(W-087) Comparison of empirical and physiologically-based modeling approaches to explore mechanisms of drug-drug interaction between abemaciclib and olaparib in a small cohort of patients with ovarian cancer

Kasey Hill, PhD – Researcher, The Ohio State University; Nicole Abbott, PhD – Researcher, The Ohio State University; Linda Duska, PhD – Professor, University of Virginia School of Medicine; Andrea E. Wahner Hendrickson, PhD – Professor, Mayo Clinic Comprehensive Cancer Center; John L. Hays, PhD – Professor, The Ohio State University; Geoffrey Shapiro, PhD – Professor, Dana-Farber Cancer Institute; Kathleen Moore, PhD – Professor, University of Oklahoma; Mitch Phelps, PhD – Professor, The Ohio State University

Objectives: Abemaciclib, a selective inhibitor of cyclin-dependent kinase 4 and 6, and olaparib, a poly (ADP-ribose) polymerase (PARP) enzyme inhibitor, are used for cancer treatment. Early toxicity was observed in patients with recurrent platinum-resistant ovarian cancer in an ongoing trial evaluating combined abemaciclib and olaparib (NCT04633239). Since metabolism and transport pathways for abemaciclib and olaparib partially overlap, and since cycle 1 pharmacokinetic (PK) data suggested drug-drug interactions (DDI) in this patient population, we aimed to explore potential drug interaction mechanisms utilizing both population pharmacokinetic (PopPK) and physiologically-based PK (PBPK) modeling and simulation approaches.

Methods: Olaparib 200 or 250 mg was administered alone BID for 7 days and then administered in combination with abemaciclib 50 mg BID starting day 8 with 28-day cycles. A nonlinear mixed-effects model was developed using NONMEM, and a PBPK model was developed using PK-sim®. Physicochemical and absorption, distribution, metabolism, and excretion data of both agents were obtained from literature, and additional model parameters were estimated using clinical data. For both modeling approaches, model evaluation was performed by comparing predicted plasma concentration-time profiles to the observed data.

Results: Both the popPK and PBPK models properly captured the observed profiles of abemaciclib and olaparib through one cycle of treatment. Compared to historical data with abemaciclib alone, the population PK model indicated there was an approximately 70% increase in first-order absorption rate constant and approximately 20% decrease in both apparent volume of distribution and apparent clearance of abemaciclib when administered with olaparib. However, time-varying components had to be incorporated to improve fitting for Days 15 through 28 in both the PopPK and PBPK models. While time-varying clearance had previously been described for olaparib, inclusion of that alone did not account for observed changes in PK of both drugs. Furthermore, with the experimental data currently available, specific enzyme and transporter contributions to the observed DDI could not be assigned.

Conclusions: The developed PK models adequately described the observed plasma concentrations of abemaciclib and olaparib when initially combined in recurrent platinum-resistant ovarian cancer population, and with inclusion of additional time-varying components, the models could adequately describe PK behavior throughout cycle 1. The present models demonstrate underlying mechanisms, beyond those previously described, are involved in time-varying PK and DDI with combined abemaciclib and olaparib, though additional experimental data will be required to confirm the specific contributions of CYP and/or transporter-mediated DDI.

Citations:
[1] Chigutsa E, Kambhampati SRP, Karen Sykes A, Posada MM, van der Walt JS, Turner PK. Development and Application of a Mechanistic Population Modeling Approach to Describe Abemaciclib Pharmacokinetics. CPT Pharmacometrics Syst Pharmacol. 2020;9(9):523-533. doi:10.1002/psp4.12544