Population Pharmacokinetic and Pharmacodynamic Modeling of AZD4901 and Simulation to Support Dose Selection for the Phase 2a Study

Abstract

Significant and reversible reductions in testosterone levels were observed with AZD4901 in both preclinical and clinical testing. A comprehensive population pharmacokinetic/pharmacodynamic (PK/PD) modeling of AZD4901 concentration and testosterone relationship from 3 phase 1 studies was performed using NONMEM to support dose selection for phase 2a development. A 2-compartment model with first-order absorption and first-order elimination best described AZD4901 PK. Circadian rhythm of baseline testosterone concentrations was well described by a cosine function. An indirect response model with inhibition of testosterone production was used to link the AZD4901 concentration to testosterone response. The AZD4901 concentration to yield 50% maximum testosterone suppression (IC50) was estimated to be 230 ng/mL. Based on simulations, following 40 mg twice daily (BID) treatment, the AZD4901 steady-state trough concentration will be much higher compared to 80 mg once daily (QD). The AZD4901 concentration time above IC50 after 40 mg BID is 84% of the time of the dosing interval compared to only 49% after 80 mg QD. The mean predicted testosterone concentrations at steady state are lower and overall less variable over 24 hours for 40 mg BID dosing compared to 80 mg QD dosing. Population PK and PK/PD analyses demonstrated that AZD4901 40 mg BID is a better dosing strategy to more consistently suppress testosterone during the entire dosing interval. Consequently, 40 mg BID dosing was suggested in a phase 2a trial in females with polycystic ovary syndrome, and the trial resulted in a positive outcome as shown by significant testosterone decrease compared to placebo.