Amiodarone vs. ICD for Ventricular Tachyarrhythmias

abstract & commentary

Synopsis: Amiodarone therapy results in a relative risk reduction in all-cause mortality that is similar to that reported in other studies.

Source: Connolly SJ, et al. Circulation 2000;101:1297-1302.

The canadian implantable defibrillator study (CIDS) was a randomized clinical trial looking at the relative effects of amiodarone therapy vs. implantable cardioverter defibrillator (ICD) therapy in patients with a history of known or suspected ventricular arrhythmias. Patients were eligible for the study if they had an episode of one of the following: 1) cardiac arrest or ventricular fibrillation (VF), 2) sustained ventricular tachyarrhythmia (VT) causing syncope, 3) sustained VT at a rate of more than 150 bpm with an ejection fraction less than or equal to 35%, or 4) syncope of unknown origin with sustained VT documented at other times or induced by programmed stimulation. Patients were randomized to receive either ICD therapy or drug therapy with amiodarone. Initially, the ICD was implanted using a thoracotomy lead system but, early in the trial, improved technology permitted a switch to nonthoracotomy lead systems. Patients randomized to amiodarone were loaded with 1200 mg/day for one week or longer in the hospital. Subsequently, they received 400 mg/day for 10 weeks and then 300 mg/day. Additional antiarrhythmic drugs were permitted in both groups to control either supraventricular or nonsustained ventricular arrhythmias that were symptomatic. The primary outcome event was all-cause mortality. Arrhythmic death was a secondary event.

Between October 1990 and January 1997, 659 patients were entered into the trial. The group was predominantly (85%) male. Slightly less than half had survived an episode of VF or cardiac arrest. Slightly less than 40% had a history of sustained ventricular tachycardia and slightly more than 13% had unmonitored syncope. The mean left ventricular ejection fraction was 34%. About 80% had ischemic heart disease.

There were 328 patients who were randomly allocated to receive an ICD, but 18 patients did not receive one. The median time to ICD implant was seven days. A thoracotomy lead system was used in 33 patients. A nonthoracotomy system was used in 277 patients. There was one operative death in both groups. Among the 18 patients who did not receive an ICD, there were seven deaths in-hospital while awaiting surgery and 11 patients either decided against implant after randomization or a technical problem precluded a successful implant. During the course of the study, 16 patients had their ICD explanted because of infection, heart transplant, or patient preference. The percentage of patients randomized to an ICD who also began amiodarone therapy at one, three, and five years after randomization were 17.4%, 21.7%, and 28.1%, respectively.

Among patients randomized to receive amiodarone, all patients started therapy and at one, three, and five years, 88.7%, 80.3%, and 85.4% of surviving patients remained on therapy. The mean amiodarone dose at three years was 262 mg/day. Among the 331 patients in the amiodarone group, 52 (15.8%) also received an ICD.

Other antiarrhythmic drugs were added to both groups at hospital discharge. Beta-adrenergic blockers were prescribed in 21% of patients on amiodarone and in 34% of patients who received an ICD. Almost 20% of ICD patients were also receiving sotalol. This imbalance in concomitant drug therapy persisted throughout the trial. There was a 19.7% relative risk reduction with ICD therapy compared with amiodarone, with total mortality decreased from 10.2% per year to 8.3% per year (P = 0.14). There were also reductions in arrhythmic death and total cardiac death in the ICD group that did not reach statistical significance. The relative risk reduction for total mortality at one, two, and three years was 15.4%, 29.7%, and 13.7%, respectively. Net adverse effects to ICD therapy and amiodarone treatment were similar. Analysis of subgroups showed that the 95% confident intervals of hazard ratios for outcome of death from any cause overlapped for all subgroups examined.

Connolly and colleagues conclude that amiodarone therapy results in approximately a 20% relative risk reduction in all-cause mortality that is similar to that reported in other studies, particularly the Antiarrhythmics vs. Implantable Defibrillator (AVID) study.

Comment by John P. DiMarco, MD, PhD

Approximately 10 years ago, three large multicenter trials were initiated to test whether ICD therapy was superior to drug therapy for patients with life-threatening ventricular arrhythmias. AVID, a study performed primarily in the United States, randomized slightly more than 1000 patients and showed that the ICD was superior to drug therapy with amiodarone or sotalol. CIDS, reported in this paper, is the second large trial to be fully reported. There were minor differences between AVID and CIDS. AVID allowed either amiodarone or sotalol to be used in the drug treatment group but most patients received the former. AVID did not allow patients with unmonitored syncope into the trial. About 13% of the patients in CIDS had this diagnosis. It is interesting to note that the point estimate for risk reduction in this group was the lowest observed. Most of the other outcome data are similar to those previously reported in AVID. There is less benefit with the ICD in patients with higher ejection fractions. The benefit of the ICD in absolute terms is modest. In this study, the absolute risk reduction at two years was a little more than 6%.

This paper may be a closer mirror of clinical practice than was the AVID protocol. In AVID, investigators were strongly discouraged from adding an antiarrhythmic drug to patients in the ICD group. In CIDS, antiarrhythmic drugs were used in more than half of the ICD patients. Other studies have shown that addition of an antiarrhythmic drug to ICD therapy has little effect on mortality but may decrease the frequency of shocks and make ICD therapy more acceptable.

There is a third study, the Cardiac Arrest Study Hamburg (CASH), which has been reported in preliminary form only. This study enrolled only patients with cardiac arrest and did not include patients with ventricular tachycardia without cardiac arrest. CASH has also reported a trend toward reduction in mortality in patients treated with the ICD. These three trials have been subject to a meta-analysis. The meta-analysis confirmed the benefit of ICD therapy over drug therapy.

In the future, the ICD should be the primary option for patients with life-threatening arrhythmias, with drug therapy used as either an adjunct to the ICD or as primary therapy for patients who decline ICD therapy or are poor candidates for it.

Recent trials suggest that life-threatening ventricular arrhythmias are best treated with:

a. sotalol.

b. amiodarone.

c. beta blockers.

d. ICD.