Utility of recombinant enzyme kinetics in prediction of human clearance: impact of variability, CYP3A5, and CYP2C19 on CYP3A4 probe substrates.

A systematic kinetic analysis of the metabolism of five benzodiazepines (low to high clearance compounds) was performed in CYP3A4, CYP3A5, and CYP2C19 baculovirus-expressed recombinant systems. The data obtained in the expression systems were scaled and compared with human liver microsomal predicted clearance and observed in vivo values, using either cytochrome P450 relative activity factors (RAFs) or the relative abundance approach. Interindividual variability, both in content (CYP3A4, CYP3A5) and activity (CYP3A4, CYP3A5, and CYP2C19), were incorporated in the clearance prediction by bootstrap analysis. These resampling Monte Carlo-based simulations were performed to justify any distribution assumptions in the generated range of the predicted clearance due to a limited sample size. This approach allowed extrapolation of the recombinant clearance data to specific population groups and investigation of the role of "minor" forms like CYP3A5 and CYP2C19 in comparison to the most prolific CYP3A4. The use of quinidine 3-hydroxylation and alprazolam 1'-hydroxylation as RAF markers for CYP3A4 and CYP3A5 activity, respectively, and the incorporation of variability improved the clearance prediction of the selected benzodiazepines (apart from flunitrazepam) to within 2-fold of the in vivo value. Clearance estimates from the immunoquantified protein levels were approximately 8-fold lower in comparison to the RAF approach. The differences observed in the benzodiazepine metabolite pathway ratios between CYP3A4 and CYP3A5, particularly for 1'- to 4-hydroxymidazolam and alprazolam, provided a useful measure of interindividual differences within the CYP3A family.