(M-034) Population Pharmacokinetic Modeling of Niraparib to Assess Different Absorption Models

Yoko Franchetti, PhD, PhD – Associate Director, GSK, Collegeville, PA; Mathieu Desrosiers, MSc – Scientist, Certara USA, Inc., Radnor, PA; Mirjam Trame, PharmD, PhD – Vice President, Certara USA, Inc., Radnor, PA; Roxanne Jewell, PhD – Senior Director, GSK, Durham, NC

Objectives: Niraparib is an orally available, potent, highly selective poly (adenosine diphosphate-ribose) polymerase (PARP)-1 and PARP-2 inhibitor approved for maintenance treatment of advanced ovarian cancer. The objectives were 1) to update an existing population PK model for niraparib with rich PK sampling data to refine the absorption model and 2) to perform a covariate analysis on the updated model.

Methods: A previous niraparib model (refined PRIMA model) was based on data from four studies: 1) PN001, a Phase 1 study in advanced solid tumors; 2) NOVA, a Phase 3 study in recurrent ovarian cancer; 3) QUADRA, a Phase 2 study in advanced relapsed ovarian cancer; and 4) PRIMA, a Phase 3 study in first-line advanced ovarian cancer. The refined PRIMA model was a three-compartment model with linear elimination and zero-order drug release into the absorption compartment preceded by a lag time and followed by first order absorption into the central compartment. In the current work, the refined PRIMA model was updated with rich sampling data from two additional studies (HEPATIC in moderate hepatic impairment and TABLET in advanced solid tumors); further absorption model development was conducted.

Results: The resulting 2023 reference model had the same systemic structure, but the updated absorption structure included three transit compartments. Identified covariate effects were:
• Increase in apparent clearance (CL/F) with increasing baseline albumin (ALB) and baseline creatinine clearance (CRCL)
• Decrease in CL/F with increasing baseline alkaline phosphatase (ALP)
• Increase in relative bioavailability in the fed state relative to the fasted/unknown state
• Increase in mean transit time in the fed state relative to the fasted/unknown state
• Increase in apparent central volume of distribution (Vc/F) with increasing baseline ALB and baseline body weight
• Increase in apparent first peripheral volume of distribution with increasing baseline ALB
Typical value estimates for CL/F and for Vc/F for a 70-kg participant were 15.9 L/h and 450 L. Except for extremely low or high baseline values of ALB, ALP, or CRCL, the covariate effects were predicted to result in minor effects on niraparib exposure over the range of covariate values in the dataset (≤25% change in exposure relative to reference).

Conclusions: The additional rich PK sampling data enabled a thorough evaluation of the absorption phase to establish the 2023 reference model. Despite the difference in the absorption structure from the refined PRIMA model, the estimates of CL/F and the total apparent volume of distribution were very similar, and similar covariate effects were identified in the two models. Thus, the transit compartment model established in the current analysis does not represent a departure from the previous refined PRIMA model but rather a model refinement. The 2023 reference model will be utilized to guide future niraparib analyses. This analysis was supported by GSK.

Citations: None