Quantitative clinical risk assessment of CYP2C, UDP-glucuronosyltransferase, P-glycoprotein induction, and complex drug-drug interactions using TruVivo human hepatocyte triculture platform.

Abstract

Quantitative prediction and clinical risk assessment for induction of drug-metabolizing enzymes and transporters beyond CYP3A has been hindered by low dynamic response in the gold standard hepatocyte monoculture model. A gap in translation of the drug-drug interaction (DDI) potential of a compound is particularly apparent when an inducer also inhibits CYP3A, leading to uncertainty in the potential net clinical outcome for CYP3A substrates. In addition, enzymes such as CYP2C8, CYP2C9, CYP2C19, UGT1A4, and P-glycoprotein, which are coregulated with CYP3A, may result in clinically relevant induction that cannot be derisked by conducting clinical interaction studies with CYP3A substrates. Identification of an in vitro model that demonstrates consistent and well defined induction of enzymes and transporters beyond CYP3A would open the opportunity to avoid unnecessary clinical interaction studies and subsequently have high value in the drug discovery and development toolbox. The TruVivo model is a novel all-human primary cell model, containing hepatocytes plus stromal and epithelial feeder cells. Within these studies, the TruVivo model was validated as a predictive tool for clinical risk assessment for induction of CYP2C8, CYP2C9, CYP2C19, CYP3A4, UGT1A4, and P-glycoprotein using known clinical inducers. Exploration into the utility of TruVivo to delineate complex DDI involving coinducers/inhibitors was also conducted and showed immense opportunity, demonstrating the value of in situ DDI experiments when clinically relevant levels of precipitant and object drugs are used. These data highlight the potential of this in vitro tool to model induction and complex DDI. SIGNIFICANCE STATEMENT: Clinical risk assessment for induction of enzymes and transporters coregulated with CYP3A, including the CYP2C enzymes, UDP-glucuronosyltransferases, and P-gp has been hampered by the low dynamic response of available in vitro models. These studies aimed to validate a novel all-human hepatocyte model, TruVivo, as a predictive tool for induction-based DDI. In addition, the model was evaluated for in situ prediction of complex DDI and shows promise in predicting net clinical outcomes for several inducers/inhibitors against selective and nonselective CYP3A substrates.