LB-011 - DEVELOPMENT AND VERIFICATION OF MECHANISTIC PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODEL FOR ENZALUTAMIDE AND ITS N-DESMETHYL METABOLITE M2 AS INDUCERS OF CYP3A4

L. Zhou¹, N. Patel², W. Chen³, M. Yu³, L. Jin³; ¹ARCUS Biosciences, , , ²Certara Simcyp Division, Certara Inc., , , ³Arcus Bioscience, , .

Background: • Enzalutamide is an oral anticancer drug and a strong CYP3A4 inducer. Its N-desmethyl metabolite M2 is a major circulating metabolite. CYP3A4 induction potential of M2 has not been well characterized.
• Defining M2’s potential in CYP3A4 induction becomes important in mechanistically understanding CYP3A4-mediated drug-drug interaction (DDI) of enzalutamide.

Methods: • The Maximum induction effect (Emax) and half maximal effective concentration (EC50) of both enzalutamide and M2 for CYP1A2, CYP2B6 and CYP3A4 were generated in-house using human hepatocytes (n=3 donors).
• Enzalutamide PBPK model including mechanistic absorption, distribution, and enzyme kinetics-based elimination was built in Simcyp V20. Formation rate of M2 metabolite was assigned based on the measured M2 concentration in the clinical study after dosing enzalutamide 160 mg (QD).
• The predictive performance of the model was assessed in its ability to predict Cmax and AUCinf of enzalutamide after single dose (160 mg) as well as at steady state after multiple doses of enzalutamide (30 to 600 mg, QD) using PK parameters available in the literature.
• The final model was then used to predict the PK of midazolam, with and without incorporation of M2-driven CYP3A4 induction, in a DDI study with midazolam (2 mg) after multiple oral doses of enzalutamide (160 mg).

Results: • The developed model well predicted the plasma concentration-time profiles of enzalutamide and M2 after single dose of 160 mg enzalutamide.
• The model also described the non-linearity of enzalutamide PK due to auto-induction and solubility limited absorption after multiple doses of enzalutamide (30 to 600 mg, QD).
• The DDI with midazolam was under-predicted in AUC and Cmax when only induction by enzalutamide was included (>2 fold of underprediction). When induction of CYP3A4 by both enzalutamide and M2 were included, the model predicted DDI with midazolam accurately with predicted Cmax and AUCinf ratio of 0.21 and 0.15, compared to clinically observed values of 0.23 and 0.14, respectively.

Conclusion: • The comprehensive PBPK model demonstrated the significant contribution of M2 in accurately predicting the magnitude of CYP3A4 induction observed in clinical studies with enzalutamide.