The comparative effects of verapamil and a new dihydropyridine calcium channel blocker on digoxin pharmacokinetics

Transcription

1 The comparative effects of verapamil and a new dihydropyridine calcium channel blocker on digoxin pharmacokinetics Conflicting conclusions have been reported about interaction of calcium channel blockers with digoxin. The effects of verapamil (240 mg/day) and a new dihydropyridine calcium channel blocker, isradipine (15 mg/day), on the pharmacokinetics of 1 mg intravenous digoxin were compared. All 24 volunteer subjects were healthy, male, nonobese, and aged 18 to 38 years. Groups of 12 subjects received each oral agent over 15 days, with collections of blood and urine for 72 hours after intravenous digoxin. Significant (P < 0.05) reduction in nonrenal (7.01 ± 1.97 to 4.00 ± 1.86 L/hr) and total clearance (14.1 ± 2.6 to 11.5 ± 2.5 L/hr) were induced by verapamil, without change in renal clearance. A near-significant (P < 0.1) increase in peripheral volume of distribution contributed to prolonged elimination half-life (23.1 ± 4.4 to 34.3 ± 9.7 hours). By contrast, isradipine caused only a 9% reduction in volume of distribution. Verapatnil causes digoxin accumulation by reducing nonrenal elimination. No evidence of clinically relevant interaction of isradipine with digoxin was seen. (CLIN PHARMACOL THER 1987;42:66-7.) Brian F. Johnson, M.D., John Wilson, Ph.D., Raj Marwaha, M.S., Kathleen Hoch, B.S., and Johanna Johnson, B.S. Worcester, Mass. Several studies have confirmed that verapamil substantially increases serum digoxin concentration in patients continuously taking digoxin.1-5 Some evidence has been presented that verapamil treatment reduces both renal (CLR) and nonrenal clearance (CL,R) of digoxin.4 It has been concluded that reduced dosage of digoxin should be recommended to patients requiring concurrent verapamil treatment. The effects of other calcium channel blocking agents on digoxin pharmacokinetics are less established. Diltiazem appears to increase steady-state digoxin levels to a lesser extent than does verapamil, possibly by reducing CLN, of digoxin.''' The effects of nifedipine are highly controversial. Two studies"8 showed an increase in steady-state serum digoxin levels. However, reduced digoxin CLR, increased digoxin CLN, and total absence of any effect have been reported in the various studies that failed to confirm change in steady-state digoxin levels.'" From the Division of Clinical Pharmacology, University of Massachusetts Medical Center. Supported by a grant from Sandoz, Inc. Received for publication Sept. 21, 1986; accepted Dec. 22, Reprint requests: Brian F. Johnson, M.D., Director, Division of Clinical Pharmacology, University of Massachusetts Medical Center, Worcester, MA H3COOC H3C 3,5-Pyridinedicarboxylic acid, 4-(4-benzofurazanyI)-1, 4-dihydro-2, 6-dimethyl-, methyl 1-methyl-ethyl ester Chemical structure of isradipine Isradipine* (see structure) is a dihydropyridine calcium channel blocker and an analog of nifedipine. In animal studies it shares many of the characteristics of nifedipine, and it has been shown to have pharmacologic effects in patients taking doses of 2.5 mg or more. Although many clinical studies have been performed, relatively few have yet been published.12"3 The drug has been shown to be highly specific for vascular smooth muscle, with little risk of cardiac depression. Improved cardiac index and stroke volume index have *Proposed proprietary name DynaCirc (Sandoz); original code number PN

2 VOLUME 42 NUMBER 1 Isradtpine and digoxin 67 Table I. Mean ( ± SD) serum digoxin levels Time Control Verapamil Control Isradipine 10 min ± ± ± ± min ± ± ± ± min ± ± ± ± hr 7.71 ± ± ± ± hr 5.42 ± ± ± ± hr ± ± ± ± hr 2.32 ± ± ± ± hr 1.78 ± ± ± ± hr 1.55 ± ± ± ± hr 1.46 ± ± ± ± hr 1.30 ± ± ± ± hr 1.21 ± ± ± ± hr 1.00 ± ± ± ± hr 0.81 ± ± ± ± hr 0.64 ± ± ± ± hr 0.45 ± ± ± ± hr 0.34 ± ± ± ± hr 0.23 ± ± ± ± 0.12 been reported in association with reduced pulmonary artery wedge pressure and systemic vascular resistance in patients with congestive heart failure.'2 Doses of 2.5 to 10 mg have been reported to improve exercise capacity in patients with angina pectoris, without important adverse effects.'3 Antihypertensive efficacy has been demonstrated in several double-blind studies performed in several hundred patients. Single oral doses of between 2.5 and 20 mg produced dose-related reductions in blood pressure within 2 to 3 hours. Early experience suggests that a daily dosage of 15 mg should be a suitable regimen, and it was decided to study the effects of this dosage of isradipine on the pharmacokinetics of intravenously administered digoxin. As a positive control, the effects of isradipine were compared with those of verapamil in a parallel group of subjects. METHODS Our subjects were 24 healthy male volunteers aged 18 to 38 years (mean 29.7 years). Two subjects were black and the other 22 were white. Body weight ranged from 61.2 to 93.0 kg (mean 75.3 kg). No subject had clinically abnormal findings on history and physical examination, ECG, complete blood count, multichannel evaluation, urinalysis, or urinary screen for drugs. None had a history of hepatitis within the past 3 years, had donated blood within the past month, or had received any course of other drugs within the past 3 months. Subjects were randomly assigned to an open-label, analyst-blind, parallel group comparison of verapamil and isradipine. On the first day of the study, 1.0 mg digoxin was given over 15 minutes as an intravenous infusion in 20 ml of normal saline solution. Food was withheld for 10 hours before and 2 hours after the infusion. As a safety precaution, the ECG was monitored by oscilloscope during the infusion and for the next 15 minutes. Venous blood samples were collected before and at 10, 20, and 40 minutes and 1, 1.5, 2, 3, 4, 5, 6, 7, 8, 10, 24, 30, 48, 54, and 72 hours after completion of the infusion. Urine was collected in two 24- hour periods beginning from the time of completion of the infusion. After the last blood sample was collected (72 hours), subjects began to take the calcium channel blocker to which they had been randomized. Half of the subjects received oral verapamil, 80 mg every 12 hours for 4 days and 80 mg t.i.d. for the remaining 10 days of the study. The other 12 volunteers received oral isradipine, 2.5 mg every 12 hours for 2 days, 5 mg every 12 hours for 2 days, and 5 mg t.i.d. for the final 10 days. The second infusion of digoxin was administered to each subject 12 days after beginning the calcium channel blocker (i.e., 14 days after the first infusion). The infusion dose, period of infusion, ECG monitoring, and blood and urine collections were identical to those of the first infusion period. Oral calcium channel blocker therapy was continued during the collection period. Venous blood samples were allowed to clot and centrifuged immediately, and the serum was separated and

3 68 Johnson et al. CL1N PHARMACOL THER JULY 1987 Table II. Mean ( ± SD) urinary digoxin excretion Control Verapamil Control Isradipine (jig) (gg) (gg) 0.24 hr ± ± ± ± hr 91.7 ± ± ± hr ± ± ± ± 62.3 Table III. Effects of calcium channel blockers on mean ( ± SD) digoxin pharmacokinetic parameters Control Verapamil Control Isradipine CL (L/hr) ± ± 2.52*t ± CLR (L/hr) 7.04 ± ± ± ± 1.97 CLNR (L/hr) ± 1.74*t ± 2.04 Mean residence time (hr) 27.8 ± ± 12.7*t 39.5 ± ± 9.1 (L) ± ' ± * V, (L) 44.6 ± ± ± 5.6 A, (ng/ml) ± ± ± ± 2.57 A2 (ng/ml) 1.69 ± ± ± ± 0.24* X, (hr-') ± ± ± ± 0.18 X2 (hr) ± ± 0.005*t ± ± X, half-life (hr) 0.52 ± ± ± 0.08 A2 half-life (hr) ± ± 9.50*t ± ± 6.43 AUC (ng hr/m1) ± ± 20.29*t ± ± K,, (h-l) ± *t ± ± K12 (h-') ± ± ± ± K2, (h-') ± ± 0.028*t ± ± X, half-life, distribution half-life (0.693/1s1); A2 half-life, elimination half-life (0.693/X2). *Indicates P < 0.05 by comparison with baseline period. TInclicates P < 0.05 for changes from baseline by comparison with isradipine-induced changes. frozen until analyzed. Urine samples were mixed thoroughly, the volumes were measured and recorded, and aliquots were removed and frozen until analyzed. The digoxin concentration in each serum sample was quantified in triplicate by RIA. Urinary digoxin concentration was measured by modification of the same RIA kit used for serum.' Urine samples were analyzed in triplicate after dilution with buffered bovine serum albumin solution so that the concentration fell in the midportion of the calibration curve. The concentration in unknown samples was calculated by multiplication of the determined value by the dilution factor of that sample. Pharmacokinetic analysis. Serum digoxin time profiles were fit to a conventional two-compartment model with elimination from the first compartment, using NONLIN84 (Statistical Consultants, Lexington, Ky.) with reciprocal concentration weighting. Compartmental analysis provided estimates of the volume of distribution of the central compartment (V0), volume of distribution at steady state (Vs1), zero time intercept of the distribution phase of the concentration-time curve (A,), zero time intercept of the elimination phase of the concentration-time curve (A,), the distribution rate constant (a) (Xi), the elimination rate constant (p) (k2), and intercompartmental rate constants: model-dependent rate constant for elimination from the central compartment (k,o), model-dependent rate constant for transfer of drug from the central compartment to the peripheral compartment (k12), and model-dependent rate constant for transfer of drug from the peripheral compartment to the central compartment (k21). Half-lives (t112) of the distribution and elimination portions of the concentration-time curve were calculated from the relationship, = 0.693/Xn. Total body clearance (CL) was calculated by dividing the intravenous dose by the AUC, calculated using the trapezoidal rule. V, was obtained by the procedure described by Benet and Galeazzi15: D,, AUMC AUC2

4 VOLUME 42 NUMBER 1 Isradipine and digoxin CC TIME(h) TIME (h) Fig. 1. Profile of mean serum digoxin concentrations against time before ( ) and after verapamil ( ). Fig. 2. Profile of mean serum digoxin concentrations'against time before ( ) and after isradipine ( ). where D. is the intravenous dose and AUMC is the area under the first moment of the serum concentrationtime curve calculated by the trapezoidal rule. Mean residence time was calculated as the ratio of AUMC and AUC. CL,, was calculated by dividing the amount of digoxin excreted in urine over 48 hours by the serum AUC for the same collection period. CI., was calculated as the difference between CL and CI, for each subject. Treatment-induced changes in each derived or calculated parameter were tested for significance by the t test for paired observations. The effects of the two treatments were compared by ANOVA. RESULTS One subject developed a generalized rash while taking verapamil and did not complete the study. Data from 23 subjects are presented. Table I compares mean ( ± SD) values of serum digoxin and Table II compares urinary digoxin elimination before and after each treatment. Mean serum digoxin concentration-time profiles for the verapamil and isradipine groups are presented in Figs. 1 and 2. Derived group mean digoxin pharmacokinetic parameters are listed in Table III. No change in creatinine clearance (CLcR) (calculated during 48-hour urinary collection periods) occurred in either treatment group. Subjects receiving verapamil displayed an 18% decrease in digoxin CL (P < 0.01), a 49% increase in mean residence time (P < 0.05), and a 49% increase in the elimination t112 (P < 0.05). Mean digoxin CLR was unchanged; hence reduced CL was estimated to be the result of a significant mean 42% reduction in CLNR. A variable and near-significant trend for increased Vs, was also seen (0.05 < P < 0.1), the mean value being 22% greater after verapamil. Additionally, significant differences were seen in X2, AUC, Km, and K, with near-significant decrease in A, (0.05 < P < 0.1). By contrast, those subjects receiving isradipine showed only a 9% reduction in Vs, (P < 0.05), increase in A2 (P < 0.05), and a trend to increased X2 (0.05 < P < 0.1). Verapamil and isradipine differed significantly from each other in their effects on CL and CI,R, mean residence time, elimination t112, X2, AUC, km, and k21 (all at P < 0.01) and on Vs, and A2 (both at P < 0.05). Subjects varied considerably in baseline parameters such as V, CLNR, and elimination t1,2 of digoxin. However, there was no demonstrable relationship between these baseline determinations and the observed changes during calcium channel blocker treatment. DISCUSSION Although there have been several studies of the effects of calcium channel blockers on digoxin pharmacokinetics, the reported conclusions have been inconsistent. This may reflect the fact that most studies have involved small groups of volunteers or patients and that many of the studies were poorly controlled. Conclusions from these studies are most convincing in regard to the effects of verapamil. At least three studies1-3 demonstrated increases of between 53% and 100% in group mean levels of serum digoxin after sufficiently prolonged treatment with digoxin to achieve steadystate conditions. However, two of these studies offer opposing evidence on the dose-dependence of the reported effect of verapamil. Further, Pedersen et al.'

5 70 Johnson et al. CLIN PHARMACOL THER JULY 1987 reported that the effect developed quickly but was almost entirely abolished after 6 weeks of continued administration of both digoxin and verapamil. Pedersen et have also studied the impact of verapamil on the kinetics of a single intravenous dose of digoxin, reporting a 35% reduction in CL with relatively greater reduction in CLNR than in CI,. Elimination t,12 increased, although peripheral volume of distribution was unaffected. Reports of changes in V, are conflicting and probably clinically irrelevant. The same group also noted progressively increasing urinary recovery of digoxin administered continuously over 6 weeks,' suggesting that verapamil continued to inhibit CI,TR, whereas CLR was restored to normal. Relatively little is known of the clinical importance of the verapamil-digoxin interaction. Pedersen et al.' suggested that the interaction with verapamil enhances the pharmacologic effects of digoxin. They found that digoxin-induced increase in intracellular sodium levels in circulating blood cells was enhanced by verapamil in eight healthy volunteers. However, they found no change in the density or binding affinity of ouabain receptors on lymphocytes after verapamil, and Somberg et al.' reported no increased rubidium uptake in canine myocardial cells after mean circulating levels of digoxin had doubled after verapamil administration. In one study in healthy volunteers a mean 46% increase in steady-state digoxin levels was apparently induced by 10 mg nifedipine taken three times a day, and this was associated with a mean 29% reduction in calculated CI-, of digoxin.' However, several other studies concluded that nifedipine induces no change in steadystate digoxin leve1,7'1" and in two of the studies urinary digoxin excretion was determined and also found to be unchanged. Further, the kinetics of digoxin after a single intravenous dose were reported' to be little affected by nifedipine. Neither CL nor volumes of distribution showed any important change, but reduced urinary recovery of digoxin suggested that nonrenal mechanisms for digoxin elimination might actually be increased by nifedipine. However, the most recent study' showed a modest dose-independent increase in mean steady-state digoxin level, with a trend to reduced CLNR. The results of our study support the principal conclusions of other investigators in regard to the verapamil-digoxin interaction. CL was significantly reduced and this would be expected to lead to increased steady-state digoxin levels in patients continuously taking both drugs. Digoxin clearance consists of metabolic and renal components, each representing about half of the CL. Changes in digoxin clearance induced by ver- apamil have been claimed to result from reduction in both clearances. However, in our single-dose studies, reduced elimination appeared to be entirely caused by reduced nonrenal mechanisms. Our data are most compatible with a decrease in intrinsic hepatic clearance. Whereas other investigators found no trend for change in any calculated volume of distribution, most of our subjects showed a sizeable increase in V and therefore in the volume of the peripheral compartment. Because increased binding in tissues is unlikely with the administration of another drug, this might be explained by an increase in the plasma free fraction of digoxin. Although the variability between subjects was sufficient to prevent levels of statistical significance being achieved, this trend contributed to the significant prolongation of mean elimination ti,, seen after verapamil. Although the persistence and clinical relevance of the verapamil-digoxin interaction remained unclear, there are at least potential risks that concurrent use of the two drugs will cause enhanced efficacy and toxicity of digoxin. It may be anticipated that isradipine might interact with digoxin in a manner similar to that of nifedipine, in view of the close chemical similarity. If it is accepted that nifedipine does not alter CL, in this regard isradipine appears similar. However, we found no evidence of divergent effects on CI-, and CL,R, as Pedersen et al.' have previously reported. Further, unlike these authors' findings with nifedipine, we found that isradipine caused a small but significant fall in the volume of distribution of digoxin that was apparently in the peripheral compartment. This change may represent a decrease in tissue binding since change in clearance was not evident. It should be emphasized that the mean change was small and that even if larger it would have been of doubtful clinical importance. Changes in volume of distribution have no effect on steady-state drug levels. It is possible that isradipine will slightly increase peak serum digoxin levels (e.g., between 1 and 2 hours after digoxin tablets are ingested). We have previously provided evidence that the risk of cardiac toxicity is related only to steady-state digoxin levels (i.e., serum levels at a time when blood and tissue concentrations are in equilibrium)." Therefore we believe it is highly unlikely that coadministration of isradipine will have any clinically relevant effect on the cardiac actions of digoxin. It appears that no special precautions are indicated when the two drugs are given together. The assistance of Ronald Rossetti, Michaela Painchaud, and Annemarie Brahm is gratefully acknowledged.

6 VOLUME 42 NUMBER I Isradtpine and digoxin 71 References Belz GG, Doering W, Munkes R, Matthews J. Interaction between digoxin and calcium antagonists and antiarrhythmic drugs. CLIN PHARMACOL THER 1983;33: Doering W. Effect of coadministration of verapamil and quinidine on serum digoxin concentration. Eur J Clin Pharmacol 1983;25: Klein HO, Lang R, Weiss E, et al. The influence of verapamil on serum digoxin concentration. Circulation 1982;65: Pedersen KE, Dorph-Pedersen A, Hvidt S, Klitgaard NA, Nielsen-Kudsk F. Digoxin-verapamil interaction. CLIN PHARMACOL THER 1981;30: Pedersen KE, Dorph-Pedersen A, Hvidt S, Klitgaard HA, Pedersen KK. The long-term effect of verapamil on plasma digoxin concentration and renal digoxin clearance in healthy subjects. Eur J Clin Pharmacol 1982;22: Yoshida A, Fujita M, Kurosawa N, et al. Effects of diltiazem on plasma level and urinary excretion of digoxin in healthy subjects. CLIN PHARMACOL THER 1984;35: Kuhlmann J. Effects of nifedipine and diltiazem on plasma levels and renal excretion of beta-acetyldigoxin. CLIN PHARMACOL THER 1985;37: Kirch W, Hutt Hi, Dylewicz P, Graf JK, Ohnhaus EE. Dose-dependence of the nifedipine-digoxin interaction? CLIN PHARMACOL THER 1986;39:35-9. Pedersen KE, Dorph-Pedersen A, Hvidt S, Klitgaard NA, Kjaer K, Nielsen-Kudsk F. Effect of nifedipine on digoxin kinetics in healthy subjects. CLIN PHARMACOL THER 1982;32: Schwartz JB, Raizner A, Akers S. The effect of nifedipine on serum digoxin concentrations in patients. Am Heart J 1984;107; Schwartz JB, Migliore Pi. Effect of nifedipine on serum digoxin concentration and renal digoxin clearance. CLIN PHARMACOL THER 1984;36: Broudy D, Greenberg G, Siemienczuk D. Beneficial effects of the calcium antagonist PN in patients with congestive heart failure. J Am Coll Cardiol 1984; 3(suppl):479. Handler CE, Sowton E. Safety, tolerability and efficacy of PN , a new calcium antagonist in patients with angina and coronary heart disease. Eur J Clin Pharmacol 1984;27: Marwaha RK, Johnson BF. A radioimmunoassay for digoxin in human urine. J Clin Chem Clin Biochem 1984;22: Benet LZ, Galeazzi RI. Noncompartmental determination of the steady-state volume of distribution. J Pharm Sci 1979;68: Pedersen KE, Christiansen BD, Kjaer K, Klitgaard NA, Nielsen-Kudsk F. Verapamil-induced changes in digoxin kinetics and intraerythrocytic sodium concentration. CLIN PHARMACOL THER 1983;34:8-13. Somberg JC, Wellins H, Maguire W, Miura D. Verapamil-digitalis interaction: effect on cardiac Purkinje fibers and myocardium. [abstr]. Am J Cardiol 1982;49: Chapple DJ, Hughes R, Johnson BF. The relationship between cardiotoxicity and plasma digoxin concentration in conscious dogs. Br J Pharmacol 1976;57:23-7.