Abstract 3988: Beyond G12C: Pan-cancer landscape and co-mutational architecture of rare RAS alleles across 211,526 patients

Abstract Background: KRAS G12C has renewed interest in RAS biology, yet most RAS-mutant cancers harbor non-G12C variants whose therapeutic potential is uncertain and tissue-contextual. We aimed to map the prevalence, co-mutation signatures, and potential actionability of rare RAS alleles to inform next-generation clinical trial design. Methods: We performed a pan-cancer analysis of AACR Project GENIE v.18, analyzing 250,018 tumor samples from 211,526 patients to quantify the prevalence of rare KRAS alleles (G12D/V/A/R, Q61, G13D), NRAS, and HRAS alleles by lineage, to define co-mutation signatures and assess clinical actionability. Results: KRAS was altered in 32,946 tumors (16%; 35,614 patients). Rare KRAS alleles demonstrated strong codon-lineage patterns: G12D (9,457) and G12V (7,602) were the most common, each enriched across pancreatic adenocarcinoma (PDAC; 31.8%), colorectal cancer (CRC; 16.7%), and lung cancer (13.2%); G12A (1,542) occurred mainly in LUAD (41.4%) and CRC (20.5%), rarely in PDAC (1.36%); G12R (1,683) was highly pancreas-restricted (69%). G13D (2,596) was dominantly colorectal (55.9%). Q61X (2,279) occurred in PDAC (26%), LUAD (21%), and CRC (16%). HRAS mutations were rare (1,983 tumors 1%; 1,860), most frequently occurring in urothelial carcinoma (12.5%), thyroid cancer (6.3%), and HNSCC (2%). Codon-61 variants (Q61H/K/L/R; 461) accounted for 23.3% of all HRAS mutations, with enrichment in thyroid (22.8%), urothelial (17.4%), melanoma (6.3%), and HNSCC (6.1%); additional recurrent hotspots included G12 (234) and G13 (248). NRAS alterations were detected in 5,882 tumors (3%; 5,715 patients), most frequently in melanoma (1,440; 24.5%), CRC (565; 9.6%), and LUAD (300; 5.1%). Q61X accounted for 51% of all NRAS mutations, most commonly in melanoma (25%), CRC (10%), and LUAD (5%). KRAS-mutant tumors showed significant co-occurrence with TP53, SMAD4, CDKN2A, STK11, KEAP1, and APC (all p0.001); KRAS was mutually exclusive with NF1, RB1, and PTEN (all p0.001). HRAS-mutant tumors significantly co-occurred with NOTCH1, FGFR3, NF1, TERT, AKT1, ARID1A, PIK3CA, CDKN2A, CCND1, CDK4, and CDK6 (all p≤0.01), with weaker associations for STK11, SMAD4, and TP53. NRAS-mutant tumors significantly co-occurred with CDKN2A, APC, BRAF, NF1, and TERT (all p0.001) and were mutually exclusive with TP53, STK11, and PIK3CA. Conclusions: Rare RAS alleles are common across solid tumors and show distinct codon- and lineage-specific architectures. KRAS alterations are broadly pan-epithelial, whereas HRAS and NRAS cluster within discrete niches. Co-mutation patterns differ, with KRAS linked to PDAC/LUAD/CRC tumor-suppressor pathways, HRAS to NOTCH1-FGFR3-PI3K-AKT-CDK signaling, and NRAS to BRAF, NF1, and TERT. These data support co-mutation-driven combinations and lineage-tailored therapeutic strategies for next-generation RAS-targeted approaches. Citation Format: Niamh Coleman, David Hong. Beyond G12C: Pan-cancer landscape and co-mutational architecture of rare RAS alleles across 211,526 patients abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3988.