European Journal of Pharmaceutical Sciences 10 (2000) 285–294 www.elsevier.nl / locate / ejps A pharmacokinetic simulation model for ivabradine in healthy volunteers a, b b c b * Stephen B. Duffull , Sylvie Chabaud , Patrice Nony , Christian Laveille , Pascal Girard , a Leon Aarons a School of Pharmacy and Pharmaceutical Sciences, University of Manchester, Manchester, UK b Service de Pharmacologie Clinique, EA643, Lyon, France c ´ Institut de Recherche Internationales Servier, Courbevoie, France Received 2 November 1999; received in revised form 20 January 2000; accepted 27 January 2000 Abstract Ivabradine is a novel bradycardic agent that has been developed for the prevention of angina. Ivabradine has an active metabolite S-18982. The aim of this study is to develop a pharmacokinetic simulation model. Pharmacokinetic data from two studies were pooled and included data from a total of 66 healthy male volunteers. The data were collected following single dose intravenous and multiple dose oral administration of ivabradine. The multiple dose regimens were administered every 12 h and there were seven active dosing levels. The modelling was performed using the NONMEM software. The model was assessed in terms of its ability to describe the original data set used in its construction and also data arising from a different clinical pharmacology study involving 12 additional subjects. The pharmacokinetics of ivabradine and S-18982 were best described by two linked two compartment intravenous bolus and first-order input, with first-pass loss, and first-order output model. When the model was used for simulation it produced an adequate description of both the original data and data arising from a different clinical pharmacology study.  2000 Elsevier Science B.V. All rights reserved. Keywords: Pharmacokinetics; NONMEM; Ivabradine; Modelling; Healthy volunteers; Phase I clinical pharmacology study 1. Introduction The pharmacokinetic models were developed independent- ly for ivabradine and S-18982 and also independently for Ivabradine is a novel negative chronotropic agent that intravenous and oral administration. There was no evi- has been developed for the prevention of myocardial dence of non-linearity in the pharmacokinetic model. The ischaemia. It acts by reducing the slope of the slow pharmacokinetic analyses were then linked to the heart rate diastolic depolarisation in the automatic cells of the endpoint by way of a hypothetical effect compartment. The sinoatrial node by inhibiting the i pacemaker current. It is current analysis incorporates these 18 subjects with an f reported to have no effect on cardiac contractility in terms additional 60 subjects from another study. of positive or negative inotropic effects. The N-dealkylated We consider that the goals of model building with metabolite, S-18982, has also been shown to decrease heart respect to a traditional pharmacokinetic analysis, where rate in pre-clinical studies when administered intravenous- covariate identification is often an important goal, to be ly or orally to dogs (Ragueneau et al., 1998). There are no different to those that are required for simulation purposes. reports of the administration of S-18982 to humans. In the latter case, it is not a requirement that the model be An account of the pharmacokinetics and pharmaco- biologically exact as long as the model reflects accurately dynamics of ivabradine has been published previously both the population mean concentration–time course of the (Ragueneau et al., 1998). In this report the phar- drug in question and also the stochastic variation about this macokinetics and pharmacodynamics were studied from a mean associated with between- and within-subject vari- group of 18 subjects who received both oral and intraven- ability. In addition, it should be able to be expanded to ous dosing. The pharmacodynamic endpoint was heart rate. account for new scenarios (Holford et al., 1999). Based on the second of these requirements an underlying mechanis- tic basis for the model is preferred (e.g., parameterisation *Corresponding author. School of Pharmacy, University of Queens- in terms of clearances rather than rate constants) such that, land, Brisbane, QLD 4072, Australia. Tel.: 161-7-3365-8808; fax: 161- for any given change in potential covariates (that have not 7-3365-1688. E-mail address: sduffull@pharmacy.uq.edu.au (S.B. Duffull) necessarily been identified a priori), a projected effect on 0928-0987 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0928-0987(00)00086-5