- Email: [email protected]
Torsade de pointes due to quinidine: Observations in 31 patients
Torsade de pointes due to quinidine: Observations in 31 patients We performed a mail solicitation and obtained the records of 31 patients with documented torsade de pointes (TDP) due to quinidine. All 31 patients had heart disease: ischemic = 11 patients (36%) rheumatic = five patients (16%), hypertensive = four patients (13X), cardiomyopathic = four patients (13%), other = seven patients (22%). Quinidine was administered to these patients for the following reasons: atrial fibrillation or flutter = 22 patients (71%), ventricular premature beats = six patients (19%), ventricular or supraventricular tachycardia = three patients (10%). The 31 patients were receiving quinidine, 650 to 2400 (mean 1097) mg/day, and 14 patients had serum quinidine levels of 1.4 to 10.6 (mean 3.7) pg/ml. TDP occurred within 1 week of initiation of quinidine therapy in 23 (74%) of the patients. Twenty-eight (90%) of the 31 patients were receiving digoxin, and 5 (24%) of 21 patients had hypokalemia at the time of TDP. Off of quinidine therapy, corrected QT (QT.) intervals in 24 patients ranged from 390 to 580 (mean 470) msec and were prolonged in 17 patients (71%). On quinidine therapy, QT, intervals in 23 patients ranged from 390 to 830 (mean 510) msec and were prolonged in 21 patients (91%). In summary, patients with TDP due to quinidine usually had heart disease complicated by atrial fibrillation, were receiving digoxin, and were receiving moderate dosages of quinidine for less than 1 week prior to TDP. Approximately two thirds of patients with TDP due to quinidine had long QT intervals while off of quinidine. (AM HEART J 107:425, 1984.)
Jerry L. Bauman, Pharm.D., Robert A. Bauernfeind, M.D., Julie V. Hoff, Pharm.D., Boris Strasberg, M.D., Steven Swiryn, M.D., and Kenneth M. Rosen, M.D. Chicago,
Ill.
Syncope due to quinidine usually reflects the occurrence of a variety of polymorphic ventricular tachycardia known as torsade de pointes.‘-* It has been estimated that between 1% and 8% of patients receiving quinidine experience syncope.6-7 However, it has been postulated that quinidine syncope does not occur randomly and that some patients could be at increased risk. Reynolds and Vander Arks reviewed the relevant literature and concluded that quinidine syncope tends to occur in patients with atria1 fibrillation complicating rheumatic heart disease, especially when there is concomitant administration of digitalis. However, this conclusion was based upon small series of patients with quinidine syncope reported prior to 1970.
From the Department of Pharmacy Department of Medicine, University
Practice and the Section of Illinois.
of Cardiology,
Supported in part by National Heart, Lung, and Blood Institute Institutional Training Grant HL 07387, National Heart, Lung, and Blood Institute Research Grants HL 18794 and HL 23566, and a grant from the Eleanor B. Pillsbury Resident Trust Fund. Received
for publication
Reprint requests: Jerry sity of Illinois Hospital,
Nov.
18, 1982;
accepted
Dec.
12, 1982.
L. Bauman, Pharm.D., Cardiology Section, PO Box 6998, Chicago, IL 60680.
Univer-
In the present study, we used a mail solicitation to obtain the records of a large series of patients with torsade de pointes due to quinidine, in an attempt to characterize these patients. We also obtained ECGs, which were taken when these patients were on and off of quinidine therapy, and used these ECGs to analyze QT intervals. Our results are germane to identifying patients at increased risk for torsade de pointes due to quinidine. In addition, we were able to make observations regarding the onset of torsade de pointes in some of these patients. METHODS Patient selection. Criteria for inclusion in the present study were: (1) electrocardiographically documented torsade de pointes during quinidine therapy, (2) absence of torsade de pointes (and syncope) prior to quinidine therapy, (3) absence of torsade de pointes (and syncope) after the systemic elimination of quinidine, and (4) availability of sufficient medical and drug history to contribute to our analysis. In this study, we defined torsade de pointes as a rapid ventricular tachycardia with gradually changing QRS morphology (Fig. 1). We collected cases of torsade de pointes due to quinidine primarily by mail solicitation. We sent letters and questionnaires to 2052 cardiologists listed in the American College of Cardiology Directory. In the letter, we outlined
425
March,
426
Bauman
et al.
Fig.
American
1. Rhythm
strip documenting
torsade
the purpose of the study and listed the criteria for inclusion. For each suitable case, we requested copies of rhythm strips documenting torsade de pointes and completion of the questionnaire. Specific information requested in the questionnaire related to: (1) the medical history of the patient, including diagnosed organic heart disease and dysrhythmia, (2) details of quinidine therapy, namely dosage, duration of therapy, and serum levels, (3) details of concomitant drug therapy, (4) serum electrolytes at the time of torsade de pointes, and (5) therapy of torsade de pointes and subsequent course. Finally, we requested 12-lead ECGs taken prior to, during, and after quinidine therapy. In addition to the mail solicitation, we screened our files for the records of referred patients who satisfied the criteria for inclusion. QT interval analysis. Twelve-lead ECGs, taken both on and off of quinidine therapy, were used to analyze QT intervais. Patients with bundle branch block (QRS duration L 120 msec) were excluded from this analysis. QT intervals were measured from the beginning of the QRS complex to the end of the T wave. U waves were carefully excluded from these measurements by the methods of Lepeschkin and Surawicz8 In patients with sinus rhythm, the standard ECG lead displaying the longest QT interval was used to determine corrected QT (QT,) interval. In patients with atria1 fibrillation, mean QT intervals and cycle lengths were measured in the ECG lead displaying the most easily discernible T waves.8 QT, interval was calculated by the method of Bazett,a with the use of the following equation: QT, = measured QT/m The upper limit of normal for QTe, as determined by this method, was considered to be 440 msec for both men and women. QT intervals were also evaluated by the method of Simonson et al.,‘” with the use of the following equation: Calculated QT = 0.2423 + 0.140 (R-R) + 0.0003 (age) t 0.0164 (SEE). By this method, the measured QT interval was considered to be prolonged when it was greater than the calculated QT interval plus two standard errors of the estimate.‘O RESULTS
A total of 31 cases fulfilled the criteria for inclusion. Of these 31 cases, 21 were collected by the mail solicitation and 10 were retrieved from our files. Patient
characteristics
and drug
history.
Of the 31
patients, 16 were male and 15 were female, ages 17 to 80 (mean 57) years. All 31 patients had diagnosed organic heart disease (Table I). Eleven (36%) of
de pointes
due to quinidine
Heart
1984 Journal
in 1 of 31 patients.
these patients had ischemic heart disease. Five patients (16 % ) had rheumatic heart disease (four patients had mitral stenosis and one patient had both mitral stenosis and aortic stenosis). Four patients (13 % ) had hypertensive heart disease, four patients (13 % ) had congestive cardiomyopathy of unknown etiology, three patients (10%) had sick sinus syndrome, two patients (6 % ) had mitral valve prolapse, and two patients (6%) had atrial septal defects. Quinidine therapy was initiated because of (paroxysmal or sustained) atria1 fibrillation or flutter in 22 (71% ) of the 31 patients. Quinidine was administered to six patients (19%) because of ventricular premature beats, to two patients (6%) because of recurrent sustained ventricular tachycardia, and to one patient (3 % ) because of paroxysmal supraventricular tachycardia associated with the Wolff-Parkinson-white syndrome. The dosages of quinidine administered to the 31 patients ranged from 650 to 2400 mg/day (mean 1097 mg/day). Serum quinidine levels drawn within 1 hour of torsade de pointes were available in 14 of the 31 patients and ranged from 1.4 to 10.6 pg/ml (mean 3.7 pg/ml). Only 2 of these 14 patients had quinidine levels above the usual therapeutic range (2 to 6 pg/ml). The patients had received from one dose of quinidine to approximately 3 years of quinidine therapy prior the occurrence of torsade de pointes. The duration of quinidine therapy prior to this arrhythmia was less than 1 week in 23 patients (74 % ) and less than 3 months in 28 patients (90% ). Twenty-eight (90%) of the 31 patients were receiving digoxin concomitantly with quinidine at the time of torsade de pointes. Serum digoxin levels drawn within 12 hours of this arrhythmia were available in 20 patients and ranged from 0.2 to 1.9 rig/ml (mean 1.2 rig/ml). Thus, none of these 20 patients had digoxin levels above the usual therapeutic range (up to 2.0 rig/ml). Three patients were receiving either procainamide (two patients) or disopyramide (one patient) in combination with quinidine. Other medications which were being taken at the time of torsade de pointes included natriuretic diuretics (13 patients), propranolol (seven patients),
Volume Number
Table
107 3
Torsade
I. Organic Variety
disease
Ischemic Rheumatic Hypertensive Cardiomyopathic Sick sinus syndrome Mitral valve prolapse Congenital
due to quinidine
427
880~
heart disease
of heart
de pointes
No.
of patients
(%)
11 (36%) 5 (16%) 4 (13%) 4 (13%) 3 (10%) 2 (6%) 2 (6%)
methyldopa (four patients), and isosorbide dinitrate (four patients). In addition, one patient was receiving a phenothiazine (prochloroperazine). Serum potassium concentrations drawn within 24 hours of torsade de pointes were available in 21 patients. These concentrations ranged from 3.2 to 4.9 mEq/L (mean 3.9 mEq/L) and were low (less than 3.5 mEq/L) in 5 (24%) of these 21 patients. Serum calcium concentrations drawn within 24 hours of torsade de pointes were available in nine patients and ranged from 8.2 to 9.6 mg/dl (mean 8.9 mg/dl). Serum magnesium concentrations were available in only two patients and were 1.5 and 1.9 mEq/L, respectively. Analysis of QT intervals. We analyzed 12-lead ECGs, which were taken when patients were not receiving quinidine or other type I antiarrhythmic drugs. Appropriate ECGs were available for 24 of the 31 patients. Nineteen of these ECGs were taken within 7 days prior to the initiation of quinidine therapy, whereas five were taken at least 14 days after quinidine therapy was discontinued. On these ECGs, 18 patients had sinus rhythm and six patients had atria1 fibrillation. Heart rates (ventricular rates) ranged from 42 to 171 bpm (mean 71 bpm). Heart rates were less than 60 bpm in 4 (17%) of the 24 patients. QT intervals ranged from 300 to 570 msec (mean 430 msec). Utilizing Bazett’sg method, QT, intervals in these 24 patients ranged from 390 to 580 msec (mean 470 msec) and were prolonged in 17 patients (71%) (Fig. 2). By means of the method of Simonson et al.,‘O measured QT intervals were prolonged in 16 (67%) of the 24 patients. Thus, the majority of patients had long QT intervals when they were off of quinidine therapy. Prominent U waves were not present on any of the 24 ECGs which were taken off of quinidine therapy. We also analyzed 12-lead ECGs, which were taken during quinidine therapy (within 48 hours of torsade de pointes). Appropriate ECGs were available for 23 of the 31 patients. On these ECGs, all 23 patients had sinus rhythm. Heart rates ranged from 44 to 88
0
MO8005803 i
520-
f 4800 k 400380320Fig. 2. QT
intervals corrected for rate (QTJ. Off of quinidine therapy (left), QT, intervals in 24 patients ranged from 390 to 580 (mean 470) msec and were prolonged in 17 patients (71%). On quinidine therapy (right), QT, intervals in 23 patients ranged from 390 to 630 (mean 490) msec and were prolonged in 21 patients (91% 1.
bpm (mean 65 bpm). Heart rates were less than 60 bpm in 8 (35%) of the 23 patients. QT intervals ranged from 390 to 630 msec (mean 490 msec). Utilizing Bazett’s formula, QT, intervals in these 23 patients ranged from 390 to 630 msec (mean 510 msec) and were prolonged in 21 patients (91% ) (Fig. 2). By using the method of Simonson et al., again QT intervals were prolonged in 21 (91%) of the 23 patients. Thus, a large majority of patients had long QT intervals on quinidine therapy. In addition to long QT intervals, we noted that 12 of the 23 patients had prominent U waves on quinidine therapy. All 23 patients had either long QT intervals or prominent U waves on quinidine therapy. Observations
regarding
onset of torsade
de pointes.
ECG rhythm strips revealed torsade de pointes with cycle lengths ranging from 220 to 360 msec (mean 270 msec). In 11 of the 31 patients, these rhythm strips revealed the events preceding the onset of torsades. Prior to the onset of torsade de pointes, 10 of these 11 patients had sinus rhythm (cycle lengths ranging from 600 to 1180 msec, with a mean of 910 msec) and ventricular premature beats. In each of these 10 patients a similar “long-short” ventricular
March,
428
Bauman
et al.
American
&us
short cycle mean= 570
rhyttwn
meanu=aiotn0ec c
Heart
1984
Journal
mwc
\
,
\ ion@cycle
toraades da pointes mean CL= 270 msee
mean=1170msec
Fig. 3. Drawing of electrocardiographic rhythm strip, depicting long-short ventricular sequence which preceded onset of torsade de pointes in 10 of 11 patients. All of these 10 patients had sinus rhythm (mean cycle length 910 msec). Ventricular premature beat(*) was followed by compensatory pause (mean R-R interval of these long cycles was 1170 msec). Following a conducted beat there was another ventricular premature beat (mean R-R interval of these short cycles was 570 msec), which was followed by torsade de pointes (mean cycle length 270 msec).
sequence immediately preceded torsade de pointes: a ventricular premature beat was followed by a compensatory pause (R-R intervals of these long cycles ranged from 920 to 1740 msec, with a mean of 1170 msec), then following a conducted beat there was another ventricular premature beat (R-R intervals of these short cycles ranged from 510 to 680 msec, with a mean of 570 msec) (Fig. 3). Prior to the onset of torsade de pointes, one patient had monomorphic ventricular tachycardia (cycle length 420 msec) which degenerated to torsade de pointes. Follow-up. The 31 patients were managed in many different ways by their various physicians. Therefore, we found it impossible to derive much useful information regarding therapy of torsade de pointes. Quinidine therapy was discontinued after the first episode in almost all patients. Nevertheless, 18 (58%) of the 31 patients had one or more recurrences of torsade de pointes (one patient had more than 400 episodes). Seventeen (55%) of the 31 patients had episodes which were always self-terminating, whereas 14 patients (45%) required one or more direct-current cardioversions. One patient died as a result of torsade de pointes due to quinidine. Thirty patients were discharged from the hospital. Two of these patients were later rechallenged with quinidine, and both had nonfatal recurrences of torsade de pointes. Seven other patients were administered other type I ant&rhythmic agents, and four of these patients (two taking disopyramide and two taking procainamide) had nonfatal recurrences of torsade de pointes. Of the 30 patients discharged from the hospital, one patient died nonsuddenly of unknown cause, whereas 29 patients are alive without recurrences of torsade de pointes (and syncope) after from 1 to 30 (mean 11.5) months.
DISCUSSION
Although many patients with arrhythmia benefit from quinidine therapy, quinidine also has the potential to cause a life-threatening ventricular arrhythmia, namely torsade de pointes. It would be useful to be able to identify patients as being at high risk or at low risk for having torsade de pointes due to quinidine. This kind of information could best be obtained by following a large series of patients started on quinidine therapy and comparing those who have torsade de pointes to those who do not. However, if only 4% of the patients receiving quinidine have this arrhythmia, it would be necessary to follow 500 patients receiving quinidine in order to characterize 20 patients with torsade de pointes. A study of this magnitude is probably not feasible. Selection process. In the present study, we used a mail solicitation to obtain the records of a large series of patients with torsade de pointes due to quinidine, in an attempt to characterize patients with this arrhythmia. Our method of study has limitations. For a patient with torsade de pointes due to quinidine to be included in our series, this arrhythmia had to be electrocardiographically documented, the cardiologist had to attribute its occurrence to quinidine therapy, and the records had to be mailed to us (or the patient referred to us). This selection process could have affected our results. Another limitation to our study is that we do not have the records of a series of patients without torsade de pointes during quinidine therapy with which to compare our series. Therefore, although our series of 31 patients with torsade de pointes due to quinidine is considerably larger than any previously published series, and enables us to characterize patients with this arrhythmia, we are often limited in our ability to reach conclusions regarding
Volume Number
107 3
identification of patients at high risk for having torsade de pointes due to quinidine. Torsade risk factors. We found that the patients with torsade de pointes due to quinidine were heterogenous with respect to age and sex. In contrast, all of the patients in our series had diagnosed organic heart disease. It has been suggested that the presence of heart disease could predispose to the occurrence of torsade de pointes.4 Our results are consistent with this hypothesis. In our series, the most common variety of heart disease was ischemic, whereas the majority of previously published cases have had rheumatic heart disease6 (diagnosed in only 16% of our patients). This discrepancy probably reflects the change in the relative prevalences of ischemic heart disease and rheumatic heart disease over the past several decades. In any case, our results imply that the presence of rheumatic heart disease is not a specific predisposition to torsade de pointes due to quinidine. Atrial fibrillation. Most (71%) of the patients in our series were receiving quinidine because of (paroxysmal or sustained) atrial fibrillation or flutter. This finding is consistent with previous reports6 It is not clear whether the presence of atrial fibrillation is a specific predisposition to torsade de pointes due to quinidine or whether this association merely reflects the population treated with quinidine. It is also possible that physicians are less likely to attribute torsade de pointes to quinidine therapy when this arrhythmia occurs in patients who are receiving quinidine in the treatment of other ventricular arrhythmias. Quinidine dosage. Most of the patients in our series were receiving moderate dosages of quinidine, and only 2 (14% ) of 14 patients had quinidine levels which were above the therapeutic range. Therefore, it seems clear that torsade de pointes due to quinidine usually does not reflect administration of “too much” quinidine, at least not by usual standards. Onset. In our series of patients torsade de pointes usually occurred within 1 week of the initiation of quinidine therapy. This finding supports the suggestion, recently made by Denes et al.,” that morbidity and mortality due to quinidine might be reduced by hospitalizing and electrocardiographically monitoring patients for several days when quinidine therapy is initiated. However, some of our patients had received quinidine for months or years prior to the first occurrence of torsade de pointes. There were no apparent differences between patients who had torsade de pointes soon after the initiation of quinidine therapy and patients who had this arrhythmia after receiving quinidine uneventfully for a long period of
Torsade de pointes due to quinidine
429
time. We do not know why some patients have torsades after receiving quinidine uneventfully for a long period of time. Concomitant digoxin. Most (90%) of the patients in our series were receiving digoxin concomitantly with quinidine. These patients were receiving moderate dosages of digoxin and had low or moderate serum digoxin levels. This finding, which is consistent with previously published data,6 suggests that torsade de pointes due to quinidine does not reflect digitalis toxicity. However, it is possible that concomitant administration of digitalis does increase the risk for torsade de pointes. Alternatively, the fact that most of the patients in our series were receiving digoxin may only reflect an association between torsade de pointes due to quinidine and either heart disease or atrial fibrillation. QT interval. Since quinidine is known to delay ventricular repolarization, it is not surprising that all the patients in our series had either long QT intervals or prominent U waves while on quinidine therapy. However, we were surprised to find that approximately 70% of these patients also had long QT intervals when they were not taking quinidine (i.e., either prior to initiation of quinidine therapy or after systemic elimination of quinidine). Previous reports have implied that patients with torsade de pointes due to quinidine usually have normal QT intervals when they are not taking quinidine.12-l4 However, in most of these reports QT intervals of patients off of quinidine have not been specified. In contrast, in one recent report15 of nine patients with torsade de pointes due to antiarrhythmic drugs (disopyramide, procainamide, or amiodarone), five (56 % ) of the patients had long QT, intervals prior to taking these drugs. Our finding that approximately 70% of patients with torsade de pointes due to quinidine have long QT intervals while off of quinidine therapy may be less striking than it appears to be. The normal ranges for QT intervals are derived from individuals without apparent heart disease. It is likely that QT intervals tend to be longer in patients with heart disease16 and that the majority of patients started on quinidine therapy have some variety of heart disease. However, we think that it is unlikely that anywhere near 70% of patients started on quinidine therapy have long QT intervals. Therefore, we think that long QT intervals may identify patients who are at increased risk for having torsades de pointes due to quinidine. Congenital QT syndromes. It is well known that it is dangerous to administer quinidine to patients with congenital QT syndromes, as described by Roman0
430
Bauman
et al.
et a1.17 and Ward18 and by Jervell and LangeNielson.lg The patients in our series differed from patients with these congenital syndromes in that our patients did not have torsade de pointes prior to treatment with quinidine and our patients had QT intervals which were usually only moderately prolonged (rather than grossly prolonged) off of quinidine. Nevertheless, it is possible that in some of the patients in our series the (slightly) prolonged QT intervals off of quinidine reflected mild forms of congenital QT syndromes. Alternatively, these long QT intervals may have been acquired, that is, due to heart disease, hypokalemia, slow ventricular rates, etc.16,20,21 In either case, since patients with moderately prolonged QT intervals appear to be at increased risk for having torsade de pointes, it would be desirable to avoid administering quinidine (and probably other type I antiarrhythmic drugs) to these patients. Long-short ventricular sequence. One other finding merits discussion. We were able to scrutinize the ECG events which preceded torsade de pointes in 11 patients. In 10 of these patients, we noted that a long-short ventricular sequence, caused by the occurrence of ventricular premature beats, immediately preceded the onset of torsades. Although this finding has not been previously emphasized, we have noted the same sequence of events in illustrations accompanying several previous reports of torsade de pointes due to quinidine.6* 7.11,‘Z IS.22The long ventricular cycles could increase dispersion of ventricular refractoriness and thereby potentiate the initiation of torsade de pointes.‘” If this were true, it could explain why shortening of ventricular cycle length, with ventricular pacing or administration of isoproterenol, can be effective therapy for torsade de pointes due to quinidine. 4.13.24,26Alternatively, it is conceivable that the ventricular premature beats are only a secondary event, that is, that a more basic derangement (transient imbalances in autonomic nervous tone?) causes increased dispersion of ventricular refractoriness, and that this increased dispersion first causes ventricular premature beats and then torsade de pointes. ADDENDUM
After the preparation of this report 1 of the 31 patients was inadvertently treated with disopyramide and died suddenly several days later. REFERENCES
I
Selzer A, Wray W: Quinidine syncope: Paroxysmal ventricular fibrillation occurring during treatment of chronic atria1 arrhythmias. Circulation 30:17, 1964.
American
March, 1994 Heart Journal
2. Krickler DM, Curry PVL: Torsades de pointes, an atypical ventricular tachycardia. Br Heart J 38:117, 1976. 3. Smith WM, Gallagher JJ: Les torsades de pointes: An unusual ventricular arrhythmia. Ann Intern Med 93:578, 1980. 4. Sclarovsky S, Strasberg B, Lewin RF, Agmon J: Polymorphous ventricular tachycardia: Clinical features and treatment. Am J Cardiol 44:339, 1979. 5. Rokseth R, Storstein 0: Quinidine therapy of chronic auricular fibrillation. Arch Int Med 111:102, 1963. 6. Reynolds EW, Vander Ark CR: Quinidine syncope and the delayed repolarization syndromes. Mod Concepts Cardiovasc Dis 45:117, 1976. 7. Ejvinsson G, Orinius E: Prodromal ventricular premature beats preceded by diastolic wave. Acta Med Stand 208:445, 1980. 8. Lepeschkin E, Surawicz B: The measurement of the QT interval of the electrocardiogram. Circulation 6:378, 1952. 9. Bazett HC: An analysis of the time-relations of the electrocardiogram. Heart 7:353, 1920. 10. Simonson E, Cady LD, Woodbury M: The normal QT interval. AM HEART J 63:747, 1962. 11. Denes P, Gabster A, Huang SK: Clinical, electrocardiographic and follow-up observations in patients having ventricular fibrillation during Holter monitoring: Role of quinidine therapy. Am J Cardiol 48:9, 1981. 12. Jenzer HR, Hagemeijer F: Quinidine syncope: Torsades de pointes with low quinidine plasma concentrations. Eur J Cardiol 4:447, 1976. 13. DiSegni E, Klein HO, David D, Libnaber C, Kaplinsky E: Overdrive pacing in quinidine syncope and other long QTinterval syndromes. Arch Intern Med 140:1036, 1980. 14. Kaplinsky E, Yahino JN, Barzilai J, Neufeld HN: Quinidine syncope: Report of a case successfully treated with lidocaine. Chest 62~764, 1972. 15. Keren A, Tzivoni D, Gavish D, Levi J, Gottlieb S, Benhorin J, Stern S: Etiology, warning signs and therapy of torsades de pointes: A study of 10 patients. Circulation 64:1167, 1981. 16. Schwartz PJ, Wolf S: QT-interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation 57:1074, 1978. 17. Roman0 C, Gemme G, Pongiglione R: Arithmie cardiache rare delleta pediatrica. II. Accessi sincopali per fibrillazione ventricular parossistica. Clin Pediatr 45:656, 1963. 18. Ward OC: New familial cardiac syndrome in children. J Irish Med Assoc 54:103. 1964. 19. Jervell A, Lange-Nielson F: Congenital deaf mutism, functional heart disease with prolongation of the QT-interval and sudden death. AM HEART J 54:59, 1957. 20. Redleaf PD, Lerner IJ: Thiazide-induced hypokalemia with associated major ventricular arrhythmias: Report of a case and comment on therapeutic use of bretylium. JAMA 206:1302, 1968. 21. Slama R, Motte G, Coumel Ph, Dessertenne F: Le syndrome “allongement de QT et syncopes par torsades de pointes. Lava1 Med 42:353, 1971. 22. Koster RW, Wellens HJJ: Quinidine-induced ventricular flutter and fibrillation without digitalis therapy. Am J Cardiol 38:519, 1976. 23. Han d, Goel BG: Electrophysiologic precursors of ventricular tachvarrhvthmias. Arch Int Med 129:749. 1972. 24. Khan MM, Logan KR, McComb JM, Adgey AAJ: Management of recurrent ventricular tachyarrhythmias associated with QT aroloneation. Am J Cardiol 47:1301. 1981. 25. Anderson JL, Mason JW: Successful treatment by overdrive pacing of recurrent quinidine syncope due to ventricular tachpcardia. Am J Med 64:715. 1978.
Jerry L. Bauman, Pharm.D., Robert A. Bauernfeind, M.D., Julie V. Hoff, Pharm.D., Boris Strasberg, M.D., Steven Swiryn, M.D., and Kenneth M. Rosen, M.D. Chicago,
Ill.
Syncope due to quinidine usually reflects the occurrence of a variety of polymorphic ventricular tachycardia known as torsade de pointes.‘-* It has been estimated that between 1% and 8% of patients receiving quinidine experience syncope.6-7 However, it has been postulated that quinidine syncope does not occur randomly and that some patients could be at increased risk. Reynolds and Vander Arks reviewed the relevant literature and concluded that quinidine syncope tends to occur in patients with atria1 fibrillation complicating rheumatic heart disease, especially when there is concomitant administration of digitalis. However, this conclusion was based upon small series of patients with quinidine syncope reported prior to 1970.
From the Department of Pharmacy Department of Medicine, University
Practice and the Section of Illinois.
of Cardiology,
Supported in part by National Heart, Lung, and Blood Institute Institutional Training Grant HL 07387, National Heart, Lung, and Blood Institute Research Grants HL 18794 and HL 23566, and a grant from the Eleanor B. Pillsbury Resident Trust Fund. Received
for publication
Reprint requests: Jerry sity of Illinois Hospital,
Nov.
18, 1982;
accepted
Dec.
12, 1982.
L. Bauman, Pharm.D., Cardiology Section, PO Box 6998, Chicago, IL 60680.
Univer-
In the present study, we used a mail solicitation to obtain the records of a large series of patients with torsade de pointes due to quinidine, in an attempt to characterize these patients. We also obtained ECGs, which were taken when these patients were on and off of quinidine therapy, and used these ECGs to analyze QT intervals. Our results are germane to identifying patients at increased risk for torsade de pointes due to quinidine. In addition, we were able to make observations regarding the onset of torsade de pointes in some of these patients. METHODS Patient selection. Criteria for inclusion in the present study were: (1) electrocardiographically documented torsade de pointes during quinidine therapy, (2) absence of torsade de pointes (and syncope) prior to quinidine therapy, (3) absence of torsade de pointes (and syncope) after the systemic elimination of quinidine, and (4) availability of sufficient medical and drug history to contribute to our analysis. In this study, we defined torsade de pointes as a rapid ventricular tachycardia with gradually changing QRS morphology (Fig. 1). We collected cases of torsade de pointes due to quinidine primarily by mail solicitation. We sent letters and questionnaires to 2052 cardiologists listed in the American College of Cardiology Directory. In the letter, we outlined
425
March,
426
Bauman
et al.
Fig.
American
1. Rhythm
strip documenting
torsade
the purpose of the study and listed the criteria for inclusion. For each suitable case, we requested copies of rhythm strips documenting torsade de pointes and completion of the questionnaire. Specific information requested in the questionnaire related to: (1) the medical history of the patient, including diagnosed organic heart disease and dysrhythmia, (2) details of quinidine therapy, namely dosage, duration of therapy, and serum levels, (3) details of concomitant drug therapy, (4) serum electrolytes at the time of torsade de pointes, and (5) therapy of torsade de pointes and subsequent course. Finally, we requested 12-lead ECGs taken prior to, during, and after quinidine therapy. In addition to the mail solicitation, we screened our files for the records of referred patients who satisfied the criteria for inclusion. QT interval analysis. Twelve-lead ECGs, taken both on and off of quinidine therapy, were used to analyze QT intervais. Patients with bundle branch block (QRS duration L 120 msec) were excluded from this analysis. QT intervals were measured from the beginning of the QRS complex to the end of the T wave. U waves were carefully excluded from these measurements by the methods of Lepeschkin and Surawicz8 In patients with sinus rhythm, the standard ECG lead displaying the longest QT interval was used to determine corrected QT (QT,) interval. In patients with atria1 fibrillation, mean QT intervals and cycle lengths were measured in the ECG lead displaying the most easily discernible T waves.8 QT, interval was calculated by the method of Bazett,a with the use of the following equation: QT, = measured QT/m The upper limit of normal for QTe, as determined by this method, was considered to be 440 msec for both men and women. QT intervals were also evaluated by the method of Simonson et al.,‘” with the use of the following equation: Calculated QT = 0.2423 + 0.140 (R-R) + 0.0003 (age) t 0.0164 (SEE). By this method, the measured QT interval was considered to be prolonged when it was greater than the calculated QT interval plus two standard errors of the estimate.‘O RESULTS
A total of 31 cases fulfilled the criteria for inclusion. Of these 31 cases, 21 were collected by the mail solicitation and 10 were retrieved from our files. Patient
characteristics
and drug
history.
Of the 31
patients, 16 were male and 15 were female, ages 17 to 80 (mean 57) years. All 31 patients had diagnosed organic heart disease (Table I). Eleven (36%) of
de pointes
due to quinidine
Heart
1984 Journal
in 1 of 31 patients.
these patients had ischemic heart disease. Five patients (16 % ) had rheumatic heart disease (four patients had mitral stenosis and one patient had both mitral stenosis and aortic stenosis). Four patients (13 % ) had hypertensive heart disease, four patients (13 % ) had congestive cardiomyopathy of unknown etiology, three patients (10%) had sick sinus syndrome, two patients (6 % ) had mitral valve prolapse, and two patients (6%) had atrial septal defects. Quinidine therapy was initiated because of (paroxysmal or sustained) atria1 fibrillation or flutter in 22 (71% ) of the 31 patients. Quinidine was administered to six patients (19%) because of ventricular premature beats, to two patients (6%) because of recurrent sustained ventricular tachycardia, and to one patient (3 % ) because of paroxysmal supraventricular tachycardia associated with the Wolff-Parkinson-white syndrome. The dosages of quinidine administered to the 31 patients ranged from 650 to 2400 mg/day (mean 1097 mg/day). Serum quinidine levels drawn within 1 hour of torsade de pointes were available in 14 of the 31 patients and ranged from 1.4 to 10.6 pg/ml (mean 3.7 pg/ml). Only 2 of these 14 patients had quinidine levels above the usual therapeutic range (2 to 6 pg/ml). The patients had received from one dose of quinidine to approximately 3 years of quinidine therapy prior the occurrence of torsade de pointes. The duration of quinidine therapy prior to this arrhythmia was less than 1 week in 23 patients (74 % ) and less than 3 months in 28 patients (90% ). Twenty-eight (90%) of the 31 patients were receiving digoxin concomitantly with quinidine at the time of torsade de pointes. Serum digoxin levels drawn within 12 hours of this arrhythmia were available in 20 patients and ranged from 0.2 to 1.9 rig/ml (mean 1.2 rig/ml). Thus, none of these 20 patients had digoxin levels above the usual therapeutic range (up to 2.0 rig/ml). Three patients were receiving either procainamide (two patients) or disopyramide (one patient) in combination with quinidine. Other medications which were being taken at the time of torsade de pointes included natriuretic diuretics (13 patients), propranolol (seven patients),
Volume Number
Table
107 3
Torsade
I. Organic Variety
disease
Ischemic Rheumatic Hypertensive Cardiomyopathic Sick sinus syndrome Mitral valve prolapse Congenital
due to quinidine
427
880~
heart disease
of heart
de pointes
No.
of patients
(%)
11 (36%) 5 (16%) 4 (13%) 4 (13%) 3 (10%) 2 (6%) 2 (6%)
methyldopa (four patients), and isosorbide dinitrate (four patients). In addition, one patient was receiving a phenothiazine (prochloroperazine). Serum potassium concentrations drawn within 24 hours of torsade de pointes were available in 21 patients. These concentrations ranged from 3.2 to 4.9 mEq/L (mean 3.9 mEq/L) and were low (less than 3.5 mEq/L) in 5 (24%) of these 21 patients. Serum calcium concentrations drawn within 24 hours of torsade de pointes were available in nine patients and ranged from 8.2 to 9.6 mg/dl (mean 8.9 mg/dl). Serum magnesium concentrations were available in only two patients and were 1.5 and 1.9 mEq/L, respectively. Analysis of QT intervals. We analyzed 12-lead ECGs, which were taken when patients were not receiving quinidine or other type I antiarrhythmic drugs. Appropriate ECGs were available for 24 of the 31 patients. Nineteen of these ECGs were taken within 7 days prior to the initiation of quinidine therapy, whereas five were taken at least 14 days after quinidine therapy was discontinued. On these ECGs, 18 patients had sinus rhythm and six patients had atria1 fibrillation. Heart rates (ventricular rates) ranged from 42 to 171 bpm (mean 71 bpm). Heart rates were less than 60 bpm in 4 (17%) of the 24 patients. QT intervals ranged from 300 to 570 msec (mean 430 msec). Utilizing Bazett’sg method, QT, intervals in these 24 patients ranged from 390 to 580 msec (mean 470 msec) and were prolonged in 17 patients (71%) (Fig. 2). By means of the method of Simonson et al.,‘O measured QT intervals were prolonged in 16 (67%) of the 24 patients. Thus, the majority of patients had long QT intervals when they were off of quinidine therapy. Prominent U waves were not present on any of the 24 ECGs which were taken off of quinidine therapy. We also analyzed 12-lead ECGs, which were taken during quinidine therapy (within 48 hours of torsade de pointes). Appropriate ECGs were available for 23 of the 31 patients. On these ECGs, all 23 patients had sinus rhythm. Heart rates ranged from 44 to 88
0
MO8005803 i
520-
f 4800 k 400380320Fig. 2. QT
intervals corrected for rate (QTJ. Off of quinidine therapy (left), QT, intervals in 24 patients ranged from 390 to 580 (mean 470) msec and were prolonged in 17 patients (71%). On quinidine therapy (right), QT, intervals in 23 patients ranged from 390 to 630 (mean 490) msec and were prolonged in 21 patients (91% 1.
bpm (mean 65 bpm). Heart rates were less than 60 bpm in 8 (35%) of the 23 patients. QT intervals ranged from 390 to 630 msec (mean 490 msec). Utilizing Bazett’s formula, QT, intervals in these 23 patients ranged from 390 to 630 msec (mean 510 msec) and were prolonged in 21 patients (91% ) (Fig. 2). By using the method of Simonson et al., again QT intervals were prolonged in 21 (91%) of the 23 patients. Thus, a large majority of patients had long QT intervals on quinidine therapy. In addition to long QT intervals, we noted that 12 of the 23 patients had prominent U waves on quinidine therapy. All 23 patients had either long QT intervals or prominent U waves on quinidine therapy. Observations
regarding
onset of torsade
de pointes.
ECG rhythm strips revealed torsade de pointes with cycle lengths ranging from 220 to 360 msec (mean 270 msec). In 11 of the 31 patients, these rhythm strips revealed the events preceding the onset of torsades. Prior to the onset of torsade de pointes, 10 of these 11 patients had sinus rhythm (cycle lengths ranging from 600 to 1180 msec, with a mean of 910 msec) and ventricular premature beats. In each of these 10 patients a similar “long-short” ventricular
March,
428
Bauman
et al.
American
&us
short cycle mean= 570
rhyttwn
meanu=aiotn0ec c
Heart
1984
Journal
mwc
\
,
\ ion@cycle
toraades da pointes mean CL= 270 msee
mean=1170msec
Fig. 3. Drawing of electrocardiographic rhythm strip, depicting long-short ventricular sequence which preceded onset of torsade de pointes in 10 of 11 patients. All of these 10 patients had sinus rhythm (mean cycle length 910 msec). Ventricular premature beat(*) was followed by compensatory pause (mean R-R interval of these long cycles was 1170 msec). Following a conducted beat there was another ventricular premature beat (mean R-R interval of these short cycles was 570 msec), which was followed by torsade de pointes (mean cycle length 270 msec).
sequence immediately preceded torsade de pointes: a ventricular premature beat was followed by a compensatory pause (R-R intervals of these long cycles ranged from 920 to 1740 msec, with a mean of 1170 msec), then following a conducted beat there was another ventricular premature beat (R-R intervals of these short cycles ranged from 510 to 680 msec, with a mean of 570 msec) (Fig. 3). Prior to the onset of torsade de pointes, one patient had monomorphic ventricular tachycardia (cycle length 420 msec) which degenerated to torsade de pointes. Follow-up. The 31 patients were managed in many different ways by their various physicians. Therefore, we found it impossible to derive much useful information regarding therapy of torsade de pointes. Quinidine therapy was discontinued after the first episode in almost all patients. Nevertheless, 18 (58%) of the 31 patients had one or more recurrences of torsade de pointes (one patient had more than 400 episodes). Seventeen (55%) of the 31 patients had episodes which were always self-terminating, whereas 14 patients (45%) required one or more direct-current cardioversions. One patient died as a result of torsade de pointes due to quinidine. Thirty patients were discharged from the hospital. Two of these patients were later rechallenged with quinidine, and both had nonfatal recurrences of torsade de pointes. Seven other patients were administered other type I ant&rhythmic agents, and four of these patients (two taking disopyramide and two taking procainamide) had nonfatal recurrences of torsade de pointes. Of the 30 patients discharged from the hospital, one patient died nonsuddenly of unknown cause, whereas 29 patients are alive without recurrences of torsade de pointes (and syncope) after from 1 to 30 (mean 11.5) months.
DISCUSSION
Although many patients with arrhythmia benefit from quinidine therapy, quinidine also has the potential to cause a life-threatening ventricular arrhythmia, namely torsade de pointes. It would be useful to be able to identify patients as being at high risk or at low risk for having torsade de pointes due to quinidine. This kind of information could best be obtained by following a large series of patients started on quinidine therapy and comparing those who have torsade de pointes to those who do not. However, if only 4% of the patients receiving quinidine have this arrhythmia, it would be necessary to follow 500 patients receiving quinidine in order to characterize 20 patients with torsade de pointes. A study of this magnitude is probably not feasible. Selection process. In the present study, we used a mail solicitation to obtain the records of a large series of patients with torsade de pointes due to quinidine, in an attempt to characterize patients with this arrhythmia. Our method of study has limitations. For a patient with torsade de pointes due to quinidine to be included in our series, this arrhythmia had to be electrocardiographically documented, the cardiologist had to attribute its occurrence to quinidine therapy, and the records had to be mailed to us (or the patient referred to us). This selection process could have affected our results. Another limitation to our study is that we do not have the records of a series of patients without torsade de pointes during quinidine therapy with which to compare our series. Therefore, although our series of 31 patients with torsade de pointes due to quinidine is considerably larger than any previously published series, and enables us to characterize patients with this arrhythmia, we are often limited in our ability to reach conclusions regarding
Volume Number
107 3
identification of patients at high risk for having torsade de pointes due to quinidine. Torsade risk factors. We found that the patients with torsade de pointes due to quinidine were heterogenous with respect to age and sex. In contrast, all of the patients in our series had diagnosed organic heart disease. It has been suggested that the presence of heart disease could predispose to the occurrence of torsade de pointes.4 Our results are consistent with this hypothesis. In our series, the most common variety of heart disease was ischemic, whereas the majority of previously published cases have had rheumatic heart disease6 (diagnosed in only 16% of our patients). This discrepancy probably reflects the change in the relative prevalences of ischemic heart disease and rheumatic heart disease over the past several decades. In any case, our results imply that the presence of rheumatic heart disease is not a specific predisposition to torsade de pointes due to quinidine. Atrial fibrillation. Most (71%) of the patients in our series were receiving quinidine because of (paroxysmal or sustained) atrial fibrillation or flutter. This finding is consistent with previous reports6 It is not clear whether the presence of atrial fibrillation is a specific predisposition to torsade de pointes due to quinidine or whether this association merely reflects the population treated with quinidine. It is also possible that physicians are less likely to attribute torsade de pointes to quinidine therapy when this arrhythmia occurs in patients who are receiving quinidine in the treatment of other ventricular arrhythmias. Quinidine dosage. Most of the patients in our series were receiving moderate dosages of quinidine, and only 2 (14% ) of 14 patients had quinidine levels which were above the therapeutic range. Therefore, it seems clear that torsade de pointes due to quinidine usually does not reflect administration of “too much” quinidine, at least not by usual standards. Onset. In our series of patients torsade de pointes usually occurred within 1 week of the initiation of quinidine therapy. This finding supports the suggestion, recently made by Denes et al.,” that morbidity and mortality due to quinidine might be reduced by hospitalizing and electrocardiographically monitoring patients for several days when quinidine therapy is initiated. However, some of our patients had received quinidine for months or years prior to the first occurrence of torsade de pointes. There were no apparent differences between patients who had torsade de pointes soon after the initiation of quinidine therapy and patients who had this arrhythmia after receiving quinidine uneventfully for a long period of
Torsade de pointes due to quinidine
429
time. We do not know why some patients have torsades after receiving quinidine uneventfully for a long period of time. Concomitant digoxin. Most (90%) of the patients in our series were receiving digoxin concomitantly with quinidine. These patients were receiving moderate dosages of digoxin and had low or moderate serum digoxin levels. This finding, which is consistent with previously published data,6 suggests that torsade de pointes due to quinidine does not reflect digitalis toxicity. However, it is possible that concomitant administration of digitalis does increase the risk for torsade de pointes. Alternatively, the fact that most of the patients in our series were receiving digoxin may only reflect an association between torsade de pointes due to quinidine and either heart disease or atrial fibrillation. QT interval. Since quinidine is known to delay ventricular repolarization, it is not surprising that all the patients in our series had either long QT intervals or prominent U waves while on quinidine therapy. However, we were surprised to find that approximately 70% of these patients also had long QT intervals when they were not taking quinidine (i.e., either prior to initiation of quinidine therapy or after systemic elimination of quinidine). Previous reports have implied that patients with torsade de pointes due to quinidine usually have normal QT intervals when they are not taking quinidine.12-l4 However, in most of these reports QT intervals of patients off of quinidine have not been specified. In contrast, in one recent report15 of nine patients with torsade de pointes due to antiarrhythmic drugs (disopyramide, procainamide, or amiodarone), five (56 % ) of the patients had long QT, intervals prior to taking these drugs. Our finding that approximately 70% of patients with torsade de pointes due to quinidine have long QT intervals while off of quinidine therapy may be less striking than it appears to be. The normal ranges for QT intervals are derived from individuals without apparent heart disease. It is likely that QT intervals tend to be longer in patients with heart disease16 and that the majority of patients started on quinidine therapy have some variety of heart disease. However, we think that it is unlikely that anywhere near 70% of patients started on quinidine therapy have long QT intervals. Therefore, we think that long QT intervals may identify patients who are at increased risk for having torsades de pointes due to quinidine. Congenital QT syndromes. It is well known that it is dangerous to administer quinidine to patients with congenital QT syndromes, as described by Roman0
430
Bauman
et al.
et a1.17 and Ward18 and by Jervell and LangeNielson.lg The patients in our series differed from patients with these congenital syndromes in that our patients did not have torsade de pointes prior to treatment with quinidine and our patients had QT intervals which were usually only moderately prolonged (rather than grossly prolonged) off of quinidine. Nevertheless, it is possible that in some of the patients in our series the (slightly) prolonged QT intervals off of quinidine reflected mild forms of congenital QT syndromes. Alternatively, these long QT intervals may have been acquired, that is, due to heart disease, hypokalemia, slow ventricular rates, etc.16,20,21 In either case, since patients with moderately prolonged QT intervals appear to be at increased risk for having torsade de pointes, it would be desirable to avoid administering quinidine (and probably other type I antiarrhythmic drugs) to these patients. Long-short ventricular sequence. One other finding merits discussion. We were able to scrutinize the ECG events which preceded torsade de pointes in 11 patients. In 10 of these patients, we noted that a long-short ventricular sequence, caused by the occurrence of ventricular premature beats, immediately preceded the onset of torsades. Although this finding has not been previously emphasized, we have noted the same sequence of events in illustrations accompanying several previous reports of torsade de pointes due to quinidine.6* 7.11,‘Z IS.22The long ventricular cycles could increase dispersion of ventricular refractoriness and thereby potentiate the initiation of torsade de pointes.‘” If this were true, it could explain why shortening of ventricular cycle length, with ventricular pacing or administration of isoproterenol, can be effective therapy for torsade de pointes due to quinidine. 4.13.24,26Alternatively, it is conceivable that the ventricular premature beats are only a secondary event, that is, that a more basic derangement (transient imbalances in autonomic nervous tone?) causes increased dispersion of ventricular refractoriness, and that this increased dispersion first causes ventricular premature beats and then torsade de pointes. ADDENDUM
After the preparation of this report 1 of the 31 patients was inadvertently treated with disopyramide and died suddenly several days later. REFERENCES
I
Selzer A, Wray W: Quinidine syncope: Paroxysmal ventricular fibrillation occurring during treatment of chronic atria1 arrhythmias. Circulation 30:17, 1964.
American
March, 1994 Heart Journal
2. Krickler DM, Curry PVL: Torsades de pointes, an atypical ventricular tachycardia. Br Heart J 38:117, 1976. 3. Smith WM, Gallagher JJ: Les torsades de pointes: An unusual ventricular arrhythmia. Ann Intern Med 93:578, 1980. 4. Sclarovsky S, Strasberg B, Lewin RF, Agmon J: Polymorphous ventricular tachycardia: Clinical features and treatment. Am J Cardiol 44:339, 1979. 5. Rokseth R, Storstein 0: Quinidine therapy of chronic auricular fibrillation. Arch Int Med 111:102, 1963. 6. Reynolds EW, Vander Ark CR: Quinidine syncope and the delayed repolarization syndromes. Mod Concepts Cardiovasc Dis 45:117, 1976. 7. Ejvinsson G, Orinius E: Prodromal ventricular premature beats preceded by diastolic wave. Acta Med Stand 208:445, 1980. 8. Lepeschkin E, Surawicz B: The measurement of the QT interval of the electrocardiogram. Circulation 6:378, 1952. 9. Bazett HC: An analysis of the time-relations of the electrocardiogram. Heart 7:353, 1920. 10. Simonson E, Cady LD, Woodbury M: The normal QT interval. AM HEART J 63:747, 1962. 11. Denes P, Gabster A, Huang SK: Clinical, electrocardiographic and follow-up observations in patients having ventricular fibrillation during Holter monitoring: Role of quinidine therapy. Am J Cardiol 48:9, 1981. 12. Jenzer HR, Hagemeijer F: Quinidine syncope: Torsades de pointes with low quinidine plasma concentrations. Eur J Cardiol 4:447, 1976. 13. DiSegni E, Klein HO, David D, Libnaber C, Kaplinsky E: Overdrive pacing in quinidine syncope and other long QTinterval syndromes. Arch Intern Med 140:1036, 1980. 14. Kaplinsky E, Yahino JN, Barzilai J, Neufeld HN: Quinidine syncope: Report of a case successfully treated with lidocaine. Chest 62~764, 1972. 15. Keren A, Tzivoni D, Gavish D, Levi J, Gottlieb S, Benhorin J, Stern S: Etiology, warning signs and therapy of torsades de pointes: A study of 10 patients. Circulation 64:1167, 1981. 16. Schwartz PJ, Wolf S: QT-interval prolongation as predictor of sudden death in patients with myocardial infarction. Circulation 57:1074, 1978. 17. Roman0 C, Gemme G, Pongiglione R: Arithmie cardiache rare delleta pediatrica. II. Accessi sincopali per fibrillazione ventricular parossistica. Clin Pediatr 45:656, 1963. 18. Ward OC: New familial cardiac syndrome in children. J Irish Med Assoc 54:103. 1964. 19. Jervell A, Lange-Nielson F: Congenital deaf mutism, functional heart disease with prolongation of the QT-interval and sudden death. AM HEART J 54:59, 1957. 20. Redleaf PD, Lerner IJ: Thiazide-induced hypokalemia with associated major ventricular arrhythmias: Report of a case and comment on therapeutic use of bretylium. JAMA 206:1302, 1968. 21. Slama R, Motte G, Coumel Ph, Dessertenne F: Le syndrome “allongement de QT et syncopes par torsades de pointes. Lava1 Med 42:353, 1971. 22. Koster RW, Wellens HJJ: Quinidine-induced ventricular flutter and fibrillation without digitalis therapy. Am J Cardiol 38:519, 1976. 23. Han d, Goel BG: Electrophysiologic precursors of ventricular tachvarrhvthmias. Arch Int Med 129:749. 1972. 24. Khan MM, Logan KR, McComb JM, Adgey AAJ: Management of recurrent ventricular tachyarrhythmias associated with QT aroloneation. Am J Cardiol 47:1301. 1981. 25. Anderson JL, Mason JW: Successful treatment by overdrive pacing of recurrent quinidine syncope due to ventricular tachpcardia. Am J Med 64:715. 1978.