.6%) with other lesions were included in the analysis. Approximately half (52.5%) were implanted for primary prevention indications. During follow-up, 12 (20.3%) patients received appropriate ICD therapies and 13 (22%) patients received inappropriate therapies. The incidence of appropriate shocks among patients with TOF was 27.3% (9/33) compared to 11.5% (3/26) among non-TOF diagnoses during the follow-up time (p = 0.043). Conclusions: ACHD patients with non-TOF congenital lesions are significantly less likely to receive appropriate ICD therapy than those with TOF. Our analysis calls into question the validity of traditional ICD implantation guidelines in this growing and diverse patient population. (PACE 2014; 37:1492–1498) implantable cardioverter defibrillator, adults with congenital heart disease Introduction Surgical and medical advances have dramatically improved survival in those with congenital heart disease.1 There are now more adults living in the United States with congenital heart disease (ACHD) than children.2 Nonetheless, a leading cause of late mortality in the ACHD population is sudden cardiac death (SCD).3 The mechanism of SCD is thought to be primarily from malignant ventricular arrhythmias.4 Therefore, implantable cardioverter-defibrillators (ICDs) have been used in this patient population. While role of primary and secondary prevention
Disclosures: None. Address for reprints: Michael S. Lloyd M.D., F.A.C.C., F.H.R.S., Department of Electrophysiology, Emory University Hospital, 1364 Clifton Road, NE Suite F424, Atlanta, GA 30322. Fax: 404-712-4063, e-mail: [email protected] Received September 30, 2013; revised April 2, 2014; accepted April 27, 2014. doi: 10.1111/pace.12434
ICD therapy has been well established in systolic heart failure without underlying congenital heart disease, ICD use is less clear in ACHD.5 A relatively large body of work exists for those with Tetralogy of Fallot (TOF), but very little is known about ICDs in other, non-TOF lesions, such as Lor D-ransposition of the great vessels, lesions with single-ventricle physiology, and other complex lesions.6–8 In many centers, published data have been extrapolated from the SCD in Heart Failure Trial (SCD-HeFT) and the Multicenter Automatic Defibrillator Implantation Trial II (MADIT II) populations and have been applied to the remainder of the ACHD population.9,10 However, the application of traditional ICD implantation practices may not be valid, especially in the nonTOF population. A large proportion of adults with non-TOF lesions have subaortic right ventricles, and the propensity for arrhythmia in this context is largely unstudied. Recently, a multicenter registry developed a risk score for those with ACHD and ICDs.11 However, congenital lesion type was not accounted for.
©2014 Wiley Periodicals, Inc. 1492
November 2014
PACE, Vol. 37
ICD THERAPIES IN ACHD
Figure 1. Distribution of congenital heart disease in the study population. ASD = atrial septal defect; AV = atrioventricular; PV = pulmonary venous; RV = right ventricle; TGA = transposition of great arteries; TOF = tetralogy of Fallot.
We examined our ACHD cohort with ICDs to determine if there were differences in outcomes between adults with TOF and non-TOF lesions with respect to ICD therapies. Materials and Methods All patients aged ࣙ18 with ACHD followed by Emory Congenital Heart Center with an ICD were included in the study. The various congenital lesions included in the analysis are represented in Figure 1. Patients who had an intrinsic electrophysiologic or primary myopathic disorders (such as long QT syndrome, arrhythmogenic right ventricular dysplasia, noncompaction) were excluded from the study due to inherent differences in arrhythmia risk. The protocol for this study was reviewed and approved by Emory University’s Internal Review Board. Demographics, outside records (including operative notes), clinical labs, echocardiography, and electrocardiogram (ECG) and ICD data were abstracted and entered in a database. The latest echocardiogram near the time of ICD implantation was used to determine the subsystemic and subpulmonic ventricle ejection fraction. ECG parameters were determined similarly. Primary prevention indications were defined as ICD implantation for systemic ejection fraction ࣘ35% or inducible ventricular arrhythmia on electrophysiologic study. Secondary prevention was defined as ICD implantation after resuscitation from SCD.
PACE, Vol. 37
Table I. Baseline Characteristics of the Study Population (n = 59) Age (years) BMI (kg/m2 )
35.7 ± 9.4 27.6 ± 6.6
Sex, n (%) Male Female
41 (69.5) 18 (30.5)
Indication For ICD, n (%) Primary Secondary Missing data
31 (52.5) 23 (39.0) 5 (8.5)
Systemic Ventricle Ejection Fraction ࣘ 35%, n (%) Yes No Missing data
33 (55.9) 25 (42.4) 1 (1.7)
Subpulmonic Ventricle Function Moderate to Severely Reduced Yes No Missing data QRS (ms)
27 (45.8) 29 (49.1) 3 (5.1) 148.8.1 ± 33.4
BMI = body mass index; ICD = implantable cardioverter defibrillator.
November 2014
1493
KELLA, ET AL.
Table II. Differences between TOF and non-TOF Groups
Age (years) Sex (male) BMI (kg/m2 ) Primary prevention† Systemic EF ࣘ 35 %‡ Pulmonic ventricle EF moderate to severely reduced§ QRS (ms)
TOF Group (Mean ± SD or %) N = 33
Non-TOF Group (Mean ± SD or %) n = 26
38.4 ± 8.9 69.7 28.1 ± 6.8 46.7 30.3 62.5
32.3 ± 9.0 69.2 26.9 ± 6.3 70.8 60 29.2
0.013 0.97 0.45 0.074 0.024 0.013
157.5 ± 26.1
137.8 ± 38.5
0.031
P Value
BMI = body mass index; EF = ejection fraction; SD = standard deviation; TOF = Tetralogy of Fallot. † Data missing in n = 5. ‡ Data missing in n = 1. § Data missing in n = 3.
Device programming was at the discretion of the implanting physician and depended largely on indication type, not the congenital lesion. Appropriate therapy was defined as any ICD shock or pacing therapy delivered for termination of ventricular arrhythmia. Inappropriate therapy was defined as treatment for anything other than ventricular arrhythmia. Rhythm strips of events were reviewed by an electrophysiologist at the time of follow-up after therapy and documented in clinical notes. Where needed, remote data were obtained and rhythm strips reviewed by one of the authors blinded to the congenital lesion type. Statistical Analysis Continuous variables are reported as means with standard deviations and categorical variables as absolute numbers and percentages. Student’s ttest, Mann-Whitney U test, and χ 2 test were used to compare variables, where indicated. KaplanMeier curves were used to analyze time to appropriate or inappropriate shock. In time-toevent analysis, categorical variables, including type of lesion, indication for ICD, systemic ejection fraction (EF) ࣘ35%, and pulmonic ventricular function, were compared using the log-rank test and for continuous variables, age, and body mass index, Cox-regression was used. A P value of
Lesion-Specific Differences for Implantable Cardioverter Defibrillator Therapies in Adults with Congenital Heart Disease DANESH K. KELLA, M.D.,* FAISAL M. MERCHANT, M.D.,* EMIR VELEDAR, PH.D.,*,†,‡ WENDY BOOK, M.D.,*,§ and MICHAEL S. LLOYD, M.D.*,§ From the *Department of Medicine, Division of Cardiology, Emory University, Atlanta, Georgia; †Rollins School of Public Health, Emory University, Atlanta, Georgia; ‡Robert Stempel College of Public Health and Social Work, Department of Biostatistics, Florida International University, Miami, Florida; and §Emory Center for Adults with Congenital Heart Disease, Atlanta, Georgia
Background: Sudden cardiac death is a major cause of late mortality in adults with congenital heart disease (ACHD). While data exist for adults with repaired Tetralogy of Fallot (TOF), little is known about those with non-TOF lesions. We examined the relative rates in implantable cardioverter defibrillator (ICD) therapy according to congenital lesion type in a large-volume adult congenital heart center. Methods: A cohort of 59 individuals (median follow up time, 3.2 years range 0–10) with ACHD and ICDs was stratified according to underlying congenital lesion and implant indication. Appropriate therapies were defined as any therapy for a physician-adjudicated ventricular arrhythmia. Rates of inappropriate and appropriate ICD therapies were analyzed according to several relevant clinical variables. Results: Thirty-three (56%) TOF, 15 (25.4%) L- or D-transposition of great arteries, and 11 (18.6%) with other lesions were included in the analysis. Approximately half (52.5%) were implanted for primary prevention indications. During follow-up, 12 (20.3%) patients received appropriate ICD therapies and 13 (22%) patients received inappropriate therapies. The incidence of appropriate shocks among patients with TOF was 27.3% (9/33) compared to 11.5% (3/26) among non-TOF diagnoses during the follow-up time (p = 0.043). Conclusions: ACHD patients with non-TOF congenital lesions are significantly less likely to receive appropriate ICD therapy than those with TOF. Our analysis calls into question the validity of traditional ICD implantation guidelines in this growing and diverse patient population. (PACE 2014; 37:1492–1498) implantable cardioverter defibrillator, adults with congenital heart disease Introduction Surgical and medical advances have dramatically improved survival in those with congenital heart disease.1 There are now more adults living in the United States with congenital heart disease (ACHD) than children.2 Nonetheless, a leading cause of late mortality in the ACHD population is sudden cardiac death (SCD).3 The mechanism of SCD is thought to be primarily from malignant ventricular arrhythmias.4 Therefore, implantable cardioverter-defibrillators (ICDs) have been used in this patient population. While role of primary and secondary prevention
Disclosures: None. Address for reprints: Michael S. Lloyd M.D., F.A.C.C., F.H.R.S., Department of Electrophysiology, Emory University Hospital, 1364 Clifton Road, NE Suite F424, Atlanta, GA 30322. Fax: 404-712-4063, e-mail: [email protected] Received September 30, 2013; revised April 2, 2014; accepted April 27, 2014. doi: 10.1111/pace.12434
ICD therapy has been well established in systolic heart failure without underlying congenital heart disease, ICD use is less clear in ACHD.5 A relatively large body of work exists for those with Tetralogy of Fallot (TOF), but very little is known about ICDs in other, non-TOF lesions, such as Lor D-ransposition of the great vessels, lesions with single-ventricle physiology, and other complex lesions.6–8 In many centers, published data have been extrapolated from the SCD in Heart Failure Trial (SCD-HeFT) and the Multicenter Automatic Defibrillator Implantation Trial II (MADIT II) populations and have been applied to the remainder of the ACHD population.9,10 However, the application of traditional ICD implantation practices may not be valid, especially in the nonTOF population. A large proportion of adults with non-TOF lesions have subaortic right ventricles, and the propensity for arrhythmia in this context is largely unstudied. Recently, a multicenter registry developed a risk score for those with ACHD and ICDs.11 However, congenital lesion type was not accounted for.
©2014 Wiley Periodicals, Inc. 1492
November 2014
PACE, Vol. 37
ICD THERAPIES IN ACHD
Figure 1. Distribution of congenital heart disease in the study population. ASD = atrial septal defect; AV = atrioventricular; PV = pulmonary venous; RV = right ventricle; TGA = transposition of great arteries; TOF = tetralogy of Fallot.
We examined our ACHD cohort with ICDs to determine if there were differences in outcomes between adults with TOF and non-TOF lesions with respect to ICD therapies. Materials and Methods All patients aged ࣙ18 with ACHD followed by Emory Congenital Heart Center with an ICD were included in the study. The various congenital lesions included in the analysis are represented in Figure 1. Patients who had an intrinsic electrophysiologic or primary myopathic disorders (such as long QT syndrome, arrhythmogenic right ventricular dysplasia, noncompaction) were excluded from the study due to inherent differences in arrhythmia risk. The protocol for this study was reviewed and approved by Emory University’s Internal Review Board. Demographics, outside records (including operative notes), clinical labs, echocardiography, and electrocardiogram (ECG) and ICD data were abstracted and entered in a database. The latest echocardiogram near the time of ICD implantation was used to determine the subsystemic and subpulmonic ventricle ejection fraction. ECG parameters were determined similarly. Primary prevention indications were defined as ICD implantation for systemic ejection fraction ࣘ35% or inducible ventricular arrhythmia on electrophysiologic study. Secondary prevention was defined as ICD implantation after resuscitation from SCD.
PACE, Vol. 37
Table I. Baseline Characteristics of the Study Population (n = 59) Age (years) BMI (kg/m2 )
35.7 ± 9.4 27.6 ± 6.6
Sex, n (%) Male Female
41 (69.5) 18 (30.5)
Indication For ICD, n (%) Primary Secondary Missing data
31 (52.5) 23 (39.0) 5 (8.5)
Systemic Ventricle Ejection Fraction ࣘ 35%, n (%) Yes No Missing data
33 (55.9) 25 (42.4) 1 (1.7)
Subpulmonic Ventricle Function Moderate to Severely Reduced Yes No Missing data QRS (ms)
27 (45.8) 29 (49.1) 3 (5.1) 148.8.1 ± 33.4
BMI = body mass index; ICD = implantable cardioverter defibrillator.
November 2014
1493
KELLA, ET AL.
Table II. Differences between TOF and non-TOF Groups
Age (years) Sex (male) BMI (kg/m2 ) Primary prevention† Systemic EF ࣘ 35 %‡ Pulmonic ventricle EF moderate to severely reduced§ QRS (ms)
TOF Group (Mean ± SD or %) N = 33
Non-TOF Group (Mean ± SD or %) n = 26
38.4 ± 8.9 69.7 28.1 ± 6.8 46.7 30.3 62.5
32.3 ± 9.0 69.2 26.9 ± 6.3 70.8 60 29.2
0.013 0.97 0.45 0.074 0.024 0.013
157.5 ± 26.1
137.8 ± 38.5
0.031
P Value
BMI = body mass index; EF = ejection fraction; SD = standard deviation; TOF = Tetralogy of Fallot. † Data missing in n = 5. ‡ Data missing in n = 1. § Data missing in n = 3.
Device programming was at the discretion of the implanting physician and depended largely on indication type, not the congenital lesion. Appropriate therapy was defined as any ICD shock or pacing therapy delivered for termination of ventricular arrhythmia. Inappropriate therapy was defined as treatment for anything other than ventricular arrhythmia. Rhythm strips of events were reviewed by an electrophysiologist at the time of follow-up after therapy and documented in clinical notes. Where needed, remote data were obtained and rhythm strips reviewed by one of the authors blinded to the congenital lesion type. Statistical Analysis Continuous variables are reported as means with standard deviations and categorical variables as absolute numbers and percentages. Student’s ttest, Mann-Whitney U test, and χ 2 test were used to compare variables, where indicated. KaplanMeier curves were used to analyze time to appropriate or inappropriate shock. In time-toevent analysis, categorical variables, including type of lesion, indication for ICD, systemic ejection fraction (EF) ࣘ35%, and pulmonic ventricular function, were compared using the log-rank test and for continuous variables, age, and body mass index, Cox-regression was used. A P value of
