Br HeartJ7 1995;74:53-56 53 Concentration dependent cardiotoxicity of terodiline in patients treated for urinary incontinence Br Heart J: first published as 10.1136/hrt.74.1.53 on 1 July 1995. Downloaded from Simon H L Thomas, P Daniel Higham, Kenneth Hartigan-Go, Farhad Kamali, Peter Wood, Ronald W F Campbell, Gary A Ford

Abstract Terodiline hydrochloride is an antimuscarinic Objective-Terodiline, an antimuscarinic drug with calcium antagonist properties' and calcium antagonist drug, was used to which was used for treating urinary inconti- treat detrusor instability but was with- nence caused by detrusor instability.2 The drawn in 1991 after provoking serious drug was withdrawn in 1991 after reports of ventricular arrhythmias associated with cardiac dysrhythmia including bradycardia, increases in the corrected QT interval heart block, ventricular fibrillation, and ven- (QTc). This research was performed to tricular tachycardia, usually of the torsade de relate drug induced electrocardiographic pointes-type34 and associated with QT pro- changes in asymptomatic recipients to longation. Plasma terodiline concentrations plasma concentrations of the R( +) and were very high in one affected patient. S(-) terodiline enantiomers. Predisposing factors for terodiline associated Setting-Urological and geriatric clinics torsade de pointes were old age, coexisting and wards. ischaemic heart disease, co-prescription of Subjects-Asymptomatic patients taking other cardioactive drugs, and hypokalaemia. terodiline in stable dose. Torsade de pointes is associated with Methods-Electrocardiograms (50 mm/s) abnormal prolongation of the ventricular were collected from patients while they refractory period which results in a long QT were taking terodiline and compared with interval on the electrocardiogram and is often ECGs obtained before or after terodiline. drug induced.5 Increased dispersion of ven- QT interval, heart rate corrected QT tricular recovery may be important in the interval (QTc), and QT dispersion (QTd) development of this arrhythmia and indirect were measured. Drug induced electro- evidence indicates that this is reflected by http://heart.bmj.com/ cardiographic changes were related to increased dispersion of QT interval durations plasma concentrations of R(+) and S(-) across the standard 12 lead electrocardio- terodiline. gram.67 The relation between torsade de Results-During terodiline treatment pointes and plasma drug concentrations is mean QTc and QTd were prolonged uncertain: while torsade may result from over- (491(43) and 84 (35) ms1l2) compared with dose of several drugs,5 many patients develop- measurements made offtherapy (443 (33) ing drug induced torsade do not have and 42 (17) ms'12, paired t tests, P < 0-002 excessive plasma drug concentrations and on October 4, 2021 by guest. Protected copyright. and P < 0 01 respectively) in the 12 some consider the arrhythmia to be an idio- patients in sinus rhythm. The mean (95% synchratic phenomenon.8 confidence interval) drug induced The effects of terodiline were therefore increases were 48 (23 to 74) Ms112 for QTc investigated by collecting electrocardiograms Wolfson Dept of Clinical and 42 (13 to 70) ms'12 for QTd. These from asymptomatic patients receiving the Pharmacology, increases correlated with total plasma drug for treatment of urinary incontinence, University of terodiline (QTc: r = 0-77, P < 0-006, comparing them with those taken before or Newcastle, Newcastle QTd: r = P < and with after and the terodi- upon Tyne 0*68, 0.025) drug treatment, relating S H L Thomas plasma concentrations of both terodiline line induced electrocardiographic changes K Hartigan-Go enantiomers. with plasma concentrations of each terodiline F Kamali Conclusions-Terodiline increases enantiomer. P Wood QTc G A Ford and QTd in a concentration dependent Academic manner. It is not clear whether this is a Department of stereoselective effect and, if so, which Patients and methods Cardiology, enantiomer is responsible. The results Electrocardiograms (50 mm/s) and plasma University of Newcastle, Newcastle suggest that drug induced torsade de samples were collected from patients who were upon Tyne pointes is a type A (concentration depen- taking terodiline in stable dose between 4 and P D Higham dent) adverse drug reaction. 24 h after the last dose. This process started R W F Campbell in June 1991, when reports of dysrhythmia Correspondence to: first appeared, and ended in September 1991 Dr S H L Thomas, Wolfson (Br HeartJ7 1995;74:53-56) Department of Clinical when the drug was withdrawn. Copies of elec- Pharmacology, Claremont Place, Newcastle NE1 7RU. trocardiograms predating terodiline treatment Accepted for publication Keywords: terodiline; cardiotoxicity; electrocardio- were also collected, when available, and, if 11 January 1995 graphy possible, electrocardiography was repeated at 54 Thomas, Higham, Hartigan-Go, Kamali, Wood, Campbell, Ford

Table 1 Characteristics ofpatients in sinus rhythm FO ioo0 100- / taking terodiline 0 C 50 * . 50 * Mean (range) 0 Age (years) 73 (46-92) C'0 Height (cm) 161 (150-193) R= 0.78, P < 0.005 R 0.76, P<0.008 u -50 - -50- Weight (kg) 61 (44-90) 0 100 200 300 400 500 0 100 200 300 400 500 Daily dose (mg)* 38 (12-5-50) Br Heart J: first published as 10.1136/hrt.74.1.53 on 1 July 1995. Downloaded from Duration of treatment (days) 231 (8-1050) 150 150 Sodium (mmol/l) 139 (136-142) 100 100- Potassium (mmol/l) 4-1 (3 4-4 8) .C Creatinine (umol/l) 110 (89-116) D 50 . *. 50 0 Ionised calcium (mmol/l) 1-30 (1-25-1-32) Magnesium (mmolIl) 0-8 (0 7-0 9) CO0 01 Albumin (g/l) 41 (37-46) -5 00. R 0-69, P < 002 R 0.67, P < 0.03 0 100 200 300 400 500 0 100 200 300 400 500 Values are mean (range). *Median daily dose was 25 mg. R(+)-terodiline (ng/ml) S(-)-terodiline (ng/ml) Figure 2 Relation between plasma concentrations of terodiline enantiomers and the extent ofdrug induced least 2 months after terodiline had been dis- prolongation in corrected QT intervals (QTc) and QT continued. Each patient's medical details dispersion (QTd). were recorded and the blood sample analysed for plasma drug concentrations, electrolytes, and creatinine. Electrocardiograms were analysed blind by patients with atrial fibrillation were not used an independent observer. RR, PR, and QT as this condition has major effects on QTd. intervals were measured using a digitiser For the 15 patients in sinus rhythm (table 1) (Calcomp 9000, Scottsdale, AZ, USA). The electrocardiograms were available before QT interval was measured from onset of the terodiline was started in two patients and after QRS complex to the end of the T wave, terodiline had been stopped in 11, although defined as a return to the TP baseline. In the one of these had developed atrial fibrillation. presence of U waves, the end of the T wave Electrocardiograms in sinus rhythm were was taken at the TU nadir. Three representa- therefore available during and before or after tive complexes were analysed for those leads treatment in 12 patients, of whom 11 agreed of the standard 12 lead electrocardiogram to give a blood sample. Of these, one had where the T wave could be clearly defined right bundle branch block and four were taking and a mean taken. When the end of the T other drugs which might affect QT interval wave could not be reliably identified the lead (dothiepin, imipramine, thioridazine, sotalol), was excluded from analysis. The largest QT the doses of which were not altered during the interval on the 12 lead electrocardiogram study, and five had cardiovascular disease (QTmax) was used to derive the corrected (cerebrovascular disease (three), coronary QT interval (QTc), which was calculated for artery disease (one), and hypertension (one)). http://heart.bmj.com/ each lead using Bazett's formula (QTc = QT/ Electrocardiograms of two patients receiving V (RR interval)). QT dispersion (QTd) was terodiline were collected after only 8 and 16 determined as the longest minus the shortest days of treatment, possibly before steady state QTc on the 12 lead electrocardiogram. concentrations and the maximal electrocar- Plasma concentrations of R( + )- and diographic effects of the drug would be S(-)- terodiline enantiomers were measured achieved.

simultaneously by high performance liquid A total of 17 patients gave a blood sample. on October 4, 2021 by guest. Protected copyright. chromatography using a chiral column and an Plasma concentrations of the two terodiline ultraviolet detector. The limit of detection of enantiomers correlated closely (r = 0 99, the method is 25 ng/ml for each enantiomer P < 0 0001) with a mean (SD) R(+)-/S(-)- 600_ p < 0.002 and coefficients of variation are 5-5% for the ratio of 1 03 (0 14). No significant correla- ^550 S R(+)-enantiomer at 271 ng/ml and 7-6% for tions were observed between daily terodiline Ch dose and the plasma concentration of either 500 - the S(-)-enantiomer at 173 ng/ml. All { samples were analysed in duplicate and with- terodiline enantiomer. Apparent terodiline Q450 I 7 measured as dose divided C) 400 out knowledge of the electrocardiographic clearance, daily by - changes observed in each patient. plasma concentration, varied widely between 350 ' patients and did not correlate significantly P < 0.01 with age in this small group (data not shown). 150 Results Terodiline did not affect the PR interval or Nineteen patients taking terodiline took part heart rate but prolonged QTc and QTd 100 in this study. The electrocardiograms of four (fig 1, table 2) compared with measurements

50 0 On Off Table 2 Electrocardiographic effect of terodiline in 12 patients in sinus rhythmn Figure 1 Corrected QT Heart rate PR interval QTmax QTc QTd intervals (QTc) and QT (beatslmin) (ms) (ms) (msll2) (ms) dispersion (QTd) in 12 patients in sinus rhythm on Patients not receiving treatment 73 (16) 166 (22) 405 (49) 443 (33) 42 (17) Patients receiving treatment 72 (14) 178 (26) 456 (64) 491 (43) 84 (35) and offterodiline, together Mean difference -2 12 51 48 42 with means (SD). Four (95% CI) (-10 to 7) (-3 to 26) (17 to 85) (23 to 74) (13 to 70) patients taking other drugs Paired t test NS NS P < 0 01 P < 0-002 P < 0 01 that prolong QTc are shown as crosses. Values are mean (SD). Concentration dependent cardiotoxicity ofterodiline in patients treatedfor urinary incontinence 55

made when patients were not receiving treat- while receiving terodiline had marked prolon- ment. There was a positive correlation gation of the QTc interval,'3 4 20 however, daily between absolute values of QTc and QTd in terodiline doses were only a little higher patients receiving terodiline (r = 0 70, P < (mean 40 mg) than in our asymptomatic 0 005) and between terodiline induced group. Our results and those of a small open

increases in QTc and QTd (r = 0-88, P < study of the effects of terodiline in elderly Br Heart J: first published as 10.1136/hrt.74.1.53 on 1 July 1995. Downloaded from 0O001). Increases in QTmax, QTc, and QTd patients20 show that QTc prolongation is com- correlated significantly with the plasma con- mon in asymptomatic patients receiving the centrations of terodiline and each of its enan- drug, although this is less marked than in the tiomers (fig 2). patients with torsade de pointes. The observa- QTc was increased with terodiline in three tion that terodiline also increases the disper- patients whose electrocardiograms showed sion of ventricular recovery as measured by controlled atrial fibrillation on and off treat- the QTd has not previously been reported. ment. QTc was reduced with terodiline in one These effects correlate closely with plasma other patient, but this patient's atrial fibrilla- concentrations of the drug and its enan- tion was uncontrolled (rate 11 0/min) when tiomers confirming that this is a type A con- treatment was not given and the QTc is prob- centration related adverse drug reaction, ably invalid at this rate. although some patients, for example, those with heart disease or hypokalaemia, may be particularly susceptible. Consistent with this, Discussion plasma terodiline concentration was sub- The QT interval is an important factor in the stantially higher (2946 ng/ml) in the one genesis of ventricular arrhythmia. QTc pro- patient with torsade de pointes who had this longation in excess of 500 ms!"12 almost measured4 than in our asymptomatic patients. inevitably accompanies torsade de pointes.9 Why some patients develop high concentra- More modest prolongation is associated with tions of terodiline needs to be clarified. One an increased risk of sudden death in several risk factor is increasing age as elderly patients patient groups.'1-2 Drugs that prolong the QT metabolise terodiline more slowly, with a interval can cause ventricular arrhythmia and mean elimination half life of 130-190 h2' 22 sudden death, for example, antiarrhythmic compared with 63 h in younger patients.2' drugs, phenothiazines, and some antihista- Thus it may take several weeks for steady state mines.5 The risk of drug induced arrhythmia, plasma concentrations to be achieved and however, is not determined by the QT interval these may be higher than anticipated. In addi- alone; drugs with class III effects such as tion, there is considerable variability in the amiodarone increase QTc but are apparently terodiline elimination half life between elderly associated with a low risk of torsade."3 individuals.2' 22 Variations in clearance proba- evidence Furthermore, of the relation bly reflect differences in hepatic metabolism http://heart.bmj.com/ between the risk of sudden death and the QTc via p-hydroxyterodiline, the major route of interval in apparently healthy populations is elimination. This process may be affected by conflicting. 14 15 genetic polymorphism as clearance of R(+)- An important factor in the genesis of terodiline was reduced in a slow hydroxylator arrhythmia is the heterogeneity or dispersion ofofdebrisoquine.4~~~~~24 of the duration of ventricular repolarisation. The terodiline enantiomers have stereose- When dispersion is increased some areas of lective actions. The R( + )-enantiomer is a

the ventricle will be refractory at times when more potent antimuscarinic agent,2526 while on October 4, 2021 by guest. Protected copyright. others can allow unidirectional or bidirec- the S(-)-enantiomer has more marked cal- tional impulse propogation. This may estab- cium antagonist properties.25 In rats hydroxy- lish local re-entry circuits and ventricular lation of the two enantiomers occurs at tachycardia or fibrillation may then be trig- different rates, being more rapid for R(+)- gered by critically placed ventricular prema- terodiline.27 If this was also true in humans ture contractions.'6 Indirect evidence suggests then it would be expected that plasma con- that the interlead variability in QT interval centrations of the S(-) enantiomer would be durations obtained from the standard 12 lead higher at steady state. In our patients, how- electrocardiogram reflects increased disper- ever, plasma concentrations of each enan- sion of ventricular recovery and the risk of tiomer were similar, and it is therefore arrhythmia67 1718 and sudden death.'9 QT unlikely that differential clearance occurs in interval dispersion does not necessarily humans. This is consistent with a previous increase as the QT interval lengthens: for observation that clearance of R( + )-terodiline example, amiodarone prolongs QTc without is similar to that of the racemic drug.24 increasing QTd9 and this may explain its This study does not indicate which of the infrequent association with torsade. Early enantiomers is most responsible for the afterdepolarisations also seem to be important observed electrophysiological effects, nor can in the genesis of torsade and these might also it exclude the possibility that these are caused be reflected in the 12 lead electrocardiogram by a metabolite rather than the parent drug. as increased QTd.69 It is therefore important Increases in QTc and QTd correlated signifi- that QT interval and dispersion are measured cantly with both enantiomers, but the concen- when investigating the effects of drugs on ven- trations of the enantiomers are closely related tricular repolarisation. to each other. The most likely mechanism is Nine of the 10 patients described in the lit- potassium channel inhibition, a property of erature who developed torsade de pointes other QT prolonging drugs such as quinidine, 56 Thomas, Higham, Hartigan-Go, Kamali, Wood, Campbell, Ford

amiodarone, and terfenadine.i28 To date the 9 Hii JTY, Wyse DG, Gillis AM, Duff HJ, Solyo MA, Mitchell LB. Precordial QT interval dispersion as a effects of terodiline enantiomers on cardiac marker of torsade de pointes. Disparate effects of class potassium channels have not been studied. la antiarrhythmic drugs and amiodarone. Circulation 1992;86: 1376-82. Important lessons from the terodiline expe- 10 Laakso M, Aberg A, Savola J, Pentikalinen PJ, Pyorala K. rience should be learned. It is clear from this Diseases and drugs causing prolongation of the QT interval. AmJ Cardiol 1987;59:862-5. and other data that if an effect of the drug on 11 Algra A, Tijssen GP, Roeland RTC, Pool J, Lubsen J. Br Heart J: first published as 10.1136/hrt.74.1.53 on 1 July 1995. Downloaded from cardiac repolarisation had been sought during QTc prolongation measured by standard 12 lead electro- cardiography is an independent risk factor for sudden its development it would have been detected. death due to cardiac arrest. Circulation 1991;83: This would have limited the use of the drug in 1888-94. 12 Fioretti P, Tijssen GJP, Azar Lazzeroni E, et al. Prognostic high risk groups and resulted in closer moni- significance of predischarge 12 lead electrocardiogram toring of patients while receiving therapy. In after myocardial infarction compared with other routine clinical variables. Br Heartr 1 987;57:306-12. view of the close structural similarity to the 13 Mattioni TA, Heutlin TA, Sarmiento JJ, Parker M, Lesch proarrhythmic drug prenylamine,29 it is disap- M, Kehoe RF. Amiodarone in patients with previous drug-mediated torsade de pointes: long term safety and pointing that electrocardiographic effects were efficacy. Ann Intern Med 1989;111:574-80. not studied. Nevertheless it is reassuring that 14 Goldberg RJ, Bengtson J, Chen Z, Anderson KM, Locati E, Levy D. Duration of the QT interval and total cardio- voluntary organised adverse drug reaction vascular mortality in healthy persons (the Framingham reporting by prescribers in the United study experience). AmJ Cardiol 1991;67:55-8. 15 Schoeten EG, Dekker JM, Meppelink P, Kok FJ, Kingdom detected and characterised the Vandenbroucke JP, Pool J. QT interval prolongation problem. predicts cardiovascular mortality in an apparently healthy population. Circulation 199 1;84: 1516-23. Studies using isolated detrusor muscle 16 Han J, Millet D, Chizzonette B, et al. Temporal dispersion strips suggest that the beneficial effects of of recovery of excitability in atrium and ventricle as a function of heart rate. Am HeartJ 1966;76:481-90. terodiline on the bladder are primarily medi- 17 Day CP, McComb JM, Campbell RWF. QT dispersion in ated by the anticholinergic R( + )-enan- sinus beats and ventricular extrasystoles in normal hearts. Br HeartJ 1992;67:266-8. tiomer.26 If the adverse cardiovascular effects 18 Day CP, McComb JM, Mathews J, Campbell RWF. are caused by the S(-)-enantiomer then Reduction in QT dispersion by sotalol following myocardial infarction. EurHeartJr 1991;12:423-7. treatment with pure R( + )-terodiline might 19 Barr CS, Naas A, Freeman N, Lang CC, Struthers AD. have an acceptable risk-benefit ratio. Studies QT dispersion and sudden unexplained death in chronic heart failure. Lancet 1994;343:327-9. to characterise further the electrophysiological 20 Stewart DA, Taylor J, Ghosh S, et al. Terodiline causes effects of terodiline enantiomers in animals polymorphic ventricular tachycardia due to reduced heart rate and prolongation of QT interval. Eur J Clin and humans would be useful. Pharmacol 1992;42:577-80. 21 Hallen B, Magnusson A, Bogentoft S, Ekelund P. Single- and multiple-dose pharmacokinetics of terodiline in geri- We are grateful to Kabi Pharmacia for supplies of racemic atric patients. EurJ Clin Pharmacol 1988;34:291-7. terodiline and its enantiomers, and to members of the urology 22 Hallen B, Bogentoft S, Sandquist S, Stromberg S, department at the Freeman Hospital for their help with the Setterberg G, Ryd-Kjelen E. Tolerability and steady study. state pharmacokinetics of terodiline and its main metabolites in elderly patients with urinary incontinence. EurJ7 Clin Pharmacol 1989;36:487-93. 23 Karlen B, Andersson K-E, Ekman G, Stromberg S, Ulmsten U. Pharmacokinetics of terodiline in human 1 Andersson K-E. Clinical pharmacology of terodiline. volunteers. EurJ7 Clin Pharmacol 1982;23:267-70. http://heart.bmj.com/ ScandJ7 Urol Nephrol 1984;87(suppl): 13-20 24 Hallen B, Gabrielsson B, Palmer J, Ekstrom B. 2 Langtree HD, McTavish D. Terodiline: a review of its Pharmacokinetics of R-( + )-terodiline given intra- pharmacological properties and therapeutic use in the venously and orally to healthy volunteers. Pharmacol treatment of urinary incontinence. Drugs 1990;40: Toxicol 1993;73:153-8. 748-61. 25 Larsson-Backstrom C, Arrenius E, Sagge K. Comparison 3 McCloud AA, Thorogood S, Barnett S. Torsade de ofthe calcium-antagonistic effects ofterodiline, nifedipine pointes complicating treatment with terodiline. BMJ and verapamil. Acta Pharmacologica & Toxicologica 1985; 1991;302: 1469. 57:8-17. 4 Connolly MJ, Astrige PS, White EG, Morley CA, 26 Andersson K-E, Ekstrom B, Mattiasson A. Actions of Campbell Cowan J. Torsades de pointes ventricular terodiline, its isomers and main metabolite on isolated

tachycardia and terodiline. Lancet 1991;338:344-5. detrusor muscle from rabbit and man. Acta Pharmacol & on October 4, 2021 by guest. Protected copyright. 5 Thomas SHL. Drugs, QT interval prolongation and ven- Toxicol 1988-63:390-5. tricular arrhythmias. Adv Drug React Tox Rev 1994; 27 Lindeke B, Ericsson 0, Jonsson A, Noren B, Stromberg S, 13:77-102. Vangbo B. Biotransformations of terodiline. III. 6 Higham PD, Campbell RWF. QT dispersion. Br Heart Jf Opposed stereoselectivity in the benzylic and aromatic 1994;71:508-10. hydroxylations in rat liver microsomes. Xenobiotica 7 Day CP, McComb JM, Campbell RWF. QT dispersion: 1987;17: 1269-78. an indication of arrhythmia risk in patients with long QT 28 Woosley RL, Chen Y, Freiman JP, Gillis RA. Mechanism of intervals. Br HeartJ 1990;63:342-44. the cardiotoxic actions of terfenadine. _AMA 1993;269: 8 Zehender M, Hohnloser S, Just H. QT-interval prolonging 1532-6. drugs: mechanisms and clinical relevance of their 29 Puritz R, Henderson MA, Baker SN, Chamberlain DA. arrhythmogenic hazards. Cardiovascular Drugs & Ventricular arrhythmias caused by prenylamine. BM3 Therapy 1991;5:515-30. 1977;2:608-9.