Predictive In Vitro-In Vivo Extrapolation for Time Dependent Inhibition of CYP1A2, CYP2C8, CYP2C9, CYP2C19, and...

Area of Science:

• Pharmacology
• Drug Metabolism
• Drug Interactions

Background:

• Cytochrome P450 (CYP450) enzyme inactivation can significantly increase drug exposure.
• Existing in vitro to in vivo extrapolation (IVIVE) data primarily focuses on CYP3A, with limited assessment of other CYP isoforms.
• Accurate prediction of drug-drug interactions (DDIs) is essential for patient safety.

Purpose of the Study:

• To evaluate the utility of human hepatocytes (HHEP) and human liver microsomes (HLM) for predicting clinically relevant DDIs involving CYP1A2, CYP2C8, CYP2C9, CYP2C19, and CYP2D6.
• To compare the predictive accuracy of basic versus mechanistic static models for IVIVE of time-dependent inhibition (TDI).
• To provide recommendations for optimizing IVIVE of TDI for non-CYP3A enzymes.

Main Methods:

• Identified 18 inhibitors from the University of Washington Drug-Drug Interaction Database for 119 clinical interaction studies.
• Characterized in vitro TDI using pooled HHEP and HLM with increasing inhibitor concentrations.
• Incorporated kinetic parameters (k_inact, K_I) into regulatory-recommended static equations and mechanistic models.

Main Results:

• Mechanistic static models utilizing unbound hepatic inlet concentrations achieved the highest prediction accuracy (92% for HHEP, 85% for HLM within twofold of observed values).
• Basic static models recommended by regulatory agencies significantly overpredicted clinical risk.
• Time-dependent inhibition was observed in HLM for four moderate/strong inhibitors, indicating a need for incubation condition optimization for weak inhibition.

Conclusions:

• Coupling TDI parameters from HHEP and HLM with a mechanistic static model offers an accurate and straightforward method for assessing clinical DDI risk.
• Optimization of HLM incubation conditions is necessary for reliably evaluating TDI of weak and moderate inhibitors.
• The findings support the use of mechanistic models for IVIVE of TDI with non-CYP3A enzymes, improving DDI risk assessment.