KRAS amplification and mutation are independent events in gastroesophageal adenocarcinomas (GEA).

70 Background: KRAS mutations are common oncogenic events across cancers, but effective RAS-directed therapies are lacking. However, recent studies support use of PD-1 blockade in most subsets of lung cancer with KRAS short variant mutations (KRAS SV ) (PMID: 28039262), and preclinical data supports combined MEK and SHP2 inhibition in KRAS amplified ( KRAS a ) GEA (PMID: 30093730). We sought to explore the landscape of KRAS altered GEA and compare genomic profiles of KRAS-altered and KRAS wild-type (WT) cases for biomarkers of response to targeted therapies and immune checkpoint inhibitors. Methods: 6,667 tissue specimens from patients with advanced GEA were assayed using hybrid capture-based comprehensive genomic profiling. Tumor mutational burden (TMB) was determined on up to 1.1 Mbp of sequenced DNA and microsatellite instability (MSI) was determined on 95 or 114 loci. Descriptive statistics were used to compare among subgroups. Results: KRAS SV and KRAS a were identified in 11% and 5.8% of gastric adenocarcinoma (GA), respectively, and in 7.2% and 17% of esophageal adenocarcinoma (EA), respectively. KRAS a and KRAS SV were nearly mutually exclusive, co-occurring in only 4.4% of KRAS altered cases. ERBB2 alterations were less common in KRAS SV and KRAS a GEA (both 9%) as compared with KRAS WT GEA (19%) (p = 1.9E-16). EGFR a was less common in KRAS SV versus KRAS a GEA (1.9% vs. 9.3%, p = 2.6E-8), whereas PIK3CA SV was more common in KRAS SV versus KRAS a (16% vs 5.0%, p = 1.5E-11). Median TMB for all groups was similar; however, KRAS SV GEA had a higher mean TMB (9.7 mut/Mb) as compared to KRAS a (5.1 mut/Mb, p = 5.0E-12) and KRAS WT cases (5.8 mut/Mb, p = 2.2E-07). KRAS codon 12/13 accounted for > 80% of predicted pathogenic mutations. MSI-high was also more prevalent in KRAS SV (11.4%) versus KRAS a (0.9%, p = 4.8E-15) and KRAS WT GEA (3.0%, p = 1.8E-25). MSI-high KRAS SV GEA was associated with older patient age (median 72 years) and with high TMB (median 40.9 mut/mb). Conclusions: GA was enriched for KRAS mutation whereas EA was enriched for KRAS amplification. KRAS WT versus KRAS SV versus KRAS a each presented distinct genomic profiles. KRAS a in the absence of KRAS mutation exists in 11% of GEA and warrants further exploration to inform combination treatment strategies.