Abstract A017: RAS inhibition and cytotoxic chemotherapy target complementary cell states in pancreatic cancer

Abstract Multiple studies have demonstrated preclinical activity of RAS inhibitors in models of pancreatic ductal adenocarcinoma (PDAC) and early results from ongoing clinical trials show promise. Following our earlier work using RMC-7977 (a preclinical tool compound related to the investigational RAS (ON) multi-selective inhibitor daraxonrasib) in a broad range of preclinical models, we elected to study the impact of RAS (ON) inhibition and standard-of-care (SOC) cytotoxic chemotherapies on the heterogeneity of malignant cells in PDAC. We performed single cell RNA sequencing (scRNAseq) on over three dozen PDAC tumors from the KPC genetically engineered mouse model, using different treatments and timepoints, yielding over a quarter million high quality single cell expression profiles. Consistent with prior studies, we found that RMC-7977 preferentially depleted more poorly differentiated malignant cells from KPC pancreatic tumors by one week of treatment. Residual malignant cells were well differentiated and showed hyperactivation of distinct sets of gastrointestinal and pancreatic progenitor transcription factors. Spatial transcriptomics on the same KPC tumors validated these findings and elucidated clear histological associations with RAS inhibitor treatment. This phenotype was also validated in human tissue explant models using surrogate immunohistochemical markers of cell states. We then decided to investigate the molecular, cellular, and preclinical consequences of combining RAS (ON) -Multi inhibitors with SOC chemotherapy agents. We first employed the well-validated OncoTreat algorithm to predict which PDAC malignant cells may be most susceptible to different SOC agents. Strikingly, we found that nearly all SOC agents were inferred to preferentially target more well-differentiated malignant cells. Indeed, we found that treatment of tumor-bearing KPC mice with gemcitabine + nab-paclitaxel (GnP) led to a depletion of well-differentiated malignant cell states, as measured by single cell regulatory network analysis. This led us to hypothesize that combining RAS inhibition with SOC agents might target complementary sets of malignant cell states, forming a rational basis for combining these agents. Indeed, preclinical intervention studies combining daraxonrasib with GnP showed combinatorial activity in a range of xenograft, syngeneic allograft, and patient-derived xenograft models of PDAC. Ongoing studies in the KPC model system will directly assess the impacts of these combination regimens on malignant PDAC cell states and will directly address the roles of cellular plasticity versus selective cell death in the modulation of cell state in response to RAS inhibition. Together these studies inform the rationale for the combination of RAS (ON) inhibition with cytotoxic chemotherapies in PDAC. Citation Format: Lorenzo Tomassoni, Alvaro Curiel-Garcia, Harika Gundlapalli, Melina Chen, Urszula Wasko-Kornberg, Ximo Pechuan-Jorge, Rashi Raghulan, Yongxian Zhuang, Kevin Contrepois, Steven A. Sastra, Carmine F. Palermo, Ida Aronchik, Jingjing Jiang, Andrea Califano, Mallika Singh, Kenneth P. Olive. RAS inhibition and cytotoxic chemotherapy target complementary cell states in pancreatic cancer abstract. In: Proceedings of the AACR Special Conference in Cancer Research: RAS Oncogenesis and Therapeutics; 2026 Mar 5-8; Los Angeles, CA. Philadelphia (PA): AACR; Cancer Res 2026;86 (5Suppl₁): Abstract nr A017.