(M-095) Population Pharmacokinetic Modeling Informed the Dose of GDC-8264 for a Phase 1b Study in Participants with Acute Graft-Versus-Host Disease

Rachel Garvin, MD – Lead Medical Director, Genentech, Inc.; Michael Rothenberg, MD PhD – Senior Group Medical Director, Genentech, Inc.; Nastya Kassir, PhD – Senior Director & Distinguished Scientist, Genentech, Inc.; Smita Kshirsagar, PhD – Distinguished Scientist, Genentech, Inc.

Objectives: Acute graft versus host disease (GVHD) is a complication that occurs shortly after allogeneic hematopoietic cell transplantation and causes damage to target organs, mainly skin, liver, and gastrointestinal tract. Receptor-interacting protein kinase 1 (RIP1), involved in apoptosis and necroptosis, is activated during this damage process. GDC-8264 is a potent, selective, and reversible small molecule inhibitor of RIP1 activity1. The goal of this work was to use pharmacometric modeling to inform the dose of GDC-8264 to be administered to participants with acute GVHD.

Methods: The observed pharmacokinetic (PK) data in plasma from a phase 1 first-in-human clinical study2 (EudraCT: 2019–002613-19) evaluating GDC-8264 in healthy volunteers was used to build a population PK model describing the PK of GDC-8264. The data included single- (5, 25, 75, 150, and 225 mg) and multiple- (50 and 100 mg once daily (QD), up to 14 days) oral doses administered to a total of 40 subjects. The 5 and 25 mg cohorts received an oral suspension, whereas the other dose cohorts (75–225 mg) received the tablet formulation. All population PK analyses were performed in the non-linear mixed-effect modeling software NONMEM version 7.5 using first-order conditional estimation with interaction.

Results: The analysis dataset included 873 PK observations from 40 subjects. A two-compartment base model with first-order elimination and oral absorption described the PK of GDC-8264 in the plasma. The estimated population typical values were 5.05 L/h for clearance (CL/F), 56.7 L for volume of central compartment (V1/F), and 1.7 1/h for absorption rate constant (Ka). Inter-individual variability (IIV) parameters were estimated on CL/F, V1/F, and Ka. Correlations between IIV parameters and covariates were assessed. The model included the effect of body-weight on CL/F and V1/F—with exponents fixed to a value of 0.75 and 1 respectively—and formulation effect on Ka. The final model was evaluated based on objective function value, goodness of fit plots–including prediction corrected visual predictive checks–and relative standard error on parameter estimates. The population PK model was used to simulate the plasma PK of different dose-levels with variability and to calculate steady-state (SS) exposures. The recommended doses included two dose cohorts of 35 mg (lower-dose)—a dose that was not tested in the clinic—and 75 mg (higher-dose) given as a QD dose for 28 days.

Conclusions: The model described the PK data for GDC-8264 adequately and simulations guided the choice of the doses such that there was adequate separation in the exposures between the two doses. Simulations confirmed that the mean percentage of time that concentration of GDC-8264 at SS remains above predicted EC90 (from pre-clinical data) in one dosing interval was above ~75% at both dose-levels.

Citations: 1. Prado-Acosta M et al. Inhibition of RIP1 improves immune reconstitution and reduces GVHD mortality while preserving graft-versus-leukemia effects. Sci. Transl. Med. 15 (727): eadf8366 (2023).
2. Jones NS et al. A phase I, randomized, ascending-dose study to assess safety, pharmacokinetics, and activity of GDC-8264, a RIP1 inhibitor, in healthy volunteers. Clin Transl Sci. 16 (10): 1997-2009 (2023).