Clinical Features and Basic Mechanisms of Quinidine-Induced Arrhythmias
Total Page：16
File Type：pdf, Size：1020Kb
JACC Vol. 8, No I 73A July 1986:73A-78A Clinical Features and Basic Mechanisms of Quinidine-Induced Arrhythmias DAN M. RODEN, MD, FACC,* KATHERINE A. THOMPSON, MD,* BRIAN F. HOFFMAN, MD, FACC,t RAYMOND L. WOOSLEY, MD, PHD* Nashville, Tennessee and New York, New York Quinidine therapy is one of the most common causes of is lowered and the stimulation rate is slowed. More re• the acquired long QT syndrome and the morphologically cently, the effects of a number of quinidine metabolites, distinctive tachyarrhythmia torsade de pointes. Clinical as well as the commonly found impurity dihydroquini• data from our institution and others have revealed a dine, were characterized in canine Purkinje fibers in a number ofcharacteristic features: quinidine plasma con• similar fashion. Although quinidine was the most potent centrations are generally low, marked QRS prolongation of the substances tested, both dihydroquinidine and 3• is absent, hypokalemia is frequent and abrupt heart rate hydroxyquinidine prolonged action potential and produced slowing just before the initiation ofa paroxysm is almost early afterdepolarizations as did quinidine at long cycle invariable. The lack ofcorrelation between plasma quin• lengths. idine concentrations and this adverse drug effect raises Quinidine-induced torsade de pointes is a potentially the possibility either that external factors (for example, lethal adverse drug effect, occurring in 1 to 3% of pa• hypokalemia) modulate the response to quinidine in vivo tients. Hypokalemia and slow heart rates are commonly or that one or more unmeasured active metabolites play observed in a clinical setting and, in the tissue bath, a role. quinidine and several ofits metabolites induce abnormal Therefore, the effects of alterations in extracellular automatic behavior when extracellular potassium is low• potassium and stimulation rate on the electrophysiologic ered and stimulation rate is slowed. It is concluded that effects of quinidine were examined in canine Purkinje early afterdepolarizations, a form of triggered auto• fibers. It was found that a form of triggered automat• maticity, may be responsible for induction of quinidine• icity, early afterdepolarizations, is reliably produced in induced arrhythmias in humans. the presence of quinidine when extracellular potassium (} Am Coil CardioI1986;8:73A-78A) It has been recognized since the 1920s that up to 5 to 10% to all polymorphic ventricular tachycardias occurring in the of patients in whom quinidine therapy is initiated will de• setting of marked QT interval prolongation. Polymorphic velop recurrent syncopal episodes, and occasionally die, ventricular tachycardia can of course occur in the absence within the first several days of treatment. Although "quin• of QT prolongation and quinidine-like drugs may be effec• idine syncope" was initially attributed to the vasodilator tive in such cases; it is probably not appropriate to use the properties of the drug, it was recognized in 1964 (1) that term torsade de pointes in such patients (3). in fact a distinctive ventricular tachyarrhythmia was the When Dessertenne described his index case, he specu• cause. In 1966, Dessertenne (2) described a similar tachy• lated that the unusual configuration was due to ventricular cardia, in association with marked QT prolongation, in an activation by at least two automatic foci at slightly different elderly patient with recurrent bradyarrhythmias. He coined rates. Indeed, this configuration can be replicated in this the term "torsade de pointes" which has since been applied fashion in animal models (4). Clinical electrophysiologic studies have shown that premature stimulation can elicit From *The Departments of Medicine and Pharmacology, Vanderbilt polymorphic ventricular tachycardia in some patients re• University School ofMedicine, Nashville, Tennessee and tThe Department ceiving quinidine (5). Although marked QT prolongation of Pharmacology, Columbia University College of Physicians and Sur• geons, New York, New York. This study was supported in part by Grants was not seen in these studies, this initiation sequence never• GM31304 and HL32694 from the United States Public Health Service, theless raised the possibility that reentrant mechanisms were Bethesda, Maryland. Dr. Roden is a recipient of the Clinician Scientist involved. Recently, it has been shown (6) that rapid ven• Award of the American Heart Association, Dallas, Texas. Address for reprints: Dan M. Roden, MD, Department of Pharmacol• tricular pacing after administration of fairly large doses of ogy, Vanderbilt University School ofMedicine, Nashville, Tennessee 37232. quinidine intravenously can induce ventricular tachycardia ©1986 by the Amencan College of Cardiology 0735-1097/86/$3 50 74A RODEN ET AL. JACC Vol. 8, No 1 QUINIDINE-INDUCED ARRHYTHMIAS July 1986 73A-78A in a minority of dogs with coronary occlusion and on car• Examination of the initiation of the paroxysms revealed a diopulmonary bypass; mapping studies in this model sug• strikingly stereotypic pattern of cycle length changes (Fig. gested that multiple automatic foci were implicated. 1). The cardiac cycle immediately preceding the last sinus However, several lines of evidence now suggest that beat before the episode was inevitably longer than the cycle quinidine-induced torsade de pointes is not related to en• it followed, most commonly as a result of a postectopic or hanced normal automaticity or reentry, but represents a clin• posttachycardia pause. The QT interval of the sinus beat ical example of triggered activity. Digitalis intoxication in terminating this long cycle was markedly prolonged and the vitro represents the most widely studied example of trig• arrhythmia began on or just after the peak of the T wave gered activity. Digitalis produces afterdepolarizations that of this QT interval. This phenomenon has been described increase in amplitude as stimulation rates increase and, with in a number of different fashions (7,10), the central obser• sufficiently rapid stimulation rates, can reach threshold and vation being that heart rate appears to slow abruptly just initiate spontaneous action potentials. In contrast to this before the initiation of the arrhythmia. tachycardia-dependent triggered arrhythmia, available evi• Contributing factors. Of the 24 cases at Vanderbilt, 20 dence now implicates bradycardia-dependent triggered ac• occurred within the first 5 days of therapy, but in 4, torsade tivity as the initiating mechanism of quinidine-induced tor• de pointes occurred during long-term quinidine therapy. In sade de pointes. three of these, the episode was clearly associated with the development of the hypokalemia, whereas in a fourth, heart block appeared to be the initiating factor. Some potentially Clinical Studies associated clinical factor (most commonly bradyarrhythmia Incidence and initiation. Since the late 1970s, several or hypokalemia) could be associated with the development series of patients who developed torsade de pointes under of torsade de pointes in 20 ofthe 24 patients. These clinical a variety of clinical circumstances have been described (7,8). data raised several hypotheses that we have begun to test. The common clinical features include marked QT prolon• First, it is possible that this "idiosyncratic" drug reaction gation, a relatively high incidence of hypokalemia or hy• actually represents a predictable effect of quinidine when pomagnesemia, or both, relative inefficacy of standard an• external circumstances are modified (for example, by hy• titachycardia drugs and the efficacy ofmaneuvers to increase pokalemia). Alternatively, the failure of the quinidine con• heart rate (isoproterenol infusion or cardiac pacing). We centrations to be closely associated with the development have had the opportunity to review our experience with a of this adverse drug effect raises the possibility that one or large series of patients (24 patients) with this entity at our more unmeasured metabolites may contribute in part to the institution (9). On the basis ofnonreferral patients only, we variable response. estimated that 2 to 3% of patients starting quinidine therapy will develop this potentially lethal ventricular tachyarrhyth• mia. Approximately one-half of quinidine concentrations In Vitro Studies were below 2 JLg/ml, the lower limit of the usual therapeutic In these studies, action potentials from Purkinje fibers in range, and most serum potassium concentrations in our se• free-running canine false tendons from either ventricle were ries were below 4 mEq/liter, usually in association with studied (11). They were perfused with standard Tyrode's diuretic therapy. QT intervals were markedly prolonged (by solution and bubbled with a mixture of 95% oxygen and definition) whereas QRS intervals were prolonged only 5% carbon dioxide. Extracellular potassium (Ke) varied from slightly, again emphasizing the fact that this reaction is not 2.7 to 5 mM as described later. Preparations were driven a result of poisoning by high concentrations of quinidine. by field stimulation at a wide range of cycle lengths (from 300 to 8,000 ms) and action potentials as well as the max• imal upstroke slope of phase 0 CV-max) were photographed Figure 1. Electrocardiogram of a typical paroxysm of quinidine• from an oscilloscope. In each experiment, baseline mea• induced torsade de pointes demonstrating the typical cycle length surements were made first, and then the fiber was exposed changes just before an episode. The tachycardia starts (light ar• row)