Impact of CYP2C8 single nucleotide polymorphisms on in-vitro metabolism of imatinib to N-desmethyl imatinib

Abstract

Imatinib is a first line therapy for the treatment of chronic myeloid leukaemia (CML). Treatment with imatinib must be continuous and indefinite for most patients to maintain disease control. Despite excellent efficacy and tolerability, up to 50% of CML patients discontinue imatinib due to lack of efficacy and adverse events. Imatinib is metabolised to its main metabolite N-desmethyl imatinib by CYP3A4 and CYP2C8. In vitro human liver microsome (HLM) studies indicate imatinib autoinhibition of CYP3A4-mediated metabolism, suggesting a more significant role for CYP2C8 upon chronic dosing. CYP2C8 is polymorphic and functional effects of the major CYP2C8 polymorphisms CYP2C8*3 and CYP2C8*4 on N-desmethyl imatinib formation are unknown. It was hypothesised that CYP2C8*3 and CYP2C8*4 genetic polymorphisms will decrease imatinib metabolism to N-desmethyl imatinib in HLM. Therefore the aim of this study was to examine the impact of CYP2C8*3 and CYP2C8*4 on N-demethylation of imatinib in HLMs genotyped for CYP2C8*1/*1 (n=5), CYP2C8*1/*3 (n=4), CYP2C8*1/*4 (n=2), in CYP2C8*3/*3 pooled HLM, and in expressed CYP2C8 and CYP3A4 enzymes. Effects of CYP-selective chemical and antibody inhibitors on N-demethylation were also determined. A single enzyme Michaelis-Menten model with substrate inhibition best fitted wild-type CYP2C8*1/*1 HLM kinetic data (median ± SD K𝚒 = 139 ± 61 μM). Three of four CYP2C8*1/*3 HLMs showed single enzyme but no substrate inhibition kinetics. Binding affinity (K𝚖) was approximately 2-fold higher in CYP2C8*1/*3 HLMs as compared to CYP2C8*1/*1 (median ± SD K𝚖 = 6 ± 2 vs 11 ± 2 μM, p=0.04). Intrinsic clearance (Cl𝚒𝚗𝚝) was higher in CYP2C8*1/*3 HLMs compared to CYP2C8*1/*1 (median ± SD Cl𝚒𝚗𝚝 = 19 ± 8 vs 13 ± 2 μl/min/mg, p = 0.25). CYP2C8*3/*3 (pooled HLM) showed highest binding affinity (K𝚖 = 3.6 μM) and weak autoinhibition (K𝚒 = 449 μM) kinetics. N-desmethyl imatinib formation was below the limit of quantification in one CYP2C8*1/*4 HLM, whereas the other CYP2C8*1/*4 HLM showed lower intrinsic clearance (Cl𝚒𝚗𝚝 = 7 vs 11 ± 2 μl/min/mg) due to 2-fold lower catalytic activity (V𝚖𝚊𝚡) compared to the wild-type (V𝚖𝚊𝚡 = 73 vs 140 ± 31 pmol/min/mg). A single enzyme model with substrate inhibition best fitted expressed CYP2C8 kinetic data (K𝚒 = 149 μM). Expressed CYP3A4 showed two site enzyme kinetics with no evidence of autoinhibition. CYP2C8 inhibitors reduced N-demethylation in HLM by 47-75%, compared to 0-30% for CYP3A4 inhibitors. Two unidentified peaks M1 and M2 were found in expressed CYP3A4, whereas they were absent in expressed CYP2C8. These results indicate that CYP2C8*3 may enhance CYP2C8 activity by influencing autoinhibition, and that in vitro the metabolism and autoinhibition of imatinib N-demethylation appears mainly mediated by CYP2C8 and not CYP3A4. CYP2C8*4 appears a reduced functional allele for imatinib N-demethylation.