Amiodarone or an ICD for Congestive Heart Failure

Abstract & Commentary

Synopsis: Patients with New York Heart Association class II or class III congestive heart failure and a left ventricular ejection fraction of 35% or less benefit from ICD therapy but not from amiodarone therapy.

Source: Bardy GH, et al. For the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) Investigators. N Engl J Med. 2005;352:225-237.

The Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) was designed to test the hypothesis that amiodarone or an implantable cardioverter defibrillator (ICD) would decrease the risk of death in patients with mild to moderate heart failure. Between 1997 and 2001, 2521 patients were randomized to receive either placebo, amiodarone, or a single chamber ICD programmed to a shock only mode. Patients had to have New York Heart Association (NYHA) class II or class III chronic stable heart failure. They could have either ischemic or nonischemic heart disease. They also had to have left ventricular systolic dysfunction with a measured ejection fraction of £35%. The primary end point for the trial was death from any cause. Pharmacologic treatment for heart failure was monitored throughout the trial and included, whenever possible, at least a beta adrenergic blocker and an angiotensin converting enzyme inhibitor or angiotensin receptor blocker. Aldosterone antagonists, aspirin, statins, and diuretics were used as appropriate. Placebo and amiodarone were administered in a double-blind fashion. After a loading period, patients weighing more that 200 lbs received 400 mg daily, patients weighing 150-200 lbs received 300 mg daily, and patients weighing less than 150 lbs received 200 mg daily. The amiodarone dose could be adjusted for bradycardia. All patients received a single lead ICD which was programmed to shock with a VF detection cut-off of 187 bpm or more. Back-up bradycardia pacing was programmed at 34 bpm.

Pairwise comparisons of amiodarone with placebo and ICD with placebo were performed according to the intention to treat principle. All statistical tests were 2-tailed.

In SCD-HeFT, 847 patients were randomly assigned to placebo, 845 to amiodarone, and 829 to ICD therapy. At baseline, the median left ventricular ejection fraction was 25%, 70% had New York Heart Association class II congestive heart failure, and 30% had class III heart failure. Mean follow-up in the study was 45.5 months.

In the drug treatment arms, 22% of the patients on placebo and 32% of the patients on amiodarone discontinued the double-blind study drug. The only complications observed in the amiodarone group, as compared with the placebo group, were increased tremor (4%) and increased hyperthyroidism (6%). A total of 125 patients (7%) in the drug groups crossed over to open label treatment with amiodarone at some point. Although not specified in the paper, this was presumably because of symptomatic atrial or ventricular arrhythmias.

Among the 829 patients assigned to ICD therapy, 17 declined implantation, and implantation was unsuccessful in 1. An additional 32 patients had their ICD removed during follow-up. Although the cause for this was not specified, this was presumably because of infection. Cross-over to some form of ICD therapy during follow-up occurred in 188 patients (11%) in the drug groups. The median time from randomization to cross-over was 26.7 months. In the ICD group, 259 patients (31%) received shocks from their device. Of these, 177 patients received shocks for rapid ventricular tachycardia or fibrillation. During 5 years of follow-up, the average rate of ICD shock was 7.5% for all shocks and 5.1% for appropriate shocks.

There was no mortality difference between the amiodarone group (29%) and the placebo group (28%). In contrast, mortality was only 22% in the ICD group (hazard ratio vs placebo, 0.77; 97.5 confidence interval 0.62 -0.96; P = 0.007). The relative risk reduction for ICD therapy compared with placebo was 23%, and the absolute reduction in mortality at 5 years was 7.2%.

Mortality curves and hazard ratios were also reported for certain prespecified groups. There was no difference in response to ICD therapy between patients with either ischemic or nonischemic causes for their congestive heart failure. A significant interaction between amiodarone and NYHA class was noted. Among patients with NYHA class III CHF, there was a relative 44% increase in the risk of death associated with amiodarone, as compared to those in the placebo group. This excess risk of death with amiodarone was not noted in patients with class II congestive heart failure. There was also an interaction between ICD therapy and NYHA class. Among patients with NYHA class II CHF, there was a 46% relative reduction in the risk of death, with an absolute reduction in mortality at 5 years of 11.9%. In contrast, patients with NYHA class III congestive heart failure had no apparent reduction in the risk of death with ICD therapy, compared with placebo. Other subgroups were also examined. The point estimates showed no benefit with ICD therapy vs placebo among women and in those with left ventricular ejection fractions greater than 30%. However, the confidence intervals for these subgroups were wide and these estimates should be interpreted with caution.

Bardy and colleagues conclude that patients with New York Heart Association class II or class III congestive heart failure and a left ventricular ejection fraction of 35% or less benefit from ICD therapy but not from amiodarone therapy.

Comments by John P. DiMarco, MD, PhD

The SCD-HeFT trial is truly a landmark study. The size of the trial gave it sufficient statistical power to answer several important questions. First, the data clearly show that empiric amiodarone in asymptomatic individuals with heart failure does not improve survival.

This question has been examined in a number of previous trials with inconclusive results. The data from SCD-HeFT should answer this question, since clearly amiodarone had no benefit in patients treated with an appropriate heart failure regimen including ACE inhibitors and beta blockers. The data for ICD therapy are also consistent with prior trials. Most previous studies have reported a 20-35% relative reduction in mortality associated with ICD therapy. The exceptions to this were the Coronary Artery Bypass Graft Trial, which showed no benefit with epicardial ICD systems after bypass surgery and the DINAMIT Trial, which showed no benefit in patients with recent (within 6-40 days) myocardial infarction and abnormal heart rate variability. SCD-HeFT confirms that the benefit of ICD therapy is seen not only in patients with ischemic heart disease but also in patients with nonischemic cardiomyopathy. However, the observation that there was no benefit in patients with class III congestive heart failure, but benefit in class II heart failure is the opposite of the results described in the DEFINITE Trial. The latter study was a study involving over 400 patients with nonischemic cardiomyopathy and that study showed increased benefit in patients with class III heart failure. The reason for this is unknown. However, because of this discrepancy, one should not make clinical decisions based on these subgroup analyses.

Based on the results from SCD-HeFT, the Center for Medicare Services (the CMS) has recently revised guidelines for implantation of defibrillators. According to these new guidelines, all patients with left ventricular ejection fractions below 35% and class II or class III congestive heart failure will be eligible for ICD implantation. No longer will they have to have prior myocardial infarctions or prolonged QRS durations. Importantly, patients with only class I failure are now excluded. Based on the Cabbage Patch Trial and DINAMIT, patients with recent revascularization or myocardial infarction are also excluded.

In SCD-HeFT, Bardy et al report that all patients could be defibrillated with a 30 joule shock. Based on this observation, they make a statement in their discussion that defibrillation testing at the time of ICD implant might be eliminated. Unfortunately, this statement is probably not justified by their data. Limited defibrillation testing was part of the SCD-HeFT protocol. One would assume that if the initial shock did not work, the implanting physician would have repositioned the lead. The fact that all patients could be defibrillated does not mean that all were defibrillated on the first attempt, and the paper does not present evidence that such an approach would be safe. With this one exception, however, this is a truly landmark study that should strongly influence practice.

Dr. Di Markco, Professor of Medicine, Division of Cardiology, University of Virginia, Charlottesville, is on the Editorial Board of Clinical Cardiology Alert.