ABSTRACT:The prediction of clinical drug-drug interactions (DDIs) due to mechanism-based inhibitors of CYP3A is complicated when the inhibitor itself is metabolized by CYP3A, as in the case of clarithromycin. Previous attempts to predict the effects of clarithromycin on CYP3A substrates, e.g., midazolam, failed to account for nonlinear metabolism of clarithromycin. A semiphysiologically based pharmacokinetic model was developed for clarithromycin and midazolam metabolism, incorporating hepatic and intestinal metabolism by CYP3A and non-CYP3A mechanisms.