Current issues of ACP Journal Club are published in Annals of Internal Medicine


Amiodarone reduced arrhythmic death but not cardiac death in patients with MI and depressed left ventricular function

ACP J Club. 1997 Sep-Oct;127:31. doi:10.7326/ACPJC-1997-127-2-031

Related Content in this Issue
• Companion Abstract and Commentary: Amiodarone reduced ventricular fibrillation or arrhythmic death in MI with frequent ventricular premature depolarizations

Related Content in the Archives
Review: Class II (β-blockers) and class III (amiodarone) antiarrhythmic agents reduce mortality in acute myocardial infarction

Source Citation

Julian DG, Camm AJ, Frangin G, et al., for the European Myocardial Infarct Amiodarone Trial Investigators. Randomised trial of effect of amiodarone on mortality in patients with left-ventricular dysfunction after recent myocardial infarction: EMIAT. Lancet. 1997 Mar 8; 349:667-74.



To evaluate the effect of amiodarone on all-cause mortality, cardiac mortality, and arrhythmic death in survivors of a myocardial infarction (MI) with depressed left ventricular function.


Randomized, double-blind, placebo-controlled trial with a median 21-month follow-up.


75 coronary care units in 15 European countries.


1486 patients between 18 and 75 years of age (mean age 60 y, 84% men) who had had an MI and had a left ventricular ejection fraction (LVEF) ≤ 40%. Exclusion criteria were use of amiodarone in the previous 6 months, bradycardia, second- or third-degree atrioventricular block, sinus pauses > 2.5 seconds unless controlled by a pacemaker, hepatic disease, history of thyroid dysfunction, long QT syndrome, severe angina, or congestive heart failure refractory to conventional therapy, a need for antiarrhythmic therapy other than β-blockers or digoxin, need for cardiac surgery, contraindications to amiodarone, or childbearing potential. 2 patients were lost to follow-up.


743 patients were allocated to amiodarone, 800 mg/d for 14 days followed by 400 mg/d for 14 weeks and then 200 mg/d until the end of the study. 743 patients were allocated to placebo.

Main outcome measures

All-cause mortality, cardiac mortality, and arrhythmic death.

Main results

Analysis was by intention to treat. 205 patients died; 174 cardiac deaths occurred, of which 83 were arrhythmic. All-cause mortality (P = 0.96) (Table) and cardiac mortality (85 vs 89) did not differ between the treatment groups. 33 patients in the amiodarone group had an arrhythmic death compared with 50 patients in the placebo group (P = 0.05) (Table).


Amiodarone reduced arrhythmic death but not all-cause mortality or cardiac mortality in survivors of a myocardial infarction with depressed left ventricular function.

Source of funding: Sanofi Recherche.

For article reprint: Professor A.J. Camm, St. George's Hospital Medical School, London SW17 ORE, England, UK. FAX 44-181-767-7141.

Table. Amiodarone vs placebo*

Outcome at median 21 months Amiodarone EER Placebo CER RRR (95% CI) ARR |EER-CER| (CI) RRI (CI) ARI (CI)
Arrhythmic deaths 4.4% 6.7% 35% (0 to 58) 2.3% (0 to 4.7)
All-cause mortality 13.9% 13.7% 1.0% (-30.1 to 21.6) 0.2% (-3.7 to 3.4)

*Abbreviations defined in Glossary; RRR, ARR, RRI, ARI, and CI calculated from data in article.


Physicians desire adjunctive drugs to prolong life in survivors of acute MI, especially those with poor left ventricular function. Of the strategies evaluated to date, adrenergic receptor blockade and angiotensin-converting enzyme (ACE) inhibition clearly improve survival (1). However, some strategies that were thought to promote electrical stability (e.g., sodium-channel blockers) actually shorten life (2), even in low-risk patients. Earlier studies suggested that high-risk patients might benefit from antiarrhythmic therapy with amiodarone. Besides prolonging the ventricular refractory period, amiodarone has adrenergic blocking and anti-ischemic actions. These 2 trials, the Canadian Amiodarone Myocardial Infarction Arrhythmia Trial (CAMIAT) and the European Myocardial Infarct Amiodarone Trial (EMIAT), show that the outcome of treatment with amiodarone is complex and not readily predictable.

CAMIAT focused on patients with frequent ventricular ectopy, and EMIAT focused on patients with left ventricular dysfunction. In CAMIAT, the primary outcome measure was arrhythmic death or RVF and the results were assessed using a 1-tailed t-test. Patients withdrew at high and unequal rates, 36% of patients in the amiodarone group and 25% in the placebo group, possibly confounding the results. In the ITT analysis, amiodarone did not show a statistically significant decrease in total mortality or nonarrhythmic cardiac mortality. Taken together, these factors weaken the authors' conclusion that amiodarone produces "a clinically important reduction in arrhythmic death and RVF." Although the design of EMIAT was better, the trial's 33% increase in nonarrhythmic death, which may have been caused by amiodarone, offset the 35% reduction in risk for arrhythmic death. 42 (CAMIAT) and 44 (EMIAT) patients would need to be treated to prevent 1 arrhythmic death.

The 2 trials remind us of the weakness of arrhythmic death as an outcome measure. Although committees try to distinguish arrhythmic or sudden death from other types of cardiac death, assessments using these end points fail because sudden death and arrhythmic death overlap but are not identical. Some sudden deaths are caused by ventricular arrhythmia, but others are caused by re-infarction, pulmonary embolism, and bradycardia, for example (3). Inconsistency between committees in classifying deaths (4) also hampers systematic review. Therefore, the objective end point, total mortality, remains clinically the most useful.

The results of CAMIAT and EMIAT show that, unlike other antiarrhythmic drugs, amiodarone does not shorten life. Although these trials suggest that amiodarone decreases arrhythmic death, it failed to decrease total mortality in both trials. The clinical effectiveness of amiodarone in prolonging life after MI therefore remains unproved. To evaluate whether amiodarone decreases total mortality requires a larger, more costly trial, or in the interim, a meta-analysis of current trials. Physicians must also weigh the substantial adverse effects of amiodarone against any apparent benefit. The evidence does not justify the routine use of amiodarone in survivors of MI who have left ventricular dysfunction or asymptomatic ventricular arrhythmias.

On average, only half of the patients in these trials were receiving β-adrenergic blockers and ACE inhibitors, although these drugs are proven to increase survival (1). Instead of considering amiodarone to lower the risk for arrhythmic death, physicians should confidently prescribe a β-adrenergic blocker or an ACE inhibitor to survivors of MI where appropriate.

Vernon M. Oh, MD
National University HospitalRepublic of Singapore

Vernon M. Oh, MD
National University HospitalRepublic of Singapore


1. Hennekens CH, Albert CM, Godfried SL, Gaziano JM, Buring JE. N Engl J Med. 1996;335:1660-7.

2. The Cardiac Arrhythmia Suppression Trial (CAST) Investigators. N Engl J Med. 1989;321:406-12.

3. Luu M, Stevenson WG, Stevenson LW, Baron K, Walden J. Circulation. 1989; 80:1675-80.

4. Narang R, Cleland JG, Erhardt L, et al. Eur Heart J. 1996;17:1390-403.