Polymorphic CYP2D6-Mediated Stereoselective Metabolism

POLYMORPHIC CYP2D6-MEDIATED STEREOSELECTIVE METABOLISM

It appears probable that the metabolism of both terodiline and prenylamine may be mediated by the P450 cytochrome CYP2D6, the isoform responsi-ble for debrisoquine hydroxylation. This major drug-metabolizing isozyme is expressed polymorphically in all populations, resulting in two major drug-metabolizing phenotypes – extensive (EM) and poor (PM) metabolizers. The latter are unable to effect the metabolic elimination of CYP2D6 substrates, and these include antiarrhythmic agents, β-blockers, anti-hypertensive drugs, neuroleptics and antidepressants. Consequently, PM individuals are exposed to higher concentrations of the parent drug for longer duration.

The pharmacokinetics of prenylamine are enan-tioselective, favouring the elimination of the + - (S)-enantiomer (Gietl et al., 1990; Paar et al., 1990). On multiple dosing, the apparent oral clearance of the (+) -(S)-enantiomer was 4.6-fold and the renal clearance 2.4-fold higher than that of the (−) -(R)-enantiomer. The maximum plasma concentration and AUC (area under curve of plasma concentration vs. time) of the (+) -(S)-enantiomer were 4–5 times lower than those of the (−) -(R)-enantiomer. After a single dose, the mean plasma half-lives of − - (R)-prenylamine and (+) -(S)-prenylamine were 8.2 and 24 hours, respectively. On chronic dosing, the mean half-lives for (−) -(R)-prenylamine and (+) -(S)-prenylamine were reported to be 13.7 and 17.4 hours, respectively (Gietl et al., 1990). However, the appar-ently only slightly higher mean value of the half-life of (+) -(S)-enantiomer following a single dose was mainly a consequence of its extremely long plasma half-lives of 82 and 83 hours in 2 of the 8 volun-teers. The remaining 6 subjects showed an average half-life of 11 hours. Although none of these subjects had been phenotyped for their CYP2D6 metabolic capacity, prenylamine fulfils all the structural require-ments of a CYP2D6 substrate and it is worth spec-ulating whether these two individuals were PMs of CYP2D6 with an impaired ability to eliminate (+) -(S)-prenylamine. Patients with prenylamine-induced proarrhythmias have not been genotyped or pheno-typed for their CYP2D6 metabolizing capacity.

Studies with rat liver microsomes suggest that more than one CYP isoform may be involved in the metabolism of terodiline, with different isoforms mediating the metabolism of the two enantiomers (Lindeke et al., 1987). In studies using human liver microsomes, the metabolism of terodiline at high concentrations has been shown to be stereoselec-tive favouring the (+) -(R)-enantiomer (Noren et al., 1989), although the ratio of concentrations of the two enantiomers at steady-state following administration of clinical doses is close to unity (Hallen et al., 1995).

Although much of the data in man are incom-plete, puzzling or often difficult to reconcile, there is fairly persuasive evidence to suggest that the major isozyme involved in the metabolism of (+) -(R)-terodiline is CYP2D6, and therefore the metabolism of (+) -(R)-terodiline is subject to genetic polymor-phism. The formation of p-hydroxy-terodiline from -(R)-terodiline was found to be impaired in one PM of debrisoquine (Hallen et al., 1993). In this study of the pharmacokinetics of a 25 mg oral dose of + - (R)-terodiline in healthy volunteers, the mean half-life of this enantiomer in 4 EMs of debrisoquine was 42 (range 35–50) hours and in the only PM in this study, it was 117 hours. In another study (Thomas and Hartigan-Go, 1996) in healthy volunteers, which included 7 EMs and 2 PMs who were administered a single oral dose of 200 mg racemic terodiline, the maximum plasma concentrations and AUC of + - (R)-terodiline were significantly higher compared with − -(S)-terodiline, although their half-lives were similar. Even at this high dose (which would be expected to conceal the pharmacokinetic difference between the two genotypes), the PM/EM clearance ratios for (+) -(R)-terodiline and − -(S)-terodiline were 45% and 56%, respectively. In common with all drugs subject to polymorphic metabolism, the phar-macokinetic difference between the EMs and the PMs are less evident at higher doses because of increasing saturation of metabolism in EMs at higher doses.

It is worth pointing out that the (+) -(R)-enantiomer of tolterodine (a structural analogue of terodiline) with anticholinergic properties is marketed for the treat-ment of urinary incontinence. Its oxidative hydroxyla-tion has been confirmed in in vitro and in vivo studies to be mediated principally by CYP2D6 (Brynne et al., 1998; Postlind et al., 1998). CYP3A4-mediated dealkylation provides a major alternative, albeit less effective, route of elimination in those who are PMs of CYP2D6 (Brynne et al., 1999).

The consequence of this stereoselective and (most probably) polymorphic metabolism is that the calcium antagonistic − -(S)-terodiline would accumulate in all patients over time, but in addition there will also be an accumulation of the anticholinergic (+) -(R)-terodiline in the poor and intermediate metabolizers of CYP2D6 substrates. Thus, genetically determined accumulation of (+) -(R)-terodiline could constitute another risk factor. While it is true that the doses used in Sweden and Japan were generally lower, this CYP2D6-mediated metabolism of (+) -(R)-terodiline might also explain the striking inter-ethnic differences in the incidence of ventricular arrhythmias associated with its use. Whereas 9% of the UK population are PMs, the corresponding figures for Sweden and Japan are only 6.8% and less than 1%, respectively. The higher frequency of PM alleles in the UK population will necessarily result in a higher prevalence of the heterozygous CYP2D6 genotype – a subgroup most at risk of drug–drug interactions – and therefore give rise to a higher potential for drug–drug interactions in the United Kingdom between terodiline and other QT interval-prolonging substrates of CYP2D6, such as neuroleptics, antidepressants and other antiarrhyth-mic drugs.

Ford, Wood and Daly (2000) investigated the roles of CYP2D6 and CYP2C19 genotypes in eight patients who survived terodiline-induced proarrhyth-mias (six with torsade de pointes and two with ventric-ular tachycardia). One of these eight patients had a CYP2D6 PM genotype, and it was observed that CYP2D6 alleles were no more frequent in these eight individuals than in the normal population. This study also found a statistically higher frequency of the mutant CYP2C19^∗2 allele in this population. As a result, these investigators suggested that whereas CYP2D6 PM status was not a risk factor for terodiline cardiotoxicity, possession of the CYP2C19^∗2 allele might contribute to adverse cardiac reactions to terodi-line. This study, however, has serious limitations that the investigators themselves have acknowledged. Only two mutant alleles of CYP2D6 were looked for and there was no ECG evidence confirming the adverse drug response phenotype (i.e. the presence of QT interval prolongation or torsade de pointes). There was a lack of information on co-medications in 2 patients. In another 2 patients, there was co-administration of diuretics that may predispose to hypokalaemia, and therefore to torsade de pointes.

It may be speculated whether any of the 12 patients with terodiline-induced proarrhythmias reported to the CSM, and in whom there were no obvious risk factors may have had a phar-macogenetic defect in their CYP2D6-mediated drug metabolism of (+) -(R)-terodiline. Connolly et al., (1991) and Andrews and Bevan (1991) have also reported one case each of torsade de pointes in patients without any risk factors and in whom plasma terodiline levels were markedly elevated. Informa-tion on the genotypes of such patients would have  been more helpful in elucidating the role of (pharma-cokinetic) genetic susceptibility to terodiline-induced proarrhythmias.

In addition, the susceptibility role of CYP2C19^∗2 suggested by Ford, Wood and Daly (2000) does not explain either the absence of terodiline cardiotoxicity among the Japanese (in whom the frequency of the CYP2C19^∗2 allele is much higher at 0.29–0.35), or the high frequency of anticholinergic effects medi-ated by (+) -(R)-terodiline in Scandinavia (where the frequency of the CYP2C19^∗2 allele is far lower, at no more than 0.08). There is also the evidence show-ing that the frequency of this allele is not any higher among the elderly (Yamada et al., 1998), who were the target population for the use of terodiline. Neither can the closely related CYP2C9 isoform be implicated. Terodiline 50 mg daily did not influence the plasma levels of warfarin enantiomers, nor the anticoagulant effect, following continuous daily administration of a mean dose of 5.3 mg warfarin (Hoglund, Paulsen and Bogentoft, 1989).