Mevan Wijetunga, MD • S. Adam Strickberger, MD
From the Division of Cardiology, Washington Hospital Center, Washington, DC
Patients with nonischemic dilated cardiomyopathy have a high risk of sudden death caused by lethal ventricular arrhythmias. Although empiric amiodarone does not improve mortality after myocardial infarction in patients with coronary artery disease, it may improve survival in patients with nonischemic dilated cardiomyopathy.1,2 Recent studies have shown that patients with ischemic cardiomyopathy and a low ejection fraction benefit from implantable cardioverter-defibrillator (ICD) therapy.3-5 It is not clear if patients with nonischemic dilated cardiomyopathy benefit from ICD therapy.
The purpose of the Amiodarone Versus Implantable Cardioverter-Defibrillator (AMIOVIRT) study was to compare the mortality rates of patients with nonischemic dilated cardiomyopathy and asymptomatic nonsustained ventricular tachycardia who were randomized to treatment with amiodarone or an ICD in addition to standard medical treatment. It was a randomized trial conducted from August 1996 to June 2001 in 10 medical centers in the United States.
A total of 103 patients who were older than 18 years participated in AMIOVIRT. Of these, 51 were randomized to treatment with a transvenous ICD, and 52 were randomized to treatment with amiodarone. The dose of amiodarone was 800 mg daily during the first week. After the first week, the dose was reduced to 400 mg daily for 1 year. After year 1, the dose was reduced further to 300 mg daily. All patients had a clinical diagnosis of nonischemic dilated cardiomyopathy. The entire cohort had left ventricular ejection fractions less than 35%, documented nonsustained ventricular tachycardia, and New York Heart Association functional classifications of I to III. Standard medical therapy of the underlying ventricular dysfunction with angiotensin-converting enzyme (ACE) inhibitors, beta blocking agents, and spironolactone was encouraged. Patients with newly diagnosed non-ischemic dilated cardiomyopathy were excluded.
We gathered data on death, cause of death, arrhythmia-free survival, quality of life, and the cost of treatment in each patient group. Follow-up was 100%. A committee of physicians blind to the patients’ randomization group evaluated and determined the causes of death. Arrhythmia-free survival was determined from the absence of death, syncope, life-threatening ventricular arrhythmias, or appropriate ICD discharges. Quality of life was assessed with two standard questionnaires: Quality of Well Being Schedule and State Trait Anxiety Inventory. To explore the cost of inpatient and outpatient management, data were obtained from 24 patients, all of whom were treated at one center.
We performed an intention-to-treat analysis; the analysis was based on the initially assigned treatment for each patient. Survival curves for each treatment group were plotted using the Kaplan-Meier method. The Mantel test was used to compare the survival curves in each group. At the first interim analysis
of the data, the previously determined stopping rule for futility was reached. Subsequently, the study was stopped.
The baseline characteristics were similar in both treatment groups. Among all the patients, the mean age was 59 ± 11 years, 30% were women, and the mean left ventricular ejection fraction was 23% ± 9%. The duration since the diagnosis of nonischemic dilated cardiomyopathy was 3.2 ± 4.0 years.
The patients were followed up for 2.0 ± 1.3 years. Medical management between the groups was similar. Patients were generally treated with an ACE inhibitor (85%), a beta blocker (52%), and a potassium-sparing diuretic (20%). The quality-of-life assessments at baseline were similar between the patients in each group.
We found no statistically significant difference in the 1- and 3-year survival rates in the patients who received amiodarone compared with those who received an ICD (P = .8; figure 1). There was a trend toward improved arrhythmia-free survival rates (P = .1; figure 2) and cost of medical care ($8,879 versus $22,079; P = .1) in the amiodarone-treated patients. The quality of life was similar between the groups (P = .1).
The results showed that 1- and 3-year survival rates were not statistically significant in patients with nonischemic dilated cardiomyopathy and nonsustained ventricular tachycardia treated with amiodarone compared with an ICD. We observed trends toward improved arrhythmia-free survival and lower costs of medical care in patients who received amiodarone compared with those who received an ICD. These results are similar to those of Bansch and colleagues.6 In 2002, they documented no differences in survival in patients with newly diagnosed asymptomatic nonischemic dilated cardiomyopathy who were treated with an ICD compared with standard medical therapy.
Our findings in patients with nonischemic dilated cardiomyopathy are in sharp contrast to the results of recent trials that included asymptomatic patients with ischemic car-
diomyopathy and decreased left ventricular function. In patients with ischemic cardiomyopathy, prophylactic implantation of an ICD is superior to standard medical therapy.7
One possible explanation for these observations is that ischemic and nonischemic myocardial substrates are different. Hence, amiodarone may be more effective in preventing death in patients with nonischemic dilated cardiomyopathy than in patients with ischemic cardiomyopathy.
A limitation of our study is that we did not include a standard medical treatment group in addition to the amiodarone and ICD groups. We prospectively recognized that it was not feasible to have a control group because of the sample size requirements a third group would mandate. Without a control group, our study does not exclude the possibility that neither amiodarone nor an ICD affects mortality or improves clinical outcomes in nonischemic dilated cardiomyopathy. It is reasonable to think, however, that amiodarone lowers the risk of sudden cardiac death by decreasing the occurrence of ventricular tachycardia and ventricular fibrillation. This was represented in our study by the trend toward improved arrhythmia-free survival in patients treated with amiodarone.
Another limitation is the small sample size. However, the study was stopped because we used the prospective stopping rule for futility to identify the lack of difference between the treatments. The mortality rates in the study were lower than expected in both treatment groups. Based on the results, we estimated that approximately 12,000 patients would have to be randomized to achieve a statistical power of 80%.
The use of ICD therapy has grown exponentially in the United States because of the proven benefit of this therapy in patients with ischemic
cardiomyopathy. Our findings, however, suggest that ICD therapy may not improve survival in all patients with left ventricular dysfunction,
such as those with nonischemic di-
lated cardiomyopathy. Our results suggest that treatment with amiodarone may be reasonable as an initial approach to preventing sudden cardiac death in patients with nonischemic dilated cardiomyopathy and asymptomatic nonsustained ventricular tachycardia.