Development of a Korean‐specific virtual population for physiologically based pharmacokinetic modelling and simulation

Abstract

Physiologically based pharmacokinetic (PBPK) modelling and simulation is a useful tool in predicting the PK profiles of a drug, assessing the effects of covariates such as demographics, ethnicity, genetic polymorphisms and disease status on the PK, and evaluating the potential of drug-drug interactions. We developed a Korean-specific virtual population for the SimCYP® Simulator (version 15 used) and evaluated the population's predictive performance using six substrate drugs (midazolam, S-warfarin, metoprolol, omeprazole, lorazepam and rosuvastatin) of five major drug metabolizing enzymes (DMEs) and two transporters. Forty-three parameters including the proportion of phenotypes in DMEs and transporters were incorporated into the Korean-specific virtual population. The simulated concentration-time profiles in Koreans were overlapped with most of the observed concentrations for the selected substrate drugs with a < 2-fold difference in clearance. Furthermore, we found some drug models within the SimCYP® library can be improved, e.g., the minor allele frequency of ABCG2 and the fraction metabolized by UGT2B15 should be incorporated for rosuvastatin and lorazepam, respectively. The Korean-specific population can be used to evaluate the impact of ethnicity on the PKs of a drug, particularly in various stages of drug development.