1462 BRIEF REPORTS Ventricular Tachyarrhythmias Complicating Amiodarone Therapy in the Presence of Hypokalemia Hypokalemia was promptly corrected by intravenous potassium. The U wave disappeared and the QTc interval, only 10 days after stopping amiodarone, was normal (360 ms). Case 2. A 53-year-old man without demonstrable heart disease had several episodes of palpitations. Previous elec- VINCENZO SANTINELLI, MD MASSIMO CHIARIELLO, MD CRESCENZO SANTINELLI, MD MARIO CONDORELLI, MD Two patients are described who have long QT interval and syncope due to ventricular tachyarrhythmias related to the presence of hypokalemia and amiodarone. Case I. A 49-year-old woman underwent a mitral commissurotomy in 1977 for mitral stenosis. She was admitted because of syncope. Three months earlier, she began therapy with oral amiodarone (600 mg/day) for control of supraventricular arrhythmia. Two months later, in the absence of hypokalemia, an electrocardiogram (ECG) showed a long QTc interval (600 ms) with a normal QRS duration. Fifteen days before admission, clorthalidone (100 mg every 2 days) was added to amiodarone for mild systemic hypertension. Fifteen days later, syncope due to "unusual" ventricular tachyarrhythmia occurred (Fig. 1 and 2). At that time the serum potassium level was 3.2 mEq/liter and the tests of thyroid function had normal results. Amiodarone was discontinued. After arrhythmia conversion by intravenous mexiletine, a prolonged QT/U interval (600 ms) was evident. From the Istituto di Clinica Medica I, Universit~ di Napoli, II Policlinico, Via S. Pansini 5, 80131, Napoli, Italy. Manuscript received December 9, 1983; revised manuscript received February 15, 1984, accepted FebruarY 16, 1984. FIGURE 2. Patient 1. Unipolar chest leads during ventricular tachyarrhythmia resembling bidirectional ventricular tachycardia. In lead V1 the deflections are always positively directed and have the appearance of right bundle block with the morphologic features of qR; a small q wave precedes a wide, tall R wave. The initial vector is opposite in direction to the r wave seen in lead V1 during sinus beats. In leads V5 to Ve, there are paroxysms of ventricular tachycardia with alternating QRS morphologic characteristics. A markedly prolonged QT interval with prominent U wave is evident. FIGURE 1. Patient 1. Six-lead electrocardiogram showing premature ventricular beats with an alternating QRS axis. Ventricular beat 1 is of the right-axis type and ventricular beat 2 is of the left-axis type (leads I and VL). May 15, 1984 THE AMERICAN JOURNAL OF CARDIOLOGY Volume 53 trophysiologic study showed no abnormalities. Fifteen days before the present admission, therapy with amiodarone (800 rag/day) was begun; because of hyperglycemia (214 mg % ), fast-acting insulin (35 units/day, administered subcutaneously) was added to the amiodarone therapy. Before amiodarone therapy, the QTc interval was 380 ms and the serum potassium level was 4.0 mEq/liter. Fifteen days later, syncope due to ventricular flutter developed (Fig. 3) and required repeated defibrillation. At that time, the blood glucose level was 65 mg% , the serum potassium was 3.2 mEq/liter and the ECG showed a prolonged Q T / U interval (580 ms) (Fig. 3). In the 15 days before syncope, various ECGs showed progressively prolonged QTc intervals in the absence of hypokalemia (380 to 530 ms). Hypokalemia was promptly corrected, and insulin and amiodarone therapy was discontinued. Subsequent ECGs showed a progressive reduction of the QTc interval, but no ventricular irritability. Usually, the combination of quinidine-like drugs with hypokalemia may precipitate ventricular tachyarrhythmiasl; however, in patients receiving amiodarone, hypokalemia has never been reported as an exacerbating factor. Amiodarone-induced polymorphous ventricular tachyarrhythmia in association with hypokalemia has been reported in 2 patients, but ventricular tachyarrhythmia persisted despite the prompt correction of hypokalemia.2 Our experience, however, suggests that hypokalemia may indeed "precipitate" syncope also in patients who receive amiodarone in the same manner as patients receive quinidine. The "unusual" ventricular tachyarrhythmia that developed in Patient 1463 1 only rarely complicates amiodarone therapy, suggesting that hypokalemia should be considered a necessary additional factor. Low potassium levels facilitate increases in spontaneously firing Purkinje cells.3 This enhanced automaticity, together with amiodarone-induced prolongation of action potential duration, may represent a "perfect recipe" for ventricular tachyarrhythmia. Similarly, in Patient 2, despite a previous amiodarone-induced long QT interval and in the absence of hypokalemia, ventricular flutter occurred only coincidentally with the decrease in the serum potassium level. Caution should be used in patients who receive chronic amiodarone therapy and who show disproportionate repolarization changes (prolonged QT/U intervals), especially if potassium depletion is associated. Because of the drug's extremely long half-life, ventricular tachyarrhythmia may be recurrent and difficult to treat; therefore, potentially exacerbating factors such as hypokalemia should be considered serious precipitating factors and prevented as soon as possible. References 1. Khan MM, Logan KR, McComb JM, Adgey AAJ. Management of recurrent ventricular tachyarrhythmias associated with QT prolongation. Am J Cardiol 1981;47:1301-1308. 2. Sclarovsky S, Lewin RF, Krakoff O, Strasberg B, ArditU A, Agmon J. Amiodarone-induced polymorphous ventricular tachycardia. Am Heart J 1983; 105:6-11. 3. Vassalle M. The relationship among cardiac pacemakers. Circ Res 1977; 41:267-277. FIGURE 3. Patient 2. Twelve-lead electrocardiogram showing markedly prolonged QT/U interval (580 ms) with prominent U waves (arrows) while the patient was taking amiodarone (top); coincident with the decrease in the serum potassium level (3.2 mEq/liter), ventricular flutter developed (boUom).