Despite advances in cardiopulmonary resuscitation education and the use of automatic external defibrillators, the survival rate from cardiac arrest in the United States remains low (1% to 20%). Thus, the possibility of selecting high-risk patients for prophylactic implantable cardioverter defibrillator (ICD) implantation is attractive. In the past decade, several studies have suggested that subgroups of patients with ischemic cardiomyopathy may benefit from ICD implantation. The first Multicenter Automatic Defibrillator Implant Trial (MADIT)^[1] randomized patients with coronary artery disease (CAD) and inducible ventricular tachycardia (VT) at electrophysiologic testing to medical therapy or ICD implantation. The results from MADIT suggested that patients who underwent ICD implantation had a better prognosis. Subsequent studies, including the Multicenter Unsustained Tachycardia Trial (MUSTT)^[2] and MADIT II,^[3] confirmed this observation.

The MADIT II study suggested that electrophysiologic testing was not an essential component of risk stratification for patients with ischemic cardiomyopathy and an ejection fraction (EF) ≤ 30%. Thus, prophylactic ICD implantation has become the therapy of choice in patients with CAD and left ventricular (LV) dysfunction who do not have contraindications to ICD implantation. The recent Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) also helped confirm these findings. SCD-HeFT enrolled patients with congestive heart failure (CHF) and LVEF < 36% regardless of etiology. In patients with CAD, there was a 21% reduction in total mortality with the use of ICD plus best medical therapy compared with best medical therapy only. In addition, SCD-HeFT verified that amiodarone was not effective for the prevention of sudden cardiac death (SCD) in ischemic patients.

Although observational studies suggested that some patients with nonischemic dilated cardiomyopathy may be at high risk for SCD, until the last several months, prospective randomized studies regarding prophylactic ICDs and nonischemic cardiomyopathy have not been available. Initial results from the Defibrillators in Non-Ischemic Cardiomyopathy Treatment Evaluation study (DEFINITE) were presented at the American Heart Association meeting in November 2003, and initial results from the SCD-HeFT trial were presented at the American College of Cardiology meeting in 2004. Although there were some important differences between these 2 studies in terms of the results and design, overall, the results of both suggest that prophylactic ICD implantation in patients with nonischemic cardiomyopathy is appropriate and lifesaving, just as it is in patients with coronary disease. Additional results of the DEFINITE and the SCD-HeFT trials were presented at the recent Heart Rhythm Society meeting in San Francisco. Results from the DEFINITE study were also published in The New England Journal of Medicine during the week of the presentation.^[4]

DEFINITE: Trial Results

The DEFINITE study randomized 458 subjects with nonischemic dilated cardiomyopathy and LVEF < 36% to receive best medical therapy or to receive best medical therapy plus an ICD. Although all patients had to have experienced heart failure at some point, they could have been in New York Heart Association (NYHA) Class I on medical therapy at the time of randomization. The median age of patients was 58 years, 71% of the study subjects were male, two thirds of the subjects were white, and approximately 25% had a history of diabetes and/or atrial fibrillation. All of these variables were similarly distributed between the 2 groups. Patients randomized to standard therapy had a longer time from diagnosis of cardiomyopathy to randomization in the trial (3.27 vs 2.39 years) but adjusting for this difference did not alter any of the subsequent results. At the time of randomization, 22% of patients were in class I, 57% in class II, and 21% in class III; the average QRS duration was 115 ms. Ninety percent of the randomized subjects had nonsustained VT, and the rest had frequent premature ventricular contractions. The mean LVEF was 21% and the mean walk distance was 319 meters. More than 85% of patients were treated with beta-blockers and angiotensin-converting enzyme (ACE) inhibitors during the course of the trial. The majority of patients who did not receive ACE inhibitors were treated with angiotensin receptor blocking drugs.

The primary endpoint of the DEFINITE trial was total mortality. After 2 years, 14.1% of the patients in the standard therapy group and 7.9% of the patients in the ICD group died (6.2% absolute mortality reduction). There was a 35% relative mortality reduction with the ICD, but the P value only reached .08 and thus the results for the primary endpoint were not significant. Arrhythmic mortality was significantly reduced by the ICD (hazard ratio 0.2; P = .006). When cause-specific mortality was analyzed, only 35% of the patients in the standard therapy group experienced sudden death in contrast to the 50% that had been predicted. This difference likely accounted for the failure of the primary endpoint to reach statistical significance. Although statistical significance in the primary endpoint was not reached, the results of the study did suggest that the ICD may have utility in selected patients with nonischemic dilated cardiomyopathy. On subgroup analysis, the only major subgroup that was different from the others was patients with NYHA Class III heart failure who appeared to have a greater ICD benefit. QRS duration did not have a major effect on outcome, and there was only a very small trend for patients with a longer QRS to have a greater benefit from the ICD, a trend that could have been due to chance since the confidence intervals showed wide overlap.

SCD-HeFT: Trial Results

The SCD-HeFT trial randomized patients with both coronary disease and nonischemic cardiomyopathy into 3 treatment groups: (1) placebo, (2) amiodarone in a weight-adjusted dose, (3) an ICD. To be included in SCD-HeFT, patients had to have class II or III CHF despite treatment with ACE inhibitors and beta-blockers (approximately 75% to 80% of patients in SCD-HeFT received ACE inhibitors or beta-blockers). The primary endpoint for SCD-HeFT was total mortality. There was a 23% reduction in total mortality in the ICD group compared with placebo. Of the 1676 patients who were randomized to either ICD or placebo, 792 had nonischemic cardiomyopathy. Thus, patients with presumed nonischemic cardiomyopathy were over-represented in the SCD-HeFT trial compared with the general population. Whether this difference was due to referral bias or due to an incomplete characterization of the etiology of underlying structural heart disease is not entirely clear. Both patients with ischemic and those with nonischemic cardiomyopathy appeared to benefit from the ICD. There was a 21% reduction in mortality in patients with ischemic disease and a 27% reduction in mortality in patients with nonischemic dilated cardiomyopathy.

Surprisingly, patients with NYHA Class II CHF appeared to have a greater benefit from the ICD than those with class III. These results are not similar to the observations in the DEFINITE trial and are also not similar to several other prior studies on ICD therapy. There is a clear hypothesis that could explain these results, which is that patients with class III CHF had progressive pump failure as their main mechanism of death and thus did not benefit from ICD implantation. However, this observation has not been made in other clinical trials.

The initial SCD-HeFT results did not break down the class II and class III results on the basis of etiology and underlying structural heart disease, and thus it is impossible to be certain that these results held true for both patients with ischemic and nonischemic cardiomyopathy, although it has been suggested that this is the case.

QRS Subanalysis

At the Heart Rhythm Society meeting, the SCD-HeFT investigators presented follow-up data on an important subgroup analysis, QRS duration. This analysis is important because, although the MADIT II study in patients with CAD was not prospectively designed to stratify patients on the basis of QRS duration, there was a trend in the MADIT data for patients with a longer QRS duration to have a greater degree of benefit from the ICD. Thus, the US Centers for Medicare and Medicaid Services (CMS), the organization that evaluates reimbursement for medical devices, approved the ICD for patients with CAD, but only for those patients with prolonged QRS duration and not those with narrow QRS duration.

To address this issue the SCD-HeFT investigators performed what was essentially a sensitivity analysis, in which they examined the effect of QRS duration on outcome and ICD benefit in the SCD-HeFT trial. They found that the relationship between QRS duration and ICD benefit altered depending on how they classified their data. Changing patients with a QRS duration of 120 ms from the narrow QRS group to the wide QRS group completely altered the statistical analysis from one where QRS duration was suggested to be a predictor of outcome, to one where there was actually an opposite effect, in that patients with a long QRS duration had less of a benefit from the ICD. This analysis has suggested that post hoc examination of the effects of QRS duration on outcome was fraught with danger and that since the study was not specifically designed to examine the effects of QRS duration on outcome, subgroup analysis and posthoc analysis were not appropriate. The investigators concluded that the ICD should be approved for patients with nonischemic and ischemic dilated cardiomyopathy who do not have CHF, regardless of QRS duration.

DEFINITE vs SCD-HeFT

The results of SCD-HeFT are generally similar to those seen in other studies. They confirm a significant reduction in mortality among patients receiving an ICD who have CAD. They also confirm the strong trend (that did not quite reach significance) that was seen in the DEFINITE trial in patients with nonischemic cardiomyopathy. There are some important differences, however. The mortality benefits seen with the ICD in the SCD-HeFT trial were not as significant as those seen in prior studies of patients with coronary disease. The absolute mortality benefit and the relative benefit in SCD-HeFT were inferior to those seen in MADIT II. By contrast, the relative mortality benefit in SCD-HeFT was very similar to that seen in DEFINITE (27% vs 35%), but with an important difference. In the DEFINITE trial, there was a trend toward the benefit being greater in patients with class III CHF, whereas in SCD-HeFT, patients with class II CHF did significantly better with the ICD than those in class III.

The reason for this difference is not immediately obvious from the preliminary presentations of data from the SCD-HeFT trial, and a careful analysis of the factors that account for this difference must await publication of the full data set.

What About Cost and Risk Stratification?

What about cost? Although the MADIT II trial strongly suggested that patients with coronary disease and LV dysfunction should be treated with ICD therapy, there was some reluctance to proceed with blanket reimbursement if it confined reimbursement to patients with a wide QRS complex. ICDs currently cost approximately $20,000 per patient, and if ICD therapy were widely applied to the target population it would result in a significant increase in healthcare costs for patients with heart disease. Cost-effective analyses for both the SCD-HeFT trial and the DEFINITE trial are still under way. A decision on coverage by CMS is also pending. It is hoped that these additional pieces to the puzzle will be available by the fall of this year to allow clinicians to make a well-informed, as well as reimbursable, decision on the cost-effectiveness of ICD therapy in these various groups.

What about risk stratification? In contrast to prior trials, SCD-HeFT used only EF in CHF as risk-stratifying markers for death and therapy. With these markers, the total mortality at 5 years in SCD-HeFT was relatively modest (25%) for a group of patients with an EF of 25%. The ICD reduced mortality by only 1% to 2% per year, a smaller benefit than was seen in the DEFINITE trial. Thus, techniques for risk stratification could have potential utility in selecting high-risk patients for ICD implantation. A large number of noninvasive techniques are currently available. However, prospective studies have not yet been completed that demonstrate the rationale for treating patients on the basis of a single average electrocardiogram, heart rate variability, T-wave alternans, QT dispersion, or other factors. In addition, on the basis of the data from the MADIT II trial, the role of electrophysiologic testing to identify high-risk patients with coronary disease still remains unclear. At the present time, prospectively validated risk-stratification techniques other than EF and the presence of heart failure are not available. They likely should be used only as an adjunct for decision making in cases where physicians are "on the fence" about what needs to be done.

Overall, the presentations at the Heart Rhythm Society meeting suggest that ICDs should be more widely utilized in the primary prevention of SCD in patients with LV dysfunction of any cause. It is hoped that the US Food and Drug Administration and CMS will approve this expanded indication in the forthcoming year.

References

1. Moss AJ, Hall WJ, Cannom DS, et al. Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia. Multicenter Automatic Defibrillator Implantation Trial Investigators. N Engl J Med. 1996;335:1933-1940.
2. Buxton AE, Lee KL, Fisher JD, et al. A randomized study of the prevention of sudden death in patients with coronary artery disease. Multicenter Unsustained Tachycardia Trial Investigators. N Engl J Med. 1999;341:1882-1890.
3. Moss SJ, Zareba W, Hall WJ, et al. Prophylactic implantation of a defibrillator in patients with myocardial infarction and reduced ejection fraction. N Engl J Med. 2002;346:877-883.
4. Kadish A, Dyer A, Daubert JP, et al. Prophylactic defibrillator implantation in patients with nonischemic dilated cardiomyopathy. N Engl J Med. 2004;350:2151-2158.