Development of a physiologically-based pharmacokinetic model of enzalutamide, a strong CYP3A4 inducer
Neel Deferm, Xian Pan, Cong Liu, Jia Ning, Charlotte Cross, Jean Dinh, Iain Gardner
Certara UK Limited, Simcyp Division
Introduction: Enzalutamide (Xtandi®) is a second-generation androgen receptor inhibitor and is approved in more than 40 countries for treating patients suffering from metastatic castration-resistant prostate cancer (mCPRC) [1]. Enzalutamide is rapidly absorbed from the gastrointestinal tract, with a time-to-peak plasma concentration of approximately 1.5h [1]. Moreover, the compound is significantly bound to plasma proteins (> 97%) and is primarily metabolized in the liver by both CYP3A4 and CYP2C8. The FDA lists enzalutamide as a moderate inducer of CYP2C9 and CYP2C19 and a strong inducer of CYP3A4 [2]. Consequently, co-administration of enzalutamide with a CYP3A4 substrate could potentially lead to significant drug-drug interactions (DDIs) in the clinic.
Objectives: This study aimed to develop a physiologically-based pharmacokinetic (PBPK) model for enzalutamide and to evaluate its impact on the exposure of the CYP3A4 substrate midazolam.
Methods: A minimal PBPK model with a single adjusting compartment (SAC) was developed for enzalutamide in Simcyp V22. Qsac and Vsac were estimated by parameter estimation against clinical data from Gibbons et al. [1]. The absorption of enzalutamide was described with a first-order absorption model with human intestinal permeability predicted using experimentally obtained Caco-2 data [3]. As no fm values of CYP3A4 or CYP2C8 were reported in the literature, the total (liver) clearance was first estimated using the retrograde calculator. Subsequently, CYP3A4 and CYP2C8 CLint values were optimized to recover the results of the reported DDIs of enzalutamide with gemfibrozil and its metabolite gemfibrozil glucuronide (CYP2C8 inhibitors) and itraconazole and its hydroxy metabolite (CYP3A4 inhibitors) [2]. For induction of CYP2C8 and CYP3A4, the reported in vitro Indmax and IndC50 values were used (after calibration), the CYP2C8 Ki value for enzalutamide was optimized to recover the clinical DDI with pioglitazone [2-3].
Results: The predicted plasma concentration-time profiles following single and multiple oral administrations of enzalutamide at different doses (60, 150, and 360 mg) in healthy volunteers, hepatic impairment patients, and mCRPC patients were consistent with the observed profiles, indicating that the model accurately predicts the PK of enzalutamide. In addition, our results also showed that the predicted versus observed Cmax and AUCinf ratios following co-administration of enzalutamide with gemfibrozil (1.00 and 0.67) and itraconazole (1.03 and 0.99) were within 1.5-fold error. Similarly, the predicted versus observed Cmax and AUCinf ratios of the DDI between enzalutamide and pioglitazone (0.98 and 0.93), as well as the DDI between enzalutamide and midazolam (0.87 and 1.11) were accurately predicted.
Conclusion: The developed enzalutamide PBPK model is able to recover observed clinical single and multiple-dose data in healthy volunteers, hepatic impairment patients and mCRPC patients. In addition, the predicted Cmax and AUCinf ratios of the DDIs between enzalutamide and various CYP2C8 and CYP3A4 substrates and inhibitors were within 1.5-fold error. More importantly, the interaction between enzalutamide and midazolam was well-predicted, indicating that model performance is sufficient and that it can be employed as a CYP3A4 inducer.
References:
[1] Gibbons, J.A., et al., Clinical Pharmacokinetic Studies of Enzalutamide. Clin Pharmacokinet , 2015. 54 (10): p.1043 55.
[2] Gibbons, J.A., et al., Pharmacokinetic Drug Interaction Studies with Enzalutamide. Clin Pharmacokinet , 2015. 54 (10): p. 1057 69.
[3] FDA,Xtandi Pharmacology Review. 2012.