Is amiodarone safe in heart failure

Heart failure is a complex clinical syndrome that may lead to sudden cardiac death. Several trials have studied whether the antiarrhythmic agent amiodarone reduces such mortality in heart failure,1 and two trials in particular have been at the centre of the debate: GESICA and CHF-STAT. In the GESICA trial, treatment with amiodarone was associated with reduced mortality, but the study was carried out in Argentina, with its genetically distinct population, where conditions other than a prior myocardial infarction—for example, Chagas disease—may be associated with heart failure. Therefore, caution should be used in extending the results to North American and European settings. Furthermore, GESICA was not blinded, and results, including an exceedingly optimistic 3% dropout rate, might have been affected by a treatment bias.1 The CHF-STAT trial was placebo controlled, and more than 70% of the participants had heart failure of ischaemic origin. Despite the beneficial effect of amiodarone on left ventricular systolic function in this trial, the drug had no impact on survival. The conflicting results of the GESICA and CHF-STAT trials have been attributed to differences in the participants' severity of disease.1 Patients in the GESICA study presented with advanced heart failure, and this may have increased the chance that amiodarone would reduce mortality. This explanation does, however, not accord with the results of the recently published SCD-HeFT trial,2 which, besides having confirmed the importance of implantable cardioverter defibrillators in preventing sudden cardiac death, reported a significant increase in mortality among patients with New York Heart Association class III, but not class II, heart failure in the group treated with amiodarone. In addition, the trial found no association between aetiology of heart failure and outcome in the amiodarone arm of SCD-HeFT. This contradicts the findings of GESICA and CHF-STAT, that amiodarone provides greater benefit in non-ischaemic than in ischaemic cardiomyopathies.1 The SCD-HeFT trial had good statistical power, with 2521 patients followed up for a median of 3.8 years. GESICA and CHF-STAT, on the other hand, enrolled 1190 patients combined, and both trials lasted only two years. Subgroup analysis may be misleading, but can the seemingly adverse action of amiodarone observed in SCD-HeFT really be dismissed? The answer to this question may be found by analysing the factors that may underlie any such undesired outcome. Amiodarone interferes with ion channels for sodium, potassium, and calcium as well as with α and β adrenergic receptors. These actions lead to diverse electrophysiological effects, including reduced automatism and decreased conduction velocity. Amiodarone's pharmacodynamics, with an exceedingly large distribution volume leading to accumulation in the myocardium, may also play a part. Bradyarrhythmias could ensue, leading to sudden cardiac death in advanced heart failure.3 Amiodarone has been associated with tachyarrhythmias as well, such as torsades de pointes,4 which occurs typically in patients with hypokalaemia and hypomagnesaemia. Many patients with heart failure are treated with diuretics, and maintaining their balance of plasma electrolytes is not always easy. Moreover, evidence from large randomised trials shows that diuretics may increase the risk of sudden cardiac death in patients with left ventricular dysfunction.5 Amiodarone can also convert a non-sustained ventricular tachycardia into a sustained one and facilitate the induction of this potentially fatal arrhythmia.4 Another potential complication may arise from using amiodarone with digoxin, a drug with a notoriously narrow therapeutic window. The plasma concentration of digoxin is increased by amiodarone, thus increasing the risk of toxicity and of both bradyarrhythmias and tachyarrhythmias.6 Compounding these difficulties are the common problems in heart failure of electrolyte abnormalities, impaired renal function, altered acid-base balance, and increased sympathetic tone, all of which may also predispose to toxicity. Coincidentally, one should note that the DIG trial,7 validating the use of digoxin in heart failure, lasted in excess of three years and enrolled 6800 patients. Any problem associated with using amiodarone is important, given its widespread use in heart failure, going well beyond trying to prevent sudden cardiac death. For example, amiodarone is the only antiarrhythmic agent that can be used in prophylaxis against atrial fibrillation since others, such as class I agents, are contraindicated.8 The SCD-HeFT trial was not designed to test the safety of amiodarone in heart failure. Nevertheless, unanticipated results of clinical trials—for example, those associated with cyclo-oxygenase 2 inhibitors—should not be ignored. It must also be borne in mind that the current selection process for implantation of a cardioverter defibrillator, based essentially on the evaluation of left ventricular systolic function, is not cost effective.9 Consequently, many patients could be exposed to amiodarone without protection via an implantable cardioverter defibrillator against bradyarrhythmias and tachyarrhythmias. Lastly, it is important to remember past negative experiences with antiarrhythmic agents10 as well as the recent discontinuation of the ANDROMEDA trial owing to greater mortality associated with use of dronedarone, a non-iodinated derivative of amiodarone, in patients affected by left ventricular ventricular dysfuntion.11 On the basis of all of these considerations we believe that a further large scale, prospective trial is warranted to evaluate the safety of amiodarone in heart failure.