C. Wagner,1 Y. Pan,2 V. Hsu,1 V. Sinha,1 P. Zhao1; 1Office of Clinical Pharmacology, US Food and Drug Administration, Silver Spring, MD, 2Office of Generic Drugs, US Food and Drug Administration, Silver Spring, MD

BACKGROUND: We have demonstrated successful predictions of PBPK models submitted by drug developers (sponsors) in predicting the effect of CYP inhibitors on substrate drugs. Using the same PBPK knowledgebase maintained by the Office of Clinical Pharmacology at the US FDA, we evaluated predictive performance of sponsors’ PBPK models in predicting the effect of CYP3A inducers on substrates’ PK.
METHODS: Eligible substrate drugs (n=10) and associated clinical interaction cases using CYP3A inducers (n=12) were selected from the PPBK knowledgebase (July 2008 - August 2014) based on criteria defined previously (Wagner, Pan, et al. Clin Pharmacokinet, In Press). Sponsors developed substrate PBPK models without using clinical induction data. The metric for assessing predictive performance was the R value (R = [predicted AUC or Cmax ratio (+/- inducer)]/[observed AUC or Cmax ratio (+/- inducer)], with R within a range of either 0.80-1.25 or 0.50-2.00.
RESULTS: See Figure 1.
CONCLUSION: Based on the 12 cases, the effect of CYP3A induction was well predicted. Additional refinement of the rifampin PBPK model may be necessary.