Dr. Cheng describes differences between our article [1] and that of Mickleborough and associates [2]. The latter study retrospectively compared the outcome of two groups of patients with coronary artery disease who had map-guided surgery for ventricular tachycardia; the investigators found that patients receiving amiodarone had a worse clinical outcome after surgery than those who did not receive amiodarone [2]. The investigators concluded that amiodarone should be withheld in patients with ventricular tachycardia until they have been assessed for surgery. Mickleborough and associates are too cautious about amiodarone therapy because they also state that data on "increased" pulmonary complications in patients receiving long-term amiodarone therapy were obtained from anecdotal reports and small uncontrolled series. In addition, their own series is small. The available data, therefore, do not justify the recommendation that surgery for arrhythmias should always be considered before amiodarone therapy is used. For these reasons, our conclusions [1] and the data from Mickleborough and colleagues [2] are not contradictory. Clearly, the low arrhythmogenic potential of amiodarone makes the drug particularly suitable for secondary prevention of sudden death in patients who survive a myocardial infarction [3].

Mickleborough and associates studied a high-risk cohort. The two groups of patients described in their article are not necessarily comparable because patients treated with amiodarone had a significantly lower preoperative left ventricular ejection fraction (0.23% ±11% compared with 0.33% ±10%), had failed significantly more antiarrhythmic drug trials before amiodarone administration, and had had cardioversion more frequently. Of particular importance, their study was retrospective, and observations were made between 1982 and 1991, when relatively high amiodarone doses were administered (on average, 555 ±142 mg/d). The incidence of amiodarone-induced side effects (particularly pulmonary toxicity) is usually dose-related [4]. Mickleborough and associates also selected a cohort with drug-refractory ventricular tachycardia who were at the highest risk regardless of the therapy chosen. Several reports have shown that amiodarone can be safely administered after surgery [5] or even in patients awaiting cardiac transplantation without increasing intra- or postoperative mortality [4].

We agree with Mickleborough and colleagues that patients with serious ventricular tachyarrhythmia should be treated with nonpharmacologic therapy (that is, implanted cardioverter-defibrillator, radiofrequency ablation, or surgery for ventricular tachycardia), whenever possible. However, this is usually difficult in individual patients with ventricular tachycardia or ventricular fibrillation (or both). The relative merit of medical therapy (which includes amiodarone) and of nonpharmacologic therapy is being investigated in several controlled clinical trials.