A prospective study of the clarithromycin digoxin interaction in elderly patients

Transcription

1 Journal of Antimicrobial Chemotherapy (2002) 50, DOI: /jac/dkf176 A prospective study of the clarithromycin digoxin interaction in elderly patients P. Zapater 1 *, S. Reus 2, A. Tello 2, D. Torrús 2, M. Pérez-Mateo 2 and J. F. Horga 1 1 Department of Clinical Pharmacology, University General Hospital of Alicante, Maestro Alonso 109, Alicante; 2 Department of Internal Medicine, University General Hospital of Alicante, Alicante, Spain Received 9 January 2002; returned 14 May 2002; revised 1 July 2002; accepted 15 July 2002 The study was a prospective observational trial carried out to assess the clarithromycin digoxin interaction in elderly patients chronically taking digoxin. Digoxin concentrations were determined before and after concomitant treatment with clarithromycin. A Bayesian approach was used to calculate digoxin pharmacokinetics. In the seven patients who were studied there was a significant increase in digoxin concentration after 4 7 days of clarithromycin treatment; digoxin clearance and elimination rate constant were 56 60% lower and elimination half-life was 82% longer. The pharmacokinetic clarithromycin digoxin interaction in the elderly may be much more frequent than has been assumed up to now. Keywords: clarithromycin, digoxin, pharmacokinetic interaction, Bayesian Introduction Concurrent therapy with oral antibiotics and digoxin has been associated with changes in serum digoxin concentration; tetracycline and macrolides increase serum digoxin concentrations whereas rifampicin and neomycin have the opposite effect. 1 Several mechanisms have been proposed to explain this interaction. There is evidence that the formation of dihydro (reduced) metabolites (DRM) is accomplished by intestinal bacteria in humans. It has been estimated that this metabolic pathway for digoxin represents a substantial fraction of the total digoxin metabolism in 10% of the general population. 2 One hypothesis is that in these patients macrolides kill off most of the intestinal bacteria, especially Eubacterium lentum, a microorganism that is capable of metabolizing digoxin to inactive DRM. 3 A second mechanism that is possibly involved in this interaction is inhibition of P-glycoprotein. Macrolides inhibit P-glycoprotein in the intestine and kidney, with a consequent net increase in intestinal drug absorption and a reduction in active tubular secretion. 4,5 There have been several reports of patients taking digoxin and clarithromycin simultaneously in whom serum digoxin concentrations rose; some developed digoxin toxicity. 1,6 We describe the results of a prospective open observational study carried out to assess the frequency and clinical characteristics of the digoxin clarithromycin interaction in elderly patients chronically taking digoxin. Materials and methods The study was a prospective observational trial of all consecutive patients admitted between 1 December 1999 and 31 March 2000 to the Internal Medicine Service of the University General Hospital of Alicante, and who fulfilled inclusion and exclusion criteria. Inclusion criteria Patients older than 65 years treated with a fixed dose of oral digoxin (capsules) for at least 2 weeks before hospital admission, who were treated with oral clarithromycin after hospital admission and who gave informed consent for blood samples to be taken for monitoring of digoxin levels, were included in the study. Exclusion criteria Exclusion criteria were: treatment with macrolides during the 2 weeks before hospital admission, hepatic insufficiency,... *Corresponding author. Tel: ; Fax: ; zapater_ped@gva.es The British Society for Antimicrobial Chemotherapy

2 P. Zapater et al. serum bilirubin > 25.7 µmol/l, acute renal insufficiency (serum creatinine > 110 µmol/l), nephrotic syndrome, hypothyroidism, pregnancy or contraindications to macrolides. Patients with serum digoxin concentrations >2 ng/ml on admission were also excluded. Steady-state serum digoxin concentrations were determined on the first day, before administration of the first clarithromycin dose and after 5 6 days of concomitant treatment with digoxin and clarithromycin. On the first day, blood samples were obtained before the digoxin dose (pre-clarithromycin and digoxin trough serum concentration) and at 6 h after the digoxin and clarithromycin dose. On days 5 6, a blood sample was obtained at least 6 h after an oral dose of digoxin and another was collected just before the next dose of digoxin (trough serum concentration). Blood samples obtained 6 h after an oral dose of digoxin were named sample 1 and samples collected just before the next dose of digoxin were named sample 2. On other days serum digoxin concentrations were monitored according to the clinical progress of the patient. When digoxin toxicity was suspected or the serum digoxin concentration in a trough sample exceeded 2 ng/ml the drug was withdrawn. Serum digoxin concentrations were determined using a fluorescence polarization immunoassay (TDX; Abbott Laboratories), with limited cross-reactivity between dihydrodigoxin and digitoxin (DRM). 7 A Bayesian approach was used to calculate digoxin pharmacokinetic parameters in every patient with either one or two samples available. Initial systemic clearance (CL) and volume of distribution (V d ) of patients older than 65 years were obtained from the literature: V d = 4.1 ± 0.9 L/kg, CL = 0.8 ± 0.2 ml/min/kg. 8 Absorption was assumed to be first-order with a rate constant of 0.82/h and a bioavailability (F) of 71.5 ± 8.6%. 9 A minimization Marquardt Levenberg algorithm with a convergence criterion of was used to fit pharmacokinetic parameters to individual data. Pharmacokinetic parameter estimates were determined for every patient using the dosage history and the concentrations measured on the first day of digoxin clarithromycin treatment. Calculated estimates were then combined with the dosage history for 4 8 days of concomitant treatment to predict concentrations at the times of sampling on days 4 8. Predicted concentrations were compared with the measured concentrations. Finally, a fit was made for every patient using the dosage history and the concentrations measured after 4 8 days of digoxin clarithromycin treatment to calculate V d, CL, elimination rate constant (K) and elimination half-life (t ½ ). Gender, age, weight, height, digoxin dosage history, the presence of congestive heart failure, and serum creatinine, were all included in every fit. All calculations were made using the Abbottbase Pharmacokinetic System (version 1.00) software. All the observations are reported as mean ± standard deviation (S.D.). V d, CL, K and t ½ statistical differences between groups before and after digoxin clarithromycin treatment were analysed using a paired t-test when data from all patients were available. A P value of <0.05 was considered to indicate statistical significance. Results Seven patients were studied (Table 1). Two men and five women (mean age 76.8 ± 6.6 years) treated with digoxin, six for chronic congestive heart failure with atrial fibrillation and one (patient 5) for isolated atrial fibrillation, were included. Four patients received clarithromycin for bronchitis and three for pneumonia. Digoxin dosage before admission to hospital was five to seven capsules (0.25 mg) per week and the oral clarithromycin dose was the same in all cases: 500 mg twice a day. All patients, except patient 1, took other drugs in addition to digoxin before hospital admission. The drugs that patients began to take in hospital are indicated in Table 1. Patient 4 developed acute renal insufficiency (serum creatinine µmol/l) on day 7 of concomitant treatment with digoxin and clarithromycin and simultaneously developed an increased serum digoxin concentration with electrocardiographic signs of digoxin toxicity (depressed ST segments and inversion of T waves in V5 and V6). Digoxin was discontinued and renal function recovered in 72 h. Patients 6 and 7, treated with amiodarone and verapamil respectively, had sample 1 digoxin concentrations lower than sample 2 concentrations after 7 days of digoxin clarithromycin treatment. The Bayesian method acceptably predicted the digoxin concentrations in all patients on the first day of clarithromycin treatment (Table 2). After fitting the model to the observed concentrations, the calculated V d, CL and t ½ were 4.05 ± 0.08 L/kg, 0.65 ± 0.1 ml/min/kg and 59.8 ± 10.3 h, respectively (Table 2). However, on day 7 of clarithromycin treatment the predicted serum digoxin concentrations before the fit were lower than the observed digoxin concentrations. All observed digoxin concentrations were > 2 ng/ml (Table 2). Pharmacokinetic parameters estimated after the best fit to observed digoxin concentrations were CL = 0.39 ± 0.07 ml/ min/kg (lower than CL calculated on day 1; P = 0.001), K = ± 0.07/h (P = 0.001) and t ½ = ± 27.3 h (greater than t ½ calculated on day 1; P = 0.003). Discussion In patients over 65 years old a higher prevalence of congestive heart failure and atrial fibrillation has been described and they are often treated with chronic digoxin. 10 In our study, a week of concomitant treatment with digoxin and clarithromycin induced a significant increase in serum digoxin concentrations in this patient group. Only one patient with acute renal 602

3 Table 1. Basal characteristics of patients included in the study Patient no. Gender Age Creatinine (years) (µmol/l) Potassium (mmol/l) Calcium Magnesium (mmol/l) (mmol/l) ph po 2 (FIO 2 21%) (mmhg) pco 2 New drugs added (mmhg) Drugs taken by the patient a during hospitalization M none furosemide, spironolactone, captopril, amoxicillin clavulanic acid 2 F acetylsalicylic acid captopril, ranitidine, cefuroxime 3 F acenocoumarol, furosemide, cefotaxime potassium 4 F acenocoumarol, furosemide, enalapril, prednisolone, spironolactone 5 F acenocoumarol, carvedilol, carbimazole, amiloride, hydrochlorothiazide, cerivastatin, glimepiride 6 M acenocoumarol, furosemide, glyceryl trinitrate (transdermal) 7 F furosemide, spironolactone, ranitidine levofloxacin none spironolactone, amiodarone verapamil Clarithromycin digoxin interaction a Drugs taken chronically by the patient before admission to hospital. All drugs and doses were unchanged during the hospital stay.

4 P. Zapater et al. insufficiency (patient 4) had electrocardiographic signs of digoxin toxicity simultaneously with very high digoxin concentrations, out of the detection range. However, all patients had trough digoxin concentrations, at 24 h post-dose, of >2 ng/ml. During hospitalization, four patients began treatment with drugs that could have increased serum digoxin concentrations (spironolactone, which inhibits the excretion of digoxin by the kidney; captopril, by reduction in the loss of digoxin through the renal tubules; verapamil, due to reductions in renal and especially biliary clearance and amiodarone, which reduces both renal and non-renal excretion of digoxin 1 ), but the other three patients did not receive drugs, other than clarithromycin, that could have explained the interaction. We used a Bayesian approach using initial parameters described in previous studies for patients over 65 years old. 10 This approach permitted us to calculate the pharmacokinetic parameters in those patients with only a single digoxin concentration measurement (patients 1 and 5). For all patients, a significant reduction in calculated digoxin clearance was observed after 4 7 days of concomitant treatment with clarithromycin. Interestingly, patients 6 and 7, treated with amiodarone and verapamil respectively, had digoxin concentrations in sample 1 (6 7 h post-dose) lower than those in sample 2 ( h post-dose) after 7 days of digoxin clarithromycin treatment and we are unable to offer a definitive explanation for these results. It is possible that changes in the absorption phase (for example, increased bioavailability of the P-glycoprotein substrate digoxin, as has been observed in vivo with erythromycin 1,4 ) and in renal and non-renal elimination could have occurred. However, confirmation of these observations and determination of the reasons for the changes observed in these patients requires a separate pharmacokinetic trial designed specifically to address this. Patients over 65 years are frequently treated chronically with digoxin. If they receive clarithromycin for respiratory tract infection, the results of this study show that elevated digoxin concentrations may be observed after 5 7 days of concomitant treatment, but only a relatively small number of patients will develop signs of clinical toxicity (in our study one patient with reduction in renal function). However, we emphasize the need for careful monitoring of serum digoxin concentrations in all elderly patients with normal renal function when treated with clarithromycin, with reduction in digoxin dosage as necessary. In patients with reduced renal function or those receiving other drugs that could impair digoxin clearance, clarithromycin should only be used with caution. The only macrolide that has been reported not to affect serum digoxin levels has been rokitamycin 1 and this could be an alternative macrolide to clarithromyin in such patients. However, this would need to be confirmed from clinical trials addressing both its efficacy and potential drug interactions. Table 2. Pharmacokinetic parameters determined using a Bayesian approach after fit (a) (ng/ml) V d (L/kg) CL (ml/min/kg) K (/h) t ½ (h) (a) (ng/ml) Obs. conc. before clarithromycin treatment (ng/ml) T after dose (h) No. T (days) Mean (S.D.) 4.05 (0.08) 0.65 (0.10) (0.0028) 59.8 (10.3) 604

5 Table 2. (Continued) No. T (days) T after dose (h) Obs. conc. after clarithromycin treatment (ng/ml) (b) (ng/ml) after fit (b) (ng/ml) V d (L/kg) CL (ml/min/kg) K (/h) t ½ (h) > > Mean (S.D.) 4.19 (0.11) 0.39 (0.07)* (0.0015)* (27.3)* Clarithromycin digoxin interaction No., number of patient; T, time of concomitant digoxin clarithromycin treatment (days); T after dose, time after digoxin dose at which a sample is obtained (h); Obs. conc. before clarithromycin treatment, observed digoxin concentration before treatment with clarithromycin; (a), predicted concentration of digoxin using initial pharmacokinetic parameters from previously published data: V d = 4.1 ± 0.9 L/kg, CL = 0.8 ± 0.2 ml/min/kg. 8 after fit (a), predicted concentration of digoxin after the best fit of the model to the observed digoxin concentrations; Obs conc. after clarithromycin treatment, observed digoxin concentration after 4 7 days of treatment with clarithromycin; (b), predicted concentration of digoxin after 4 7 days of treatment with clarithromycin using estimates obtained in (a); after fit (b), predicted concentration of digoxin after 4 7 days of treatment with clarithromycin using the best fit of the model to the observed digoxin concentrations; V d = volume of distribution; CL, systemic clearance; K, elimination constant; t ½, elimination half-life. Absorption was assumed to be first order with a rate constant of 0.82/h and a bioavailability of 71.5 ± 8.6%. 9 *P < 0.05 compared with data before clarithromycin treatment.

6 P. Zapater et al. References 1. Stockley, I. H. (1999). Digitalis glycoside drug interactions. In Drug Interactions, 5th edn (Stockley, I. H., Ed.), pp The Pharmaceutical Press, London. 2. Lindenbaum, J., Rund, D. G., Butler, V. P., Jr, Tse-Eng, D. & Saha, J. R. (1981). Inactivation of digoxin by the gut flora: reversal by antibiotic therapy. New England Journal of Medicine 305, Saha, J. R., Butler, V. P., Jr, Neu, H. C. & Lindenbaum, J. (1983). Digoxin-inactivating bacteria: identification in human gut flora. Science 220, Schwarz, U. I., Gramatte, T., Krappweis, J., Oertel, R. & Kirch, W. (2000). P-glycoprotein inhibitor erythromycin increases oral bioavailability of talinolol in humans. International Journal of Clinical Pharmacology and Therapeutics 38, Wakasugi, H., Yano, I., Ito, T., Hashida, T., Futami, T., Nohara, R., Sasayama, S. & Inui, K. (1998). Effect of clarithromycin on renal excretion of digoxin: interaction with P-glycoprotein. Clinical Pharmacology and Therapeutics 64, Xu, H. & Rashkow, A. (2001). Clarithromycin-induced digoxin toxicity: a case report and a review of the literature. Connecticut Medicine 65, Al-Fares, A. M., Mira, S. A. & el-sayed, Y. M. (1984). Evaluation of the fluorescence polarization immunoassay for quantitation of digoxin in serum. Therapeutic Drug Monitoring 6, Cusack, B., Kelly, J., O Malley, K., Noel, J., Lavan, J. & Horgan, J. (1979). Digoxin in the elderly: pharmacokinetic consequences of old age. Clinical Pharmacology and Therapeutics 25, Cohen, A. F., Kroon, R., Schoemaker, H. C., Breimer, D. D., Van Vliet-Verbeek, A. & Brandenburg, H. C. (1993). The bioavailability of digoxin from three oral formulations measured by a specific h.p.l.c. assay. British Journal of Clinical Pharmacology 35, Hanratty, C. G., McGlinchey, P., Johnston, G. D. & Passmore, A. P. (2000) Differential pharmacokinetics of digoxin in elderly patients. Drugs and Aging 17,