Jialin Mao, Fang Ma, Ivy Chen, Susan Wong, Hank La, Emile Plise, Liling Liu, Ryan Johnson and Yuan Chen
Department of Drug Metabolism and Pharmacokinetics, Genentech, A Member of the Roche Group
Objectives: Mutation in the gene that encodes Kirsten rat sarcoma viral oncogene homolog (KRAS) is the most common oncogenic driver in advanced non–small cell lung cancer, occurring in approximately 30% of lung adenocarcinomas. The KRAS glycine-to-cysteine mutation (G12C) composes approximately 44% of KRAS mutations in non–small cell lung cancer, with mutant KRasG12C present in approximately 13% of all patients with lung adenocarcinoma [1]. GDC-6036, AMG-510 and MRTX-849 with acrylamides warheads are designed covalent inhibitors for Kras G12C [2-4]. However, the prospective human PK prediction of covalent inhibitors represented as a challenge due to their instability in various in vitro ADME assays, disconnect from in vitro and in vivo extrapolation (IVIVE), and known nontraditional metabolic pathways [5]. The current work is focusing on how human PK of GDC-6036 can be prospectively predicted using the PBPK approach.
Methods: GDC-6036, AMG-510 and MRTX-849 were tested in various in vitro ADME assays, and their pharmacokinetics were characterized in mouse, rat, dog and cynomolgus monkey. Physicochemical parameters and in vitro data were integrated using PBPK modeling to predict the preclinical PK. Preclinical PBPK model development, using a bottom-up approach, for clearance and volume of distribution prediction was evaluated to understand cross-species IVIVE and to gain knowledge regarding how to bridge the IVIVE gap [6]. Such information was subsequently applied to prospective human PK prediction using human in vitro data for GDC-6036. Similar PBPK strategy was applied to AMG-510 and MRTX-849 and the predictions were compared to the reported human PK, in order to gauge such PBPK strategy for covalent inhibitors. Simcyp was used as the PBPK platform for the current investigation.
Results: The preclinical PBPK models were successfully developed for mouse, rat, dog and cynomolgus monkey for GDC-6036, AMG-510 and MRTX-849. The disconnect of clearance and volume of distribution using in vitro data to predict the preclincal observed were observed across the species for all three molecules. The consistent IVIVE observed in preclinical species in GDC-6036 and the successful retrospective human PK prediction of AMG-510 and MRTX-849 using similar strategy increased the confidence of human PK prediction of GDC-6036. The prospective predicted human PK was indeed aligned with the clinically observed human PK of GDC-6036 at 400 mg. It is important to notice on the observed PK profile particularly the reported half life was very consistent with the prediction. The predicted Cmax and AUC was within 2-fold of the reported.
Conclusion: The learning of the current work is that the proposed PBPK strategy provides the understanding of trends of preclinical IVIVE and a practical solution for the prospective human PK prediction of covalent inhibitors. It also demonstrated the possibility of increasing confidence in human PK prediction for covalent inhibitors in the presence of IVIVE disconnect due to insufficient mechanistic understanding.
References:
[1] Kim D., et al., Targeting KRAS G12C: from inhibitory mechanism to modulation of antitumor effect in patients. Cell. 2020 Nov 12; 183(4): 850–859.
[2] Sacher A., et al., Single-Agent Divarasib (GDC-6036) in Solid Tumors with a KRAS G12C Mutation. N Engl J Med, 2023 Aug 24;389(8):710-721.
[3] Lanman et al., Discovery of a Covalent Inhibitor of KRASG12C (AMG 510) for the Treatment of Solid Tumors. Journal of Medicinal Chemistry 2020 63 (1), 52-65.
[4] Fell J et al., Identification of the Clinical Development Candidate MRTX849, a Covalent KRASG12C Inhibitor for the Treatment of Cancer. Journal of Medicinal Chemistry 2020 63 (13), 6679-6693.
[5] Leung L., et al., Clearance Prediction of Targeted Covalent Inhibitors by In Vitro-In Vivo Extrapolation of Hepatic and Extrahepatic Clearance Mechanisms. Drug Metab Dispos. 2017 Jan;45(1):1-7.
[6] Mao J., et al., Shared Learning from a Physiologically-Based Pharmacokinetic Modeling Strategy for Human Pharmacokinetics Prediction through Retrospective Analysis of Genentech Compounds. Biopharm Drug Dispos. 2023 Aug;44(4):315-334
Reference: PAGE 32 (2024) Abstr 11102 [www.page-meeting.org/?abstract=11102]
Poster: Methodology - New Modelling Approaches