Characterization of intrinsic and acquired resistance to KRAS G12C inhibitors across a broad collection of cancer cell line models

Mutations in KRAS are among the most common oncogenic driver events in human cancer. A glycine-to-cysteine mutation at position 12 (KRAS G12C ) impairs GTP hydrolysis, keeping KRAS in the GTP-bound active state. For decades, the KRAS G12C mutation was considered undruggable, until the discovery of a new pocket beneath the effector switch II region of KRAS. Covalent binding of inhibitors into this pocket locks KRAS in its inactive, GDP-bound state. This breakthrough led to the accelerated approval of the first-generation inhibitors sotorasib (AMG 510) and adagrasib (MRTX849). However, their clinical benefit remains limited by modest response rates and the rapid emergence of drug resistance. To address these limitations, next-generation KRAS G12C inhibitors such as divarasib (GDC-6036) have been developed, showing improved selectivity, potency and clinical response rates. Nevertheless, resistance is expected to arise, underscoring the importance of early insights into both intrinsic and acquired resistance to guide the most effective use of these emerging agents in the clinic. To investigate intrinsic resistance, adagrasib, sotorasib and divarasib were profiled across 140 cancer cell lines using nine-point dose ranges in cell viability assays. KRAS G12C -mutant cell lines exhibiting limited potency or efficacy were classified as intrinsically resistant. To study acquired resistance, inhibitor-sensitive KRAS G12C -mutant cell lines were cultured with escalating doses of inhibitor. For both intrinsic and acquired resistant cell lines, resistance mechanisms were characterized using genomic and transcriptomic analyses, determination of MAPK and PI3K pathway phosphorylation after drug exposure, and broad therapeutic and combination profiling to identify pathway dependencies and strategies to overcome resistance. Our analyses revealed heterogeneous responses of KRAS G12C -mutant cell lines to the inhibitors, mirroring clinical responses. The KRAS G12C -mutant cell lines SW1573 and OV56 were intrinsically resistant to all three KRAS G12C inhibitors. Co-mutation analysis revealed a PTEN loss-of-function mutation in OV56, suggesting enhanced PI3K pathway activity. Although PTEN loss was restricted to OV56, both OV56 and SW1573 were resistant to inhibitors of various MAPK pathway components, including RAF (tovorafenib, belvarafenib), MEK (trametinib), and ERK (ulixertinib), suggesting MAPK pathway independence. Cross-resistance to adagrasib, sotorasib and divarasib was also observed in the acquired resistant models. Intrinsically KRAS G12C inhibitor-resistant cell lines, along with acquired resistant models, represent valuable systems for evaluating next-generation KRAS inhibitors and identifying new drug combinations to optimize therapeutic benefit of KRAS G12C inhibitors in patients.