- Email: [email protected]
Preserved heart rate variability identifies low-risk patients with nonischemic dilated cardiomyopathy: Results from the DEFINITE trial
Preserved heart rate variability identifies low-risk patients with nonischemic dilated cardiomyopathy: Results from the DEFINITE trial Eric J. Rashba, MD,* N.A. Mark Estes, MD† Paul Wang, MD,‡ Andi Schaechter, BSN, RN,§ Adam Howard,§ Wojciech Zareba, MD, PhD,# Jean-Philippe Couderc, PhD,# Juha Perkiomaki, MD,** Joseph Levine, MD,†† Alan Kadish, MD,§ for the Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) Investigators From the *Division of Cardiology, Department of Medicine, University of Maryland at Baltimore, Baltimore, Maryland, † New England Medical Center, Boston, Massachusetts, ‡ Stanford University, Palo Alto, California, § Clinical Cardiology Trials Office, Division of Cardiology, Department of Medicine, Northwestern University Medical School, Chicago, Illinois, # University of Rochester Medical Center, Rochester, New York, ** Division of Cardiology, Department of Internal Medicine, University of Oulu, Oulu, Finland, and †† St. Francis Hospital, Roslyn, New York. BACKGROUND The recent expansion of indications for prophylactic implantable cardioverter-defibrillator (ICD) placement in subjects with nonischemic dilated cardiomyopathy has raised concerns about the cost-effectiveness of this therapy. OBJECTIVES The purpose of this study was to identify low-risk patients with nonischemic dilated cardiomyopathy who may not require prophylactic ICD placement. METHODS This was a prospective study of 274 participants in the Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial, a randomized controlled trial that evaluated the role of prophylactic ICD placement in patients with nonischemic dilated cardiomyopathy. The patients underwent 24-hour Holter recording for analysis of heart rate variability (HRV). The primary HRV variable was the standard deviation of normal R-R intervals (SDNN). Patients with atrial fibrillation and frequent ventricular ectopy (⬎25% of beats) were excluded from HRV analysis (23% of patients). SDNN was categorized in tertiles, and Kaplan-Meier analysis was performed to compare survival in the three tertiles and excluded patients. RESULTS The study population was 73% male, with a mean age of 59 ⫾ 12 years and mean left ventricular ejection fraction of 21% ⫾ 6%. After 3-year follow-up, significant differences in mortality rates were observed: SDNN ⬎113 ms: 0 (0%), SDNN 81–113 ms: 5 (7%), SDNN ⬍81 ms: 7 (10%), excluded patients: 11 (17%) (P ⫽ .03). There were no deaths in the tertile with SDNN ⬎113 ms regardless of treatment assignment (ICD vs control). CONCLUSION Patients with nonischemic dilated cardiomyopathy and preserved HRV have an excellent prognosis and may not benefit from prophylactic ICD placement. Patients with severely depressed HRV and patients who are excluded from HRV analysis because of atrial fibrillation and frequent ventricular ectopy have the highest mortality. KEYWORDS Heart rate variability; Implantable cardioverter-defibrillator; Sudden cardiac death (Heart Rhythm 2006;3:281–286) © 2006 Heart Rhythm Society. All rights reserved.
This research was supported by a K23 grant (HL-067198) from the National Heart, Lung, and Blood Institute to Dr. Rashba; and by St. Jude Medical, CRMD. Address reprint requests and correspondence: Dr. Eric J. Rashba, Division of Cardiology, University of Maryland Medical Center, 22 South Greene Street, Room N3W77, Baltimore, Maryland 21201. E-mail address: [email protected]. (Received October 14, 2005; accepted November 30, 2005.)
1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2005.11.028
282
Introduction Sudden cardiac death remains common in the United States, with an estimated 400,000 cases reported annually.1,2 Most patients who suffer an out-of-hospital cardiac arrest do not survive the initial event.1,2 Consequently, there has been great interest in identifying patients who are at risk for sudden cardiac death so that prophylactic implantable cardioverter-defibrillator (ICD) therapy can be instituted. The landmark DEFINITE (Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation),3 MADIT-2 (Multicenter Automatic Defibrillator Implantation Trial 2),4 and SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial)5 trials conclusively demonstrated that prophylactic ICD placement reduces mortality among subjects with ischemic and nonischemic cardiomyopathy. The scientific conclusions of these trials have not been seriously disputed, but there has been a vigorous debate regarding the cost implications of implementing their findings into clinical practice, as well as the potential unnecessary exposure of low-risk patients to the complications of ICD placement.6 As a result, the focus of risk stratification efforts has shifted to identifying low-risk patients who may not require prophylactic ICD placement. Heart rate variability (HRV) is a potent predictor of survival in subjects with congestive heart failure and systolic dysfunction.7,8 We hypothesized that nonischemic dilated cardiomyopathy patients with preserved HRV would have an excellent prognosis and designed a prospective study to test this hypothesis using the DEFINITE patient cohort.
Heart Rhythm, Vol 3, No 3, March 2006 sin-II receptor blockers. In addition, beta-blocker therapy was required unless patients were unable to tolerate it. The use of antiarrhythmic drugs such as amiodarone was discouraged. However, a small number of patients were treated with amiodarone to control symptomatic atrial arrhythmias. The DEFINITE trial commenced enrollment in July 1998. Funding for the digital Holter equipment was obtained after the trial was initiated, and the first recording was obtained in August 1999. The Holter recordings were obtained at baseline whenever possible or at the next scheduled follow-up visit for patients who were already enrolled in the study. The DEFINITE trial was approved by the Institutional Review Boards at participating centers, and written informed consent was obtained from all patients.
HRV analysis Burdick model 6632 digital Holter recorders were used for the study. The data were stored on 350-MB PC cards, which were shipped to the Holter Core Lab at the University of Maryland Medical Center for downloading and analysis. Each recording was manually edited to exclude segments with excessive noise and to correctly identify each beat. All Holter recordings were overread by the principal investigator of the study (E.J.R.). The recording was excluded from HRV analysis if atrial fibrillation or frequent ventricular ectopic beats (⬎25% of total beats) was present or if the recording lasted ⬍18 hours.11 The standard deviation of normal R-R intervals (SDNN) was prespecified as the primary HRV variable.11
Randomization and follow-up
Methods Study design The study design and primary results of the DEFINITE trial have been reported previously.9 In brief, DEFINITE was a prospective, randomized comparison of standard oral medical therapy for heart failure vs standard oral medical therapy plus an ICD. Inclusion criteria were left ventricular ejection fraction (LVEF) ⬍36%, presence of ambient arrhythmias,10 history of symptomatic heart failure, and presence of nonischemic dilated cardiomyopathy. Ambient arrhythmias were defined by an episode of nonsustained ventricular tachycardia on Holter or telemetry monitoring (3–15 beats at a rate ⬎120 bpm) or an average of at least 10 premature ventricular complexes per hour on 24-hour Holter monitoring. The absence of clinically significant coronary artery disease as the cause of the cardiomyopathy was confirmed by coronary angiography or by a negative stress imaging study. Detailed inclusion and exclusion criteria were previously reported.9 All patients received angiotensin-converting enzyme (ACE) inhibitors unless they were contraindicated. Patients who were unable to tolerate ACE inhibitors received hydralazine, nitrates, or angioten-
Patients were randomly assigned to receive either standard oral medical therapy for heart failure or standard oral medical therapy plus an ICD. Patients who were randomized to the ICD group received a single-chamber device that was programmed to detect ventricular fibrillation at a rate of 180 bpm. All patients were evaluated at 3-month intervals. The cause of death of patients who died was determined by an events committee whose members were unaware of patients’ treatment assignments. The cause of death was determined according to the criteria of Epstein et al.12 ICD shocks were classified as appropriate or inappropriate by a committee of participating investigators.
Statistical analysis The baseline characteristics of participants in the Holter substudy and nonparticipants were compared using unpaired t-tests for continuous variables and chi-square tests for categorical variables. The prespecified analytic plan was developed before acquisition of the data. The SDNN data were categorized into tertiles. Patients who had a Holter but were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy (⬎25% of beats) were
Rashba et al
HRV Identifies Low-Risk Patients with Nonischemic Dilated Cardiomyopathy
Figure 1 Flow of participants in the study. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
analyzed in a fourth group. The primary endpoint was total mortality. Secondary endpoints were cardiac mortality, appropriate ICD shocks, and the combined incidence of appropriate ICD shocks and sudden cardiac death. The stored electrograms for all ICD shocks were reviewed by a committee of investigators to determine if the shock was appropriate. The DEFINITE trial design specified that follow-up would continue until the 68th death occurred (May 25, 2003). Data on patients who received a heart transplant were censored at the time of transplantation. The duration of follow-up was computed from the date of the Holter to death for patients who died and to the date of the 68th death for patients who did not die. The same approach was used for analyses of secondary endpoints. Kaplan-Meier survival curves were used to estimate the cumulative percentage of patients surviving free from endpoint events over time. The incidence of endpoint events was calculated at 3-year follow-up. Comparisons between the survival curves were made using the log rank statistic using all available followup. Cox proportional hazards analysis was conducted to determine if the predictive value of the categorical SDNN variable was independent of age, gender, LVEF, and QRS duration. All reported P values are two-tailed.
Results A description of the patient flow for the study is shown in Figure 1. Of the 458 patients who participated in the DEFINITE trial, 303 (66%) had a digital Holter monitor performed. Thirty-one of the 40 DEFINITE enrolling centers contributed at least one patient to the study. Twenty-nine patients were excluded because fewer than 18 hours of analyzable data were present on the Holter recording, yielding 274 patients for further analysis. Of the 274 patients with a complete digital Holter recording, 16% had atrial fibrillation and 7% had frequent ventricular ectopy. These patients were excluded from HRV analysis and were analyzed in a separate group. For the remaining patients SDNN was calculated and the patients categorized into three groups based on the tertile cutpoints for SDNN (SDNN ⬍ 81ms, SDNN 81–113 ms, SDNN ⬎113 ms). The digital
283
Holter recording was performed at 7 ⫾ 10 months after study enrollment (within 3 months of study enrollment in 42% of patients). Mean duration of follow-up was 2.0 ⫾ 1.0 years. The clinical characteristics of the patients who did and did not have a digital Holter performed are compared in Table 1. Baseline characteristics were similar in the two groups, except that patients who had a Holter had better New York Heart Association functional class, were more likely to be treated with -adrenergic blockers, and less likely to be a racial minority. Patients who had a digital Holter had a lower mortality rate than patients who did not [26 (9%) vs 42 (23%), respectively; P ⬍ .05), with similar trends observed for cause-specific mortality (Table 1). The incidence of appropriate ICD shocks was similar in the two groups [22 (15%) vs 13 (16%); P ⫽ NS]. The predictive value of SDNN for all-cause mortality is illustrated in Figure 2. There were significant differences in the mortality rates at 3-year follow-up: SDNN ⬎113 ms: 0 (0%), SDNN 81–113 ms: 5 (7%), SDNN ⬍81 ms: 7 (10%), excluded patients: 11 (17%) (P ⫽ .03). No deaths occurred in the tertile with the most preserved HRV (SDNN ⬎113 ms), regardless of treatment randomization (ICD ⫹ standard medical therapy for heart failure or standard medical therapy for heart failure alone). Because the primary goal of this study was to identify low-risk patients who may not
Table 1 Baseline clinical characteristics of the study population Holter (n ⫽ 274) Age (years) Male gender Nonwhite race* Diabetes Left ventricular ejection fraction (%) New York Heart Association class* I II III Cardiovascular medication Amiodarone Angiotensin-converting enzyme inhibitor Angiotensin-II receptor blocker Beta-blocker* Deaths* Cause of death Sudden cardiac Cardiac, not sudden Noncardiac Unknown Appropriate implantable cardioverterdefibrillator shocks
*P ⬍.05.
59 199 56 64 21
⫾ 12 (73%) (22%) (23%) ⫾6
No Holter (n ⫽ 184) 58 127 62 41 22
⫾ 14 (69%) (36%) (22%) ⫾6
71 (26%) 149 (54%) 54 (20%)
28 (15%) 114 (62%) 41 (22%)
14 (5%) 235 (86%)
9 (5%) 157 (85%)
31 (11%)
20 (11%)
245 (89%) 26 (9%)
144 (78%) 42 (23%)
4 10 12 0 22
(2%) (4%) (4%) (0%) (15%)
13 11 14 4 13
(7%) (6%) (8%) (2%) (16%)
284
Figure 2 Risk stratification for total mortality, implantable cardioverter-defibrillator (ICD) and medical therapy arms combined. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
require an ICD, the two treatment groups were combined for subsequent analysis. Lower levels of SDNN were associated with increased mortality during follow-up. Patients who were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy had the worst prognosis. Similar results were obtained when the predictive value of SDNN was examined for cardiac mortality (Figure 3), the composite endpoint of sudden cardiac death ⫹ appropriate ICD shocks (Figure 4), and for appropriate ICD shocks
Figure 3 Risk stratification for cardiac mortality, implantable cardioverter-defibrillator (ICD) and medical therapy arms combined. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
Heart Rhythm, Vol 3, No 3, March 2006
Figure 4 Risk stratification for the combined incidence of sudden cardiac death (SCD) and implantable cardioverter-defibrillator (ICD shocks), ICD and medical therapy arms combined. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability (HRV) analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
(Figure 5). The categorical SDNN variable remained an independent predictor of mortality (P ⫽ .002) after adjustment for age, gender, LVEF, and QRS duration.
Discussion The major finding of this study is that patients having nonischemic dilated cardiomyopathy with preserved SDNN
Figure 5 Risk stratification for appropriate implantable cardioverter-defibrillator (ICD shocks), ICD arm only. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability (HRV) analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
Rashba et al
HRV Identifies Low-Risk Patients with Nonischemic Dilated Cardiomyopathy
have an excellent prognosis and may not require prophylactic ICD placement. Lower levels of SDNN were associated with progressively increased mortality risk. Importantly, patients who were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy were successfully risk stratified, as the presence of either of these findings was associated with the worst prognosis.
Previous studies Several observational studies have examined the prognostic value of HRV in subjects with nonischemic dilated cardiomyopathy. Fauchier et al8 reported that SDNN was an independent predictor of arrhythmic events in a cohort of 116 patients. Several other groups confirmed the prognostic value of SDNN in similar patient populations.7,13,14 In contrast, Hohnloser et al15 and Grimm et al16 reported that SDNN has no prognostic value in patients with nonischemic cardiomyopathy. The most likely reasons for these discrepancies are differences in the patient populations, such as inclusion of patients with ischemic and nonischemic disease and inclusion of secondary prevention patients who have already presented with sustained ventricular arrhythmias.7,13–15 The present study differs from these prior reports in several important respects. The DEFINITE trial was a multicenter, randomized controlled trial with 40 enrolling sites worldwide, whereas most previous studies were conducted at a single center. Because nonischemic dilated cardiomyopathy is a heterogeneous disorder, the inclusion of patients from many different centers enhances the generalizability of our results. Second, the mean LVEF was ⬃35% in most of the prior studies, whereas an inclusion criterion for DEFINITE was an ejection fraction ⱕ35%, and the mean ejection fraction was 21%. Indeed, only 25% of the patients in the Marburg cardiomyopathy study16 had clinical characteristics that approximated the DEFINITE entry criteria (LVEF ⬍30% and nonsustained ventricular tachycardia present). Because the performance of risk stratification techniques can be importantly influenced by the severity of left ventricular dysfunction,17 these differences in clinical characteristics likely impacted the study results. Finally, it is important to recognize that the design of the present study was different from that of previous studies of HRV in chronic congestive heart failure, which attempted to identify high-risk patients who would benefit from prophylactic ICD placement. The goals of risk stratification have changed in the era of positive primary prevention trials, as physicians and payers confront concerns about the cost-effectiveness of prophylactic ICD therapy in an expanding pool of patients. Accordingly, the present study was designed to identify low-risk patients who may not require prophylactic ICD placement. Our results indicate that patients with nonischemic dilated cardiomyopathy who have preserved HRV have an excellent prognosis. Importantly, this low-risk subgroup composed 25% of the present study population, which could translate into a substantial reduction in health
285
care costs if prophylactic ICD placement could be safely deferred. This strategy also would avoid the unnecessary exposure of low-risk patients to the potential complications of ICD placement, such as the surgical risks of implantation and subsequent device replacement procedures, device infection, inappropriate shocks, and malfunction of the leads or pulse generator. The utility of risk stratification is potentially limited by the exclusion of patients who are not eligible for testing. In the present study, 23% of patients were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy. However, the event rate of the excluded patients was comparable with that of patients with severely impaired HRV. Our findings are consistent with previous reports that correlated the presence of atrial fibrillation and ventricular ectopy with an adverse prognosis in subjects with congestive heart failure.10,18,19 These data support the concept that the presence of atrial fibrillation or frequent ventricular ectopy in subjects with nonischemic dilated cardiomyopathy is sufficient evidence of an adverse prognosis that further risk stratification is not useful or necessary.
Study limitations The major limitation of the present study is that patients who had a digital Holter had a lower mortality rate than patients who did not. The most likely explanation for this finding is that Holter recordings were not obtained at enrollment in all patients, which resulted in missed events. It is possible that this inadvertent selection bias influenced the study results because it is easier to identify very low-risk patients within a cohort with a lower event rate. However, it is reassuring that the incidence of appropriate ICD therapy was similar among patients who did and did not have a digital Holter. The study results were consistent for the mortality and appropriate ICD therapy endpoints, suggesting that our findings are applicable to all patients who meet DEFINITE entry criteria. Finally, it is possible that the substrate for ventricular arrhythmias may change over time as a result of progressive remodeling, alteration in the neurohormonal milieu, or other factors. Additional studies are required to determine if there is a “warranty period” for a low-risk HRV measurement that would mandate periodic reassessment of arrhythmia risk if ICD therapy was initially withheld.
Clinical implications Our results suggest that prophylactic ICD therapy could be safely withheld from nonischemic dilated cardiomyopathy patients who have preserved HRV. However, the decision to withhold ICD therapy from patients who meet an approved indication has substantial ethical and medicolegal implications, and our data are subject to the study limitations cited. For these reasons, additional prospective studies are required to confirm our results in a large patient cohort with
286 implanted ICDs before implementing these findings into clinical practice.
Heart Rhythm, Vol 3, No 3, March 2006
9.
10.
Acknowledgment We thank Mario Dimitrov for expert technical assistance with editing the Holter data.
11.
12.
References 13. 1. Zheng ZJ, Croft JB, Giles WH, Mensah GA. Sudden cardiac death in the United States, 1989 –1998. Circulation 2001;104:2158 –2163. 2. Huikuri H, Castellanos A, Myerburg RJ. Sudden death due to cardiac arrhythmias. N Engl J Med 2001;345:1473–1482. 3. Kadish A, Dyer A, Daubert JP, Quigg R, Estes NAM, Anderson KP, Calkins H, Hoch D, Goldberger D, Shalaby A, Sanders WE, Schaechter A, and Levine JH, for the Defibrillators in Non-ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) Investigators. Prophylactic defibrillator implantation in patients with nonischemic dilated cardiomyopathy. N Engl J Med 2004;350:2151–2158. 4. Moss AJ, Zareba WJ, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert JP, Higgins SL, Brown MW, Andrews ML. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877– 883. 5. Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R, Domanski M, Troutman C, Anderson J, Johnson G, McNulty SE, Clapp-Channing N, Davidson-Ray LD, Fraulo ES, Fishbein DP, Luceri RM, Ip JH, for the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) Investigators. Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med 2005; 352:225–237. 6. Bigger JT. Expanding indications for implantable cardiac defibrillators. N Engl J Med 2002;346:931–932. 7. Nolan J, Batin PD, Andrews R, Lindsay SJ, Brooksby P, Mullen M, Baig W, Flapan AD, Cowley A, Prescott RJ, Neilson JM, Fox KA. Prospective study of heart rate variability and mortality in chronic heart failure: results of the United Kingdom heart failure evaluation and assessment of risk trial (UK-Heart). Circulation 1998;98:1510 – 1516. 8. Fauchier L, Babuty D, Cosnay P, Fauchier JP. Prognostic value of heart rate variability for sudden death and major arrhythmic events in
14.
15.
16.
17.
18.
19.
patients with idiopathic dilated cardiomyopathy. J Am Coll Cardiol 1999;33:1203–1207. Kadish A, Quigg R, Schaechter A, Anderson KP, Estes NAM, Levine J. Defibrillators in nonischemic cardiomyopathy treatment evaluation. Pacing Clin Electrophysiol 2000;23:338 –343. Doval HC, Nul DR, Grancelli HO, Varini SD, Soifer S, Corrado G, Dubner S, Scapin O, Perrone SV. Nonsustained ventricular tachycardia in severe heart failure: independent marker of increased mortality due to sudden death. Circulation 1996;94:3198 –3203. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiologic interpretation, and clinical use. Circulation 1996;93:1043–1065. Epstein AE, Carlson MD, Fogoros RN, Higgins SL, Venditti FJ Jr. Classification of death in antiarrhythmia trials. J Am Coll Cardiol 1996;27:433– 442. Guzzetti S, La Rovere MT, Pinna GD, Maestri R, Borroni E, Porta A, Mortara A, Malliani A. Different spectral components of 24 h heart rate variability are related to different modes of death in chronic heart failure. Eur Heart J 2005;26:357–362. Boveda S, Galinier M, Pathak A, Fourcade J, Dongay B, Benchendikh D, Massabuau P, Fauvel JM, Senard JM, Bounhoure JP. Prognostic value of heart rate variability in time domain analysis in congestive heart failure. J Interv Card Electrophysiol 2001;5:181–187. Hohnloser SH, Klingenheben T, Bloomfield D, Dabbous O, Cohen RJ. Usefulness of microvolt T-wave alternans for prediction of ventricular tachyarrhythmic events in patients with dilated cardiomyopathy: results from a prospective observational study. J Am Coll Cardiol 2003; 41:2220 –2224. Grimm W, Christ M, Bach J, Muller HH, Maisch B. Noninvasive arrhythmia risk stratification in idiopathic dilated cardiomyopathy. Results of the Marburg Cardiomyopathy Study. Circulation 2003;108: 2883–2891. Rashba EJ, Osman AF, Macmurdy K, Kirk MM, Sarang SE, Peters RW, Shorofsky SR, Gold MR. Enhanced detection of arrhythmia vulnerability using T wave alternans, left ventricular ejection fraction, and programmed ventricular stimulation: a prospective study in subjects with chronic ischemic heart disease. J Cardiovasc Electrophysiol 2004;15:170 –176. Dries DL, Exner DV, Gersh BJ, Domanski MJ, Waclawiw MA, Stevenson LW. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998;32:695–703. Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991;84:40 – 48.
This research was supported by a K23 grant (HL-067198) from the National Heart, Lung, and Blood Institute to Dr. Rashba; and by St. Jude Medical, CRMD. Address reprint requests and correspondence: Dr. Eric J. Rashba, Division of Cardiology, University of Maryland Medical Center, 22 South Greene Street, Room N3W77, Baltimore, Maryland 21201. E-mail address: [email protected]. (Received October 14, 2005; accepted November 30, 2005.)
1547-5271/$ -see front matter © 2006 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2005.11.028
282
Introduction Sudden cardiac death remains common in the United States, with an estimated 400,000 cases reported annually.1,2 Most patients who suffer an out-of-hospital cardiac arrest do not survive the initial event.1,2 Consequently, there has been great interest in identifying patients who are at risk for sudden cardiac death so that prophylactic implantable cardioverter-defibrillator (ICD) therapy can be instituted. The landmark DEFINITE (Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation),3 MADIT-2 (Multicenter Automatic Defibrillator Implantation Trial 2),4 and SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial)5 trials conclusively demonstrated that prophylactic ICD placement reduces mortality among subjects with ischemic and nonischemic cardiomyopathy. The scientific conclusions of these trials have not been seriously disputed, but there has been a vigorous debate regarding the cost implications of implementing their findings into clinical practice, as well as the potential unnecessary exposure of low-risk patients to the complications of ICD placement.6 As a result, the focus of risk stratification efforts has shifted to identifying low-risk patients who may not require prophylactic ICD placement. Heart rate variability (HRV) is a potent predictor of survival in subjects with congestive heart failure and systolic dysfunction.7,8 We hypothesized that nonischemic dilated cardiomyopathy patients with preserved HRV would have an excellent prognosis and designed a prospective study to test this hypothesis using the DEFINITE patient cohort.
Heart Rhythm, Vol 3, No 3, March 2006 sin-II receptor blockers. In addition, beta-blocker therapy was required unless patients were unable to tolerate it. The use of antiarrhythmic drugs such as amiodarone was discouraged. However, a small number of patients were treated with amiodarone to control symptomatic atrial arrhythmias. The DEFINITE trial commenced enrollment in July 1998. Funding for the digital Holter equipment was obtained after the trial was initiated, and the first recording was obtained in August 1999. The Holter recordings were obtained at baseline whenever possible or at the next scheduled follow-up visit for patients who were already enrolled in the study. The DEFINITE trial was approved by the Institutional Review Boards at participating centers, and written informed consent was obtained from all patients.
HRV analysis Burdick model 6632 digital Holter recorders were used for the study. The data were stored on 350-MB PC cards, which were shipped to the Holter Core Lab at the University of Maryland Medical Center for downloading and analysis. Each recording was manually edited to exclude segments with excessive noise and to correctly identify each beat. All Holter recordings were overread by the principal investigator of the study (E.J.R.). The recording was excluded from HRV analysis if atrial fibrillation or frequent ventricular ectopic beats (⬎25% of total beats) was present or if the recording lasted ⬍18 hours.11 The standard deviation of normal R-R intervals (SDNN) was prespecified as the primary HRV variable.11
Randomization and follow-up
Methods Study design The study design and primary results of the DEFINITE trial have been reported previously.9 In brief, DEFINITE was a prospective, randomized comparison of standard oral medical therapy for heart failure vs standard oral medical therapy plus an ICD. Inclusion criteria were left ventricular ejection fraction (LVEF) ⬍36%, presence of ambient arrhythmias,10 history of symptomatic heart failure, and presence of nonischemic dilated cardiomyopathy. Ambient arrhythmias were defined by an episode of nonsustained ventricular tachycardia on Holter or telemetry monitoring (3–15 beats at a rate ⬎120 bpm) or an average of at least 10 premature ventricular complexes per hour on 24-hour Holter monitoring. The absence of clinically significant coronary artery disease as the cause of the cardiomyopathy was confirmed by coronary angiography or by a negative stress imaging study. Detailed inclusion and exclusion criteria were previously reported.9 All patients received angiotensin-converting enzyme (ACE) inhibitors unless they were contraindicated. Patients who were unable to tolerate ACE inhibitors received hydralazine, nitrates, or angioten-
Patients were randomly assigned to receive either standard oral medical therapy for heart failure or standard oral medical therapy plus an ICD. Patients who were randomized to the ICD group received a single-chamber device that was programmed to detect ventricular fibrillation at a rate of 180 bpm. All patients were evaluated at 3-month intervals. The cause of death of patients who died was determined by an events committee whose members were unaware of patients’ treatment assignments. The cause of death was determined according to the criteria of Epstein et al.12 ICD shocks were classified as appropriate or inappropriate by a committee of participating investigators.
Statistical analysis The baseline characteristics of participants in the Holter substudy and nonparticipants were compared using unpaired t-tests for continuous variables and chi-square tests for categorical variables. The prespecified analytic plan was developed before acquisition of the data. The SDNN data were categorized into tertiles. Patients who had a Holter but were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy (⬎25% of beats) were
Rashba et al
HRV Identifies Low-Risk Patients with Nonischemic Dilated Cardiomyopathy
Figure 1 Flow of participants in the study. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
analyzed in a fourth group. The primary endpoint was total mortality. Secondary endpoints were cardiac mortality, appropriate ICD shocks, and the combined incidence of appropriate ICD shocks and sudden cardiac death. The stored electrograms for all ICD shocks were reviewed by a committee of investigators to determine if the shock was appropriate. The DEFINITE trial design specified that follow-up would continue until the 68th death occurred (May 25, 2003). Data on patients who received a heart transplant were censored at the time of transplantation. The duration of follow-up was computed from the date of the Holter to death for patients who died and to the date of the 68th death for patients who did not die. The same approach was used for analyses of secondary endpoints. Kaplan-Meier survival curves were used to estimate the cumulative percentage of patients surviving free from endpoint events over time. The incidence of endpoint events was calculated at 3-year follow-up. Comparisons between the survival curves were made using the log rank statistic using all available followup. Cox proportional hazards analysis was conducted to determine if the predictive value of the categorical SDNN variable was independent of age, gender, LVEF, and QRS duration. All reported P values are two-tailed.
Results A description of the patient flow for the study is shown in Figure 1. Of the 458 patients who participated in the DEFINITE trial, 303 (66%) had a digital Holter monitor performed. Thirty-one of the 40 DEFINITE enrolling centers contributed at least one patient to the study. Twenty-nine patients were excluded because fewer than 18 hours of analyzable data were present on the Holter recording, yielding 274 patients for further analysis. Of the 274 patients with a complete digital Holter recording, 16% had atrial fibrillation and 7% had frequent ventricular ectopy. These patients were excluded from HRV analysis and were analyzed in a separate group. For the remaining patients SDNN was calculated and the patients categorized into three groups based on the tertile cutpoints for SDNN (SDNN ⬍ 81ms, SDNN 81–113 ms, SDNN ⬎113 ms). The digital
283
Holter recording was performed at 7 ⫾ 10 months after study enrollment (within 3 months of study enrollment in 42% of patients). Mean duration of follow-up was 2.0 ⫾ 1.0 years. The clinical characteristics of the patients who did and did not have a digital Holter performed are compared in Table 1. Baseline characteristics were similar in the two groups, except that patients who had a Holter had better New York Heart Association functional class, were more likely to be treated with -adrenergic blockers, and less likely to be a racial minority. Patients who had a digital Holter had a lower mortality rate than patients who did not [26 (9%) vs 42 (23%), respectively; P ⬍ .05), with similar trends observed for cause-specific mortality (Table 1). The incidence of appropriate ICD shocks was similar in the two groups [22 (15%) vs 13 (16%); P ⫽ NS]. The predictive value of SDNN for all-cause mortality is illustrated in Figure 2. There were significant differences in the mortality rates at 3-year follow-up: SDNN ⬎113 ms: 0 (0%), SDNN 81–113 ms: 5 (7%), SDNN ⬍81 ms: 7 (10%), excluded patients: 11 (17%) (P ⫽ .03). No deaths occurred in the tertile with the most preserved HRV (SDNN ⬎113 ms), regardless of treatment randomization (ICD ⫹ standard medical therapy for heart failure or standard medical therapy for heart failure alone). Because the primary goal of this study was to identify low-risk patients who may not
Table 1 Baseline clinical characteristics of the study population Holter (n ⫽ 274) Age (years) Male gender Nonwhite race* Diabetes Left ventricular ejection fraction (%) New York Heart Association class* I II III Cardiovascular medication Amiodarone Angiotensin-converting enzyme inhibitor Angiotensin-II receptor blocker Beta-blocker* Deaths* Cause of death Sudden cardiac Cardiac, not sudden Noncardiac Unknown Appropriate implantable cardioverterdefibrillator shocks
*P ⬍.05.
59 199 56 64 21
⫾ 12 (73%) (22%) (23%) ⫾6
No Holter (n ⫽ 184) 58 127 62 41 22
⫾ 14 (69%) (36%) (22%) ⫾6
71 (26%) 149 (54%) 54 (20%)
28 (15%) 114 (62%) 41 (22%)
14 (5%) 235 (86%)
9 (5%) 157 (85%)
31 (11%)
20 (11%)
245 (89%) 26 (9%)
144 (78%) 42 (23%)
4 10 12 0 22
(2%) (4%) (4%) (0%) (15%)
13 11 14 4 13
(7%) (6%) (8%) (2%) (16%)
284
Figure 2 Risk stratification for total mortality, implantable cardioverter-defibrillator (ICD) and medical therapy arms combined. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
require an ICD, the two treatment groups were combined for subsequent analysis. Lower levels of SDNN were associated with increased mortality during follow-up. Patients who were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy had the worst prognosis. Similar results were obtained when the predictive value of SDNN was examined for cardiac mortality (Figure 3), the composite endpoint of sudden cardiac death ⫹ appropriate ICD shocks (Figure 4), and for appropriate ICD shocks
Figure 3 Risk stratification for cardiac mortality, implantable cardioverter-defibrillator (ICD) and medical therapy arms combined. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
Heart Rhythm, Vol 3, No 3, March 2006
Figure 4 Risk stratification for the combined incidence of sudden cardiac death (SCD) and implantable cardioverter-defibrillator (ICD shocks), ICD and medical therapy arms combined. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability (HRV) analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
(Figure 5). The categorical SDNN variable remained an independent predictor of mortality (P ⫽ .002) after adjustment for age, gender, LVEF, and QRS duration.
Discussion The major finding of this study is that patients having nonischemic dilated cardiomyopathy with preserved SDNN
Figure 5 Risk stratification for appropriate implantable cardioverter-defibrillator (ICD shocks), ICD arm only. SDNN is categorized in tertiles. Patients who were excluded from heart rate variability (HRV) analysis because of atrial fibrillation (AF) or frequent ventricular ectopy are analyzed in a separate group. Kaplan-Meier curves of event-free survival are shown, truncated at 3 years. PVCs ⫽ frequent premature ventricular contractions (⬎25% of beats); SDNN ⫽ standard deviation of normal R-R intervals.
Rashba et al
HRV Identifies Low-Risk Patients with Nonischemic Dilated Cardiomyopathy
have an excellent prognosis and may not require prophylactic ICD placement. Lower levels of SDNN were associated with progressively increased mortality risk. Importantly, patients who were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy were successfully risk stratified, as the presence of either of these findings was associated with the worst prognosis.
Previous studies Several observational studies have examined the prognostic value of HRV in subjects with nonischemic dilated cardiomyopathy. Fauchier et al8 reported that SDNN was an independent predictor of arrhythmic events in a cohort of 116 patients. Several other groups confirmed the prognostic value of SDNN in similar patient populations.7,13,14 In contrast, Hohnloser et al15 and Grimm et al16 reported that SDNN has no prognostic value in patients with nonischemic cardiomyopathy. The most likely reasons for these discrepancies are differences in the patient populations, such as inclusion of patients with ischemic and nonischemic disease and inclusion of secondary prevention patients who have already presented with sustained ventricular arrhythmias.7,13–15 The present study differs from these prior reports in several important respects. The DEFINITE trial was a multicenter, randomized controlled trial with 40 enrolling sites worldwide, whereas most previous studies were conducted at a single center. Because nonischemic dilated cardiomyopathy is a heterogeneous disorder, the inclusion of patients from many different centers enhances the generalizability of our results. Second, the mean LVEF was ⬃35% in most of the prior studies, whereas an inclusion criterion for DEFINITE was an ejection fraction ⱕ35%, and the mean ejection fraction was 21%. Indeed, only 25% of the patients in the Marburg cardiomyopathy study16 had clinical characteristics that approximated the DEFINITE entry criteria (LVEF ⬍30% and nonsustained ventricular tachycardia present). Because the performance of risk stratification techniques can be importantly influenced by the severity of left ventricular dysfunction,17 these differences in clinical characteristics likely impacted the study results. Finally, it is important to recognize that the design of the present study was different from that of previous studies of HRV in chronic congestive heart failure, which attempted to identify high-risk patients who would benefit from prophylactic ICD placement. The goals of risk stratification have changed in the era of positive primary prevention trials, as physicians and payers confront concerns about the cost-effectiveness of prophylactic ICD therapy in an expanding pool of patients. Accordingly, the present study was designed to identify low-risk patients who may not require prophylactic ICD placement. Our results indicate that patients with nonischemic dilated cardiomyopathy who have preserved HRV have an excellent prognosis. Importantly, this low-risk subgroup composed 25% of the present study population, which could translate into a substantial reduction in health
285
care costs if prophylactic ICD placement could be safely deferred. This strategy also would avoid the unnecessary exposure of low-risk patients to the potential complications of ICD placement, such as the surgical risks of implantation and subsequent device replacement procedures, device infection, inappropriate shocks, and malfunction of the leads or pulse generator. The utility of risk stratification is potentially limited by the exclusion of patients who are not eligible for testing. In the present study, 23% of patients were excluded from HRV analysis because of atrial fibrillation or frequent ventricular ectopy. However, the event rate of the excluded patients was comparable with that of patients with severely impaired HRV. Our findings are consistent with previous reports that correlated the presence of atrial fibrillation and ventricular ectopy with an adverse prognosis in subjects with congestive heart failure.10,18,19 These data support the concept that the presence of atrial fibrillation or frequent ventricular ectopy in subjects with nonischemic dilated cardiomyopathy is sufficient evidence of an adverse prognosis that further risk stratification is not useful or necessary.
Study limitations The major limitation of the present study is that patients who had a digital Holter had a lower mortality rate than patients who did not. The most likely explanation for this finding is that Holter recordings were not obtained at enrollment in all patients, which resulted in missed events. It is possible that this inadvertent selection bias influenced the study results because it is easier to identify very low-risk patients within a cohort with a lower event rate. However, it is reassuring that the incidence of appropriate ICD therapy was similar among patients who did and did not have a digital Holter. The study results were consistent for the mortality and appropriate ICD therapy endpoints, suggesting that our findings are applicable to all patients who meet DEFINITE entry criteria. Finally, it is possible that the substrate for ventricular arrhythmias may change over time as a result of progressive remodeling, alteration in the neurohormonal milieu, or other factors. Additional studies are required to determine if there is a “warranty period” for a low-risk HRV measurement that would mandate periodic reassessment of arrhythmia risk if ICD therapy was initially withheld.
Clinical implications Our results suggest that prophylactic ICD therapy could be safely withheld from nonischemic dilated cardiomyopathy patients who have preserved HRV. However, the decision to withhold ICD therapy from patients who meet an approved indication has substantial ethical and medicolegal implications, and our data are subject to the study limitations cited. For these reasons, additional prospective studies are required to confirm our results in a large patient cohort with
286 implanted ICDs before implementing these findings into clinical practice.
Heart Rhythm, Vol 3, No 3, March 2006
9.
10.
Acknowledgment We thank Mario Dimitrov for expert technical assistance with editing the Holter data.
11.
12.
References 13. 1. Zheng ZJ, Croft JB, Giles WH, Mensah GA. Sudden cardiac death in the United States, 1989 –1998. Circulation 2001;104:2158 –2163. 2. Huikuri H, Castellanos A, Myerburg RJ. Sudden death due to cardiac arrhythmias. N Engl J Med 2001;345:1473–1482. 3. Kadish A, Dyer A, Daubert JP, Quigg R, Estes NAM, Anderson KP, Calkins H, Hoch D, Goldberger D, Shalaby A, Sanders WE, Schaechter A, and Levine JH, for the Defibrillators in Non-ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) Investigators. Prophylactic defibrillator implantation in patients with nonischemic dilated cardiomyopathy. N Engl J Med 2004;350:2151–2158. 4. Moss AJ, Zareba WJ, Hall WJ, Klein H, Wilber DJ, Cannom DS, Daubert JP, Higgins SL, Brown MW, Andrews ML. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med 2002;346:877– 883. 5. Bardy GH, Lee KL, Mark DB, Poole JE, Packer DL, Boineau R, Domanski M, Troutman C, Anderson J, Johnson G, McNulty SE, Clapp-Channing N, Davidson-Ray LD, Fraulo ES, Fishbein DP, Luceri RM, Ip JH, for the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) Investigators. Amiodarone or an implantable cardioverter-defibrillator for congestive heart failure. N Engl J Med 2005; 352:225–237. 6. Bigger JT. Expanding indications for implantable cardiac defibrillators. N Engl J Med 2002;346:931–932. 7. Nolan J, Batin PD, Andrews R, Lindsay SJ, Brooksby P, Mullen M, Baig W, Flapan AD, Cowley A, Prescott RJ, Neilson JM, Fox KA. Prospective study of heart rate variability and mortality in chronic heart failure: results of the United Kingdom heart failure evaluation and assessment of risk trial (UK-Heart). Circulation 1998;98:1510 – 1516. 8. Fauchier L, Babuty D, Cosnay P, Fauchier JP. Prognostic value of heart rate variability for sudden death and major arrhythmic events in
14.
15.
16.
17.
18.
19.
patients with idiopathic dilated cardiomyopathy. J Am Coll Cardiol 1999;33:1203–1207. Kadish A, Quigg R, Schaechter A, Anderson KP, Estes NAM, Levine J. Defibrillators in nonischemic cardiomyopathy treatment evaluation. Pacing Clin Electrophysiol 2000;23:338 –343. Doval HC, Nul DR, Grancelli HO, Varini SD, Soifer S, Corrado G, Dubner S, Scapin O, Perrone SV. Nonsustained ventricular tachycardia in severe heart failure: independent marker of increased mortality due to sudden death. Circulation 1996;94:3198 –3203. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiologic interpretation, and clinical use. Circulation 1996;93:1043–1065. Epstein AE, Carlson MD, Fogoros RN, Higgins SL, Venditti FJ Jr. Classification of death in antiarrhythmia trials. J Am Coll Cardiol 1996;27:433– 442. Guzzetti S, La Rovere MT, Pinna GD, Maestri R, Borroni E, Porta A, Mortara A, Malliani A. Different spectral components of 24 h heart rate variability are related to different modes of death in chronic heart failure. Eur Heart J 2005;26:357–362. Boveda S, Galinier M, Pathak A, Fourcade J, Dongay B, Benchendikh D, Massabuau P, Fauvel JM, Senard JM, Bounhoure JP. Prognostic value of heart rate variability in time domain analysis in congestive heart failure. J Interv Card Electrophysiol 2001;5:181–187. Hohnloser SH, Klingenheben T, Bloomfield D, Dabbous O, Cohen RJ. Usefulness of microvolt T-wave alternans for prediction of ventricular tachyarrhythmic events in patients with dilated cardiomyopathy: results from a prospective observational study. J Am Coll Cardiol 2003; 41:2220 –2224. Grimm W, Christ M, Bach J, Muller HH, Maisch B. Noninvasive arrhythmia risk stratification in idiopathic dilated cardiomyopathy. Results of the Marburg Cardiomyopathy Study. Circulation 2003;108: 2883–2891. Rashba EJ, Osman AF, Macmurdy K, Kirk MM, Sarang SE, Peters RW, Shorofsky SR, Gold MR. Enhanced detection of arrhythmia vulnerability using T wave alternans, left ventricular ejection fraction, and programmed ventricular stimulation: a prospective study in subjects with chronic ischemic heart disease. J Cardiovasc Electrophysiol 2004;15:170 –176. Dries DL, Exner DV, Gersh BJ, Domanski MJ, Waclawiw MA, Stevenson LW. Atrial fibrillation is associated with an increased risk for mortality and heart failure progression in patients with asymptomatic and symptomatic left ventricular systolic dysfunction: a retrospective analysis of the SOLVD trials. Studies of Left Ventricular Dysfunction. J Am Coll Cardiol 1998;32:695–703. Middlekauff HR, Stevenson WG, Stevenson LW. Prognostic significance of atrial fibrillation in advanced heart failure. A study of 390 patients. Circulation 1991;84:40 – 48.