CAN AMIODARONE PULMONARY TOXICITY BE PREDICTED IN PATIENTS UNDERGOING IMPLANTABLE CARDIOVERTER-DEFIBRILLATOR IMPLANTATION

Implantable cardioverter defibrillator (ICD) implantation is rapidly becoming accepted as primary therapy for malignant ventricular arrhythmias. Many patients undergoing ICD implantation are on concomitant antiarrhythmic drugs to decrease shock frequency, slow tachycardia rate, and suppress supraventricular arrhythmias. Amiodarone is a potent antiarrhythmic agent that is also frequently used in the treatment of patients with refractory ventricular arrhythmias. Ten to forty percent of patients undergoing ICD implantation will also be taking amiodarone. It has been reported to cause pulmonary toxicity in about 5% of patients per year. Acute amiodarone toxicity presenting as adult respiratory distress syndrome has been reported much less frequently. Although perioperative morbidity due to amiodarone has been described, the risk, predictability, and consequences of acute pulmonary toxicity from amiodarone in patients undergoing ICD implantation have not been previously described. We reviewed the records of 99 consecutive patients undergoing ICD implantation at our institution from October 1987 to April 1992. Thirty-nine patients were taking 480 +/-230 mg of amiodarone (median 400 mg, lower 20th percentile 400 mg, upper 80th percentile 800 mg) for 291 +/- 554 days prior to ICD implantation. Ten patients taking amiodarone developed acute pulmonary toxicity clinically manifesting as diffuse pulmonary infiltrates on chest radiography and adult respiratory distress syndrome with hypoxia (arterial pO2 < 60 mmHg) without evidence of pneumonia or elevated pulmonary capillary wedge pressure (PCW less-than-or-equal-to 15 mmHg). Of the 60 patients not taking amiodarone none developed adult respiratory distress syndrome. There was no relationship between the clinical variables of age, type of anesthesia, ejection fraction, dose or duration of amiodarone use, number of intraoperative ICD test shocks, operative time, or intraoperative FI(O2). Compared to patients on amiodarone who did not develop toxicity, patients with amiodarone pulmonary toxicity had prolonged intensive care unit stays (17.1 +/- 16.9 days vs 3.5 +/- 1.5 days, P < 0.001), as well as prolonged ventilator dependence (9.1 +/- 18.7 days vs 0.9 +/- 1.1 days, P = 0.02). Acute amiodarone pulmonary toxicity resulted in a widely variable clinical syndrome that occurred 1-5 days after surgery. In conclusion, pulmonary toxicity was not predicted by clinical or hemodynamic variables, results in markedly prolonged ventilator dependence and intensive care unit stays, and may result in death.