Abstract
Significant inter-individual variability of exposure for CYP2C19 substrates may be only partly due to genetic polymorphism. Therefore, the in vivo inter-individual variability in hepatic intrinsic clearance (CLint,h) of CYP2C19 substrates was estimated from reported AUC values using Monte Carlo simulations. The coefficient of variation (CV) for CLint,h in poor metabolizers (PM) expected from genotypes CYP2C19*2/*2, CYP2C19*3/*3 or CYP2C19*2/*3 was estimated as 25.8% from the CV for AUC of omeprazole in PMs. With this, CVs of CLint,h in extensive metabolizers (EM: CYP2C19*1/*1), intermediate metabolizers (IM: CYP2C19*1/*2 or *3) and ultra-rapid metabolizers (UM), CYP2C19*17/*17 and *1/*17, were estimated as 66.0%, 55.8%, 6.8% and 48.0%, respectively. To validate these CVs, variability in the AUC of CYP2C19 substrates lansoprazole and rabeprazole, partially metabolized by CYP3A4 in EMs and IMs, were simulated using the CV in CLint,h for CYP2C19 EMs and IMs and 33% of the CV previously reported for CYP3A4. Published values were within 2.5–97.5 percentile range of simulated CVs for the AUC. Furthermore, simulated CVs for the AUC of omeprazole and lansoprazole in ungenotyped populations were comparable with published values. Thus, estimated CLint,h variability can predict variability in the AUC of drugs metabolized not only by CYP2C19 but also by multiple enzymes.
