Participation of CYP2C8 and CYP3A4 in the N-demethylation of imatinib in human hepatic microsomes

Abstract

Background and purpose: Imatinib is a clinically important inhibitor of tyrosine kinases that are dysregulated in chronic myelogenous leukaemia and gastrointestinal stromal tumours. Inter-individual variation in imatinib pharmacokinetics is extensive, and influences drug safety and efficacy. Hepatic cytochrome P450 (CYP) 3A4 has been implicated in imatinib N-demethylation, but the clearance of imatinib decreases during prolonged therapy. CYP3A phenotype correlates with imatinib clearance at the commencement of therapy, but not at steady state. The present study evaluated the possibility that multiple CYPs may contribute to imatinib oxidation in liver.

Experimental approach: Imatinib biotransformation in human liver microsomes (n= 20) and by cDNA-expressed CYPs was determined by LC-MS. Relationships between imatinib N-demethylation and other drug metabolizing CYPs were assessed.

Key results: N-desmethylimatinib formation was correlated with microsomal oxidation of the CYP3A4 substrates testosterone (ρ= 0.60; P < 0.01) and midazolam (ρ= 0.46; P < 0.05), and the CYP2C8 substrate paclitaxel (ρ= 0.58; P < 0.01). cDNA-derived CYPs 2C8, 3A4, 3A5 and 3A7 supported imatinib N-demethylation, but 10 other CYPs were inactive; in kinetic studies, CYP2C8 was a high-affinity enzyme with a catalytic efficiency ∼15-fold greater than those of CYPs 3A4 and 3A5. The CYP3A inhibitors ketoconazole and troleandomycin, and the CYP2C8 inhibitors quercetin and paclitaxel decreased imatinib oxidation. From molecular modelling, the imatinib structure could be superimposed on a pharmacophore for CYP2C8 substrates.

Conclusions and implications: CYP2C8 and CYPs 3A contribute to imatinib N-demethylation in human liver. The involvement of CYP2C8 may account in part for the wide inter-patient variation in imatinib pharmacokinetics observed in clinical practice.

Figure 1

CYP-dependent oxidation of imatinib to N-desmethylimatinib.

Figure 2

LC–MS/MS traces of (A) imatinib, (B) N-desmethylimatinib and (C) imatinib-d8 (internal standard) following solid-phase extraction of a human sample that had the authentic analytes added (as described in Methods).

Figure 3

Linear relationships between microsomal imatinib N-demethylation and (A) CYP3A-dependent testosterone 6β-hydroxylation, (B) CYP3A-dependent midazolam 1′-hydroxylation and (C) CYP2C8-dependent paclitaxel 6α-hydroxylation.

Figure 4

(A) Representative kinetic analysis of hepatic microsomal imatinib N-demethylation in HL42, (B) imatinib N-demethylation mediated by cDNA-expressed CYPs and (C) kinetic analysis of imatinib N-demethylation mediated by cDNA-expressed CYP2C8, CYP3A4 and CYP3A5. Data were derived in at least duplicate incubations that varied by <8% from the stated mean values.

Figure 5

(A) Effects of CYP-specific inhibitors on imatinib N-demethylation (mean ± SEM in n = 3 individual hepatic microsomes), and (B) relationship between relative microsomal CYP2C8/CYP3A4 activities and susceptibility to inhibition by the CYP2C8 inhibitor quercetin (20 µM; n = 9 individual microsomal fractions).

Figure 6

Structure of imatinib fitted to (A) the CYP2C8 and (B) the CYP3A4 pharmacophore models reported by Melet et al. (2004) and Ekins et al. (1999) respectively. Shown are hydrophobic regions (green sphere), polar groups (blue sphere) and hydrogen bond acceptors (red sphere).