Intravenous amiodarone versus propafenone for atrial fibrillation and flutter after cardiac surgery: a note of caution

Sir, We read with interest the paper written by Di Biasi et al. (Eur J Cardiothorac Surg 9: 587-591) about the efficacy of amiodarone and propafenone in converting postoperative atrial fibrillation and flutter after cardiac surgery. The conventional management of postoperative atrial fibrillation and flutter involves control of the ventricular rate with digoxin, verapamil or low-dose beta-blockers. Despite the effectiveness of these agents, there are several effects that might complicate their use, especially the pro-arrhythmic potential. For this reason, the efficacy and safety of newer anti-arrhythmic drugs, sotalol and amiodarone, have been investigated intensively in a more recent period. Although the paper by Di Biasi confirms previous observations, it does not underline some of the potentially severe complications related to amiodarone therapy, especially in the early postoperative course after cardiac surgery, such as profound hypotension, myocardial depression and atropin-resistant bradycardia [1, 2, 5]. In 1992 we started a prospective, randomized study on the efficacy of high-dose intravenous, short-term amiodarone versus digoxin combined with verapamil (1 mg intravenous digoxin during the first 24 h followed by 0.25 mg/day with, in addition, repeatable 1.25 mg i. v. verapamil to control heart rate when the pulse was higher than 110/min). However, this trial had to be interrupted after the inclusion of 23 patients, because of the occurrence of severe side effects observed in the amiodarone group. Some pertinent observations are presented briefly here. The inclusion criteria of the study were postoperative atrial fibrillation and/or flutter within 4 days following coronary bypass and/or valvular surgery, lasting for at least 15 ra in -2 h. The study included the following loading protocol: amiodarone 5 mg/kg over 15 min, followed by an infusion of 1500 mg/12 h and peroral amiodarone 2×800 mg from day 2 to day 4, reducing the dosage to 2x400 mg at day 5 and continuing therapy until hospital discharge. Reduction of dosage was planned in cases of significant bradycardia (<60/min) and when amiodarone serum levels exceeded 2 .5 3.0 gg/ml. The main exclusion criterias were preoperative atrial fibrillation or flutter, atrio-ventricular block II, sick sinus with bradycardia less than 50/min, amlodarone therapy up to 6 monhs preoperatively, treatment with digitalis, verapamil and ]3-blockers in the first 24 h postoperatively as well as prolonged QT interval. During the infusion of amiodarone, intensive care unit monitoring was carried out and a 48 h HolterECG was registered. The arterial pressure and ventricular rate were recorded continuously. Blood samples for amiodarone and desmethylamiodarone serum analysis were obtained 30 min after starting the amiodarone treatment, at 1, 3, 6 and 12 h, and at days 2 and 4 as well as at hospital discharge. The serum was analyzed using high pressure liquid chromatography. The serum concentrations obtained in the available patients are summarized in Fig. 1. A high-dose intravenous loading protocol in which amiodarone is administered in a fixed dose regimen achieved the recommended target serum concentration of 2 .0 -3 .0 gg/ml in all patients; however, the metabolite desmethylamiodarone did not reach a therapeutic level. The study was interrupted after 23 patients had been randomized, 12 in the amiodarone group and 11 in the verapamil-digitalis group. The indicence of conversion to sinus rhythm within 24 h was similar in both groups, 10/12 patients after amiodatone and 10/11 patients after verapamil-digitalis, but there were more patients converting within the 1 st hour following bolus injection of amiodarone than after starting the control treatment. The effect on ventricular rate was more evident after amiodarone treatment (mean heart