Abstract A005: Canine hemangiosarcoma as a model for RAS-mutated human cancers: Preliminary data

Abstract Canine cancer patients represent an important translational model for human cancer research with their spontaneous tumor pathogenesis and the many histological, genetic, molecular and clinical similarities between canine and human tumors. Canine tumors provide an opportunity to investigate shared biomarkers and their response to therapeutic interventions over an accelerated timeline compared to human cancers due to differences in expected lifespan and other factors. Dogs, like humans, have three related RAS genes, HRAS, KRAS, and NRAS, with extremely similar genetic sequence and resultantly high protein homology. These proteins have many key molecular functions including roles in cell proliferation and survival. Across multiple human cancer types, recurrent gain-of-function missense mutations in the RAS-family genes occur primarily at codons 12, 13, and 61. These mutations result in constitutively active RAS proteins. The frequencies of mutations across the three RAS genes and codons have been hypothesized to lead to distinct biological behaviors. Mutations are most common in KRAS, then NRAS followed by HRAS. We investigated RAS-family genetic mutations in a dataset of 213 canine patients with hemangiosarcoma (HSA), a common, aggressive malignancy derived from endothelial cells for which treatment with both surgery and traditional chemotherapies often fail to prolong survival beyond six months. Recent publications show that genomic profiling of canine HSA tumors can help group tumors into subtypes with potential for guiding therapy selection and improving outcomes. All 213 dogs had tumor tissue sequenced via the FidoCure® Precision Medicine Platform targeted sequencing panel and were followed until time of death or were censored to the date patients were last known to be living. All SNVs and Indels detected were missense mutations. Unlike human pan-cancer data, somatic mutations in NRAS (n=32) were most common with only 1 dog with a somatic KRAS mutation and no dogs with somatic HRAS mutations. Similar to human RAS mutations, affected codons were primarily codon 61 (n=28), codon 13 (n=5), and codon 12 (n=1). Pairwise survival comparisons between dogs with missense mutations in NRAS codon 61 showed that this mutation was associated with better prognosis (n=28, MST 251 [CI 213;327]) compared to dogs with missense mutations at other codons in the NRAS and KRAS genes (n=5, MST 66 [CI 64;NA], p=1.58E-05) and compared to dogs without missense mutations in the RAS family of genes (n=180, MST 123 [CI 100;149], p=0.036). These preliminary findings of similarities between canine and human tumors and associations between mutated codons and prognosis in our growing dataset create an important opportunity to accelerate cancer research, including clinical and biopharmaceutical studies, with great potential benefit to both dogs and humans. Citation Format: Michelle E. White, Garrett Harvey, Lucas Rodrigues, Chase Schwalbach, Dorothy Girimonte, Aubrey Miller, Abigail Hull, Lindsay Lambert, Christina Lopes, Gerald Post. Canine hemangiosarcoma as a model for RAS-mutated human cancers: Preliminary data [abstract]. In: Proceedings of the AACR Special Conference: Targeting RAS; 2023 Mar 5-8; Philadelphia, PA. Philadelphia (PA): AACR; Mol Cancer Res 2023;21(5_Suppl):Abstract nr A005.