Abstract B037: Comparison of RAS G12C(ON) and KRAS G12C(OFF) inhibitor activity in parental or KRAS G12C-amplified tumors via mathematical modeling

Abstract RAS cycles between inactive, GDP-bound (RAS(OFF)), and active, GTP-bound (RAS(ON)) states. Oncogenic mutations in RAS shift this equilibrium towards the ON state which interacts with downstream effectors, driving uncontrolled proliferation. G12C is a common oncogenic KRAS mutation in patients with NSCLC and CRC. Several covalent inhibitors of KRAS G12C(OFF) have received FDA approval. In contrast to these inhibitors, the investigational agent elironrasib is a RAS(ON) G12C-selective covalent inhibitor that directly binds to and inhibits G12C(ON). Elironrasib is currently being evaluated in clinical trials for the treatment of KRAS G12C-mutant solid tumors. Herein we describe a mathematical model to compare target inhibition of RAS(ON) and RAS(OFF) G12C-selective inhibitors in naïve tumors and tumors exhibiting amplification or increased GTP reloading (as may occur following upstream RTK activation) of the KRAS G12C mutant allele. This novel pharmacokinetic/target engagement (PK/TE) model describes the quantitative relationship between inhibitor concentration and KRAS G12C target engagement in xenograft tumors in vivo. We applied the mathematical model to data generated using MiaPaCa-2 parental and sotorasib acquired resistance tumors, the latter characterized by amplification of KRAS G12C. Plasma and tumor PK and tumor TE data were collected for elironrasib and two other KRAS G12C(OFF) inhibitors. Covalent targeting of either KRAS G12C(OFF) or KRAS G12C(ON) was incorporated into the model by inclusion of KRAS G12C GTPase activity. A single PK/TE model structure, utilizing inhibitor-specific crosslinking rates and tumor model-specific rates of KRAS G12C synthesis, accurately recapitulated the inhibition time profiles for all treatments. Simulations demonstrated that the depth of TE for RAS(OFF) and RAS(ON) inhibitors is a function of the inhibitor-specific crosslinking rate relative to synthesis rate. As such, faster rates of crosslinking achieved greater maximal TE and also provided greater ability to counteract KRAS G12C amplification. For RAS(OFF) inhibitors TE can be dependent on the rate of GTP hydrolysis, whereas this is not limiting for agents that directly target the RAS G12C(ON) state. Additional simulations showed that KRAS(OFF) inhibitors are more susceptible than elironrasib to increased GTP reloading rates. Taken together, the model highlights that a fast-crosslinking RAS(ON) inhibitor, such as elironrasib, is optimal to control RAS-GTP signaling in tumors with amplification and/or increased GTP reloading rates of KRAS G12C. Consistent with the model predictions, deeper TE and durable tumor stasis were observed with elironrasib relative to G12C(OFF) inhibitors in the aforementioned MiaPaCa-2 models. Collectively, our mathematical model highlights the advantage of RAS G12C(ON) inhibitors, relative to KRAS G12C(OFF), in suppressing KRAS(ON) signaling, underscoring their potential to overcome resistance to KRAS G12C(OFF) inhibitors driven by amplification and/or increased GTP reloading of KRAS G12C. Citation Format: Muhammad Ali Al-Radhawi, Urszula N. Wasko-Kornberg, Jessica Spradlin, Lei Bao, Xing Wei, Yue Huang, Kyle Seamon, David Wildes, Jingjing Jiang, Mallika Singh, Jacqueline A. M. Smith, Zhengping Wang, Lingyan Jiang, Benjamin J. Maldonato, Zhican Wang. Comparison of RAS G12C(ON) and KRAS G12C(OFF) inhibitor activity in parental or KRAS G12C-amplified tumors via mathematical modeling [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: RAS Oncogenesis and Therapeutics; 2026 Mar 5-8; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(5_Suppl_1):Abstract nr B037.