Abstract 6492: KRAS inhibitor-based combinations in pancreatic cancer identified through Optim.AI

Abstract PURPOSE: KRAS is the most frequently mutated oncogene in almost all cancers, including pancreatic ductal carcinoma (PDAC) and colorectal cancer. Increased development efforts targeting KRAS mutations have yielded recent FDA approvals of two KRAS G12C inhibitors. However, this is beneficial for only a small proportion of patients especially with indications like PDAC, where G12C mutations are rare (1%) while G12D mutations account for approximately 45%. Hence, there is a need to develop methods to target a broader range of KRAS mutations. Beyond direct gene mutation, activation of oncogenic KRAS and downstream or parallel signaling through other genomic alterations may benefit from KRAS-specific therapeutic strategies, especially combination therapy. In this study, Optim. AI™, a hybrid computational-experimental platform is harnessed to identify mutation-specific optimal therapies and to stratify efficacious drug combinations containing KRAS inhibitors in pancreatic cancer. METHODS: Optim. AI™ utilizes small datasets to rationally converge upon optimal drug combinations within a defined drug search space. By mapping experimental data points to a second-order quadratic function, Optim. AI™ predicts cell killing efficacies for all possible combinations, independent of the background of the models utilized. In this study, pancreatic cancer cell lines, with varying KRAS mutations, were tested with an array of 155 different combinations consisting of 12 drugs at varying concentrations. The drugs include KRAS G12C inhibitors, Sotorasib and Adagrasib, KRAS inhibitors in development, and approved therapies used clinically for PDAC. Post-drug treatment viability was measured and used for Optim. AI™ analysis, to rank and compare the top therapies across the different cell lines. RESULTS: Through Optim. AI™ analysis, we observed a range of sensitivities towards KRAS inhibition, even for both Sotorasib and Adagrasib. Compared against HT-29 cell line, harboring wild type KRAS, we have identified distinct combinations of KRAS inhibitors with standard of care drugs on cell lines with G12D, G12V and G12C KRAS mutants. Subsequent validation of the top and bottom ranked combinations for the respective cell lines correspond to the predicted Optim. AI™ results. CONCLUSIONS: This study has demonstrated the differential sensitivities of the KRAS inhibitors across several pancreatic cancer cell lines, with different KRAS mutational profiles. Notably, we have identified KRAS inhibitor-based combinations that are distinctly different from wild type KRAS cell line. These preliminary findings could be used to drive precision medicine by identifying suitable biomarkers to potentially stratify pancreatic cancer patients and improve treatment outcomes. Future work includes evaluating the clinical significance of these combinations by validating them on patient-derived organoids. Citation Format: Jhin Jieh Lim, Edward Kai-Hua Chow, Masturah Rashid. KRAS inhibitor-based combinations in pancreatic cancer identified through Optim. AI abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 6492.