Application of Static Models to Predict Midazolam Clinical Interactions in the Presence of Single or Multiple Hepatitis C Virus Drugs.

Asunaprevir (ASV), daclatasvir (DCV), and beclabuvir (BCV) are three drugs developed for the treatment of chronic hepatitis C virus infection. Here, we evaluated the CYP3A4 induction potential of each drug, as well as BCV-M1 (the major metabolite of BCV), in human hepatocytes by measuring CYP3A4 mRNA alteration. The induction responses were quantified as induction fold (mRNA fold change) and induction increase (mRNA fold increase), and then fitted with four nonlinear regression algorithms. Reversible inhibition and time-dependent inhibition (TDI) on CYP3A4 activity were determined to predict net drug-drug interactions (DDIs). All four compounds were CYP3A4 inducers and inhibitors, with ASV demonstrating TDI. The curve-fitting results demonstrated that fold increase is a better assessment to determine kinetic parameters for compounds inducing weak responses. By summing the contribution of each inducer, the basic static model was able to correctly predict the potential for a clinically meaningful induction signal for single or multiple perpetrators, but with over prediction of the magnitude. With the same approach, the mechanistic static model improved the prediction accuracy of DCV and BCV when including both induction and inhibition effects, but incorrectly predicted the net DDI effects for ASV alone or triple combinations. The predictions of ASV or the triple combination could be improved by only including the induction and reversible inhibition but not the ASV CYP3A4 TDI component. Those results demonstrated that static models can be applied as a tool to help project the DDI risk of multiple perpetrators using in vitro data.