Abstract B021: Pan-RAS IMM-1-104 activity in humanized 3D tumor models is independent of specific amino acid substitution

Introduction: Novel dual-MEK inhibitor IMM-1-104 is under clinical investigation for use in RAS-addicted solid tumors. Approved KRAS G12C inhibitors are available but cover a limited subset of patients. For example, the KRAS G12C substitution occurs in only 1-3% of pancreatic cancers. We assessed response to IMM-1-104 across RAS mutant preclinical models to determine whether a preference was observed for mutation position, or for specific amino acid substitutions. Experimental Procedures: Response to IMM-1-104 was measured in a humanized 3D tumor growth assay across 133 tumor models. Sixty-nine of these models have a reported RAS mutation, and all models are being mutationally profiled by whole exome sequencing, with the majority (~75%) completed to date. The RAS-mutant panel spans 11 tissue types and includes a subset of 30 confirmed KRAS G12 mutated cell lines drawn from three major indications: 12 pancreatic, 11 lung, and 7 colorectal cancer models. Based on the 3D assay, cell lines were classified into sensitive, intermediate, and resistant to IMM-1-104. The distribution of responses was then assessed across mutation position and amino acid substitutions. Summary of New Data: In the full dataset across 69 RAS-mutant models, at least one model displayed response to IMM-1-104 (sensitive or intermediate) for each observed mutation in K/N/HRAS. That is, no particular mutation position or amino acid substitution was exclusively found to be resistant. Association of response with amino acid identity was further evaluated in a subset of lines for the most frequently altered residue in KRAS, G12. A distribution of responses was observed for each amino acid substitution. The G12 substitutions having at least four cell lines each total across the three indications included G12C (8 lines), D (5 lines), R (4 lines), and V (11 lines). Across cell lines for each of these substitutions, multiple response categories were observed. In each case, half or more lines fell into the intermediate category with the rest falling into one or both of the other response categories. For example, out of the 8 KRAS G12C lines, 6 showed intermediate response, 1 showed resistance, and 1 showed sensitivity. Examining these distributions together, no significant statistical relationship was seen between the amino acid substitution and response categories by Fisher’s exact test. Conclusions: Across 69 RAS-mutated cell lines, each mutation position or amino acid substitution was associate