Predicting the Effect of CYP3A Inducers on the Pharmacokinetics of Substrate Drugs Using Physiologically Based Pharmacokinetic (PBPK) Modeling: An Analysis of PBPK Submissions to the US FDA

Abstract

Background and objective: We recently published analyses regarding the predictive performance of physiologically based pharmacokinetic (PBPK) models, submitted to the US Food and Drug Administration (FDA), for the effect of cytochrome P450 (CYP) inhibitors on the pharmacokinetics of substrate drugs. We now analyze and summarize the predictive performance of PBPK models for the effect of CYP3A inducers on a substrate's pharmacokinetics.

Methods: This analysis was based on 11 substrate PBPK models, developed by six sponsors, using a commercial PBPK software, with 13 clinical interaction studies. Four CYP3A inducers were used: rifampicin, rifabutin, carbamazepine, and efavirenz. Sponsors either directly used the software-provided inducer models or verified these models' induction magnitude prior to use. The metric for assessing predictive performance was the R predicted/observed value [R predicted/observed = (predicted mean exposure ratio)/(observed mean exposure ratio)], with the exposure ratio defined as maximum plasma concentration (C max) or area under the plasma concentration-time curve (AUC) with and without an inducer.

Results: In 77% (10/13; AUCR) and 83% (10/12; C max R) of the cases, the R predicted/observed values were within 1.25-fold of the observed data. Cases with R predicted/observed values >1.25-fold (>twofold for all three AUCR) were under-predictions as a result of using the PBPK software's default rifampicin model. Improved predictions were observed when the rifampicin model was modified by increasing the induction potency.

Conclusion: Based on submissions to the FDA, and similar to our previous findings for CYP inhibition, we observed good agreement between PBPK-predicted and observed effect of CYP3A inducers on substrate pharmacokinetics. Verification of the inducer model appears to be crucial for improved predictive performance.