Biol. Pharm. Bull. 29(3) 517—521 (2006)

nally developed as an anti-anginal drug with the properties of a calcium antagonist. The drug also blocks sodium and potassium channels, and prolongs action potentials and refractory periods of normal ventricular and arterial myocardium. These multiple ion channel-blocking properties similar to amiodarone suggest clinical potential in the management of patients with arrthythmias. Indeed, several groups in Japan have reported a beneficial effect of bepridil in the treatment of patients with atrial fibrillation (AF), though there seemed to be large interindividual variability in the clinical efficacy. That is, Yoshida et al. reported that bepridil effectively prevented paroxymal AF in 15 of 23 patients, and Nakazato et al. reported that bepridil restored sinus rhythm in 65 of 112 patients with persistent AF with a mean conversion time of 2.1 months. In addition, Fujiki et al. reported that the oral administration of bepridil alone or in combination with aprindine restored sinus rhythm in 22 of 32 patients with long-lasting persistent AF, and that the rate of termination of AF was enhanced by the coadministration of aprindine. They also found that pharmacological conversion with bepridil alone or in combination with aprindine recovered atrial mechanical function better and maintained sinus rhythms longer than electrical conversion in patients with long-lasting persistent AF. There is limited information available regarding the clinical pharmacokinetics of bepridil and the variability therein. Benet reviewed several studies, and reported that bepridil is completely metabolized presumably by hepatic oxidative processes. Wu et al. also reported that oxidative reactions at seven sites are involved in metabolic pathways for bepridil, but that the formation of metabolites is adequately described by four interrelated pathways, namely, aromatic hydroxylation, followed by N-dealkylation, N-debenzylation, and Nacetylation. In addition, unpublished data obtained by Sankyo Co., Ltd. (Tokyo, Japan) suggested that bepridil is mainly metabolized by hepatic cytochrome P450 (CYP) 2D6 and to a lesser extent by CYP3A4 (personal communication). The genetic polymorphism of CYP2D6 has been studied extensively, and more than 50 variants have been documented. Among Asian extensive/intermediate metabolizers, the three most common alleles of the CYP2D6 gene are CYP2D6*1, *2, and *10. The mutant allele of CYP2D6 (CYP2D6*2) does not affect the enzymatic activity, whereas the CYP2D6*10 allele reduces the affinity of CYP2D6 for several drugs. On the other hand, more than 30 single nucleotide polymorphisms have been identified in the CYP3A4 gene. For the most common variant, CYP3A4*1B, increased transcription was demonstrated in vitro, which may theoretically result in greater enzymatic activity in vivo. However, the allele frequency was 0% in Japanese. On the other hand, CYP3A5 is another important CYP3A protein in the liver, and exhibits significant overlap with CYP3A4 in substrate specificity. The defective allelic variant, CYP3A5*3, is frequently observed among Japanese. In addition, precisely why the use of aprindine together with bepridil increases the rate of AF termination has not been clarified. It is possible that coadministration of aprindine prevents the metabolism of bepridil, because aprindine is a competitive inhibitor of CYP2D6. The present study was designed to evaluate the variability in the pharmacokinetics of bepridil in Japanese patients with arrhythmias, and to investigate the effects of aprindine as well as the CYP2D6 and CYP3A5 genotypes on the oral clearance of bepridil. A pharmacokinetic analysis was performed using a nonlinear mixed effects model (NONMEM) program, because population pharmacokinetics based on a NONMEM analysis can simultaneously evaluate the mean pharmacokinetic parameters, the covariates affecting the pharmacokinetics of a drug, and also unknown interand intra-individual pharmacokinetic variability.