Population pharmacokinetic analyses for belzutifan to inform dosing considerations and labeling

Abstract

Abstract Belzutifan (Welireg, Merck & Co., Inc., Rahway, NJ, USA) is an oral, potent inhibitor of hypoxia‐inducible factor 2α, approved for the treatment of certain patients with von Hippel–Lindau (VHL) disease‐associated renal cell carcinoma (RCC), central nervous system hemangioblastomas, and pancreatic neuroendocrine tumors. It is primarily metabolized by the polymorphic uridine 5′‐diphospho‐glucuronosyltransferase (UGT) 2B17 and cytochrome (CYP) 2C19. A population pharmacokinetic (PK) model was built, using NONMEM version 7.3, based on demographics/PK data from three clinical pharmacology (food effect, formulation bridging, and genotype/race effect) and two clinical studies (phase I dose escalation/expansion in patients with RCC and other solid tumors; phase II in patients with VHL). Median (range) age for the combined studies was 55 years (19–84) and body weight was 73.6 kg (42.1–165.8). Belzutifan plasma PK was well‐characterized by a linear two‐compartment model with first‐order absorption and elimination. For patients with VHL, the predicted geometric mean (% coefficient of variation) apparent clearance was 7.3 L/h (51%), apparent total volume of distribution was 130 L (35%), and half‐life was 12.39 h (42%). There were no clinically relevant differences in belzutifan PK based on the individual covariates of age, sex, ethnicity, race, body weight, mild/moderate renal impairment, or mild hepatic impairment. In this model, dual UGT2B17 and CYP2C19 poor metabolizers (PMs) were estimated to have a 3.2‐fold higher area under the plasma concentration‐time curve compared to UGT2B17 extensive metabolizer and CYP2C19 non‐PM patients. This population PK analysis enabled an integrated assessment of PK characteristics with covariate effects in the overall population and subpopulations for belzutifan labeling.