Abstract B015: Leveraging patient-derived organoids and matched 2D models to elucidate resistance mechanisms to KRAS inhibition in pancreas cancer

Abstract Pancreatic ductal adenocarcinoma (PDAC) remains one of the most lethal malignancies, with limited therapeutic options and nearly universal KRAS mutations (92%). The convoluted molecular profile of PDAC, marked by coincident inactivation of tumor suppressors (TP53, CDKN2A, SMAD4), absence of recurrent targetable alterations, and striking transcriptional heterogeneity, complicates therapeutic intervention. Despite progress in KRAS-directed strategies, including allele-specific (G12C, G12D), pan-KRAS and multi-RAS inhibitors, as well as degraders and vaccines, long-term efficacy is challenged by the emergence of acquired resistance. Purpose: This study leveraged a large biobank of Human patient-derived Pancreatic Organoids (HPOs) and matched 2D models to: (i) characterize heterogeneity of responses to distinct KRASi, (ii) dissect genetic and non-genetic resistance mechanisms, and (iii) inform rational combinatorial strategies. Methods: 60 HPOs and matched 2D lines were profiled using MRTX1133 (G12D-selective), AMG510 (sotorasib, G12C-selective), BI-2865 (pan-KRAS (OFF) ), and RMC-7977 (RAS (ON) multi-selective). Comparative studies of culture conditions and media assessed the impact of extrinsic cues on KRASi response. Acquired resistance to distinct KRASi classes was generated in 40 models spanning diverse KRAS alleles. Whole-exome sequencing (WES), single-cell RNA sequencing (scRNA-seq), multiplex signaling analyses, and a 60-compound targeted combination screen were performed. Results: Baseline responses were highly heterogeneous. Loss of the wild-type KRAS allele increased KRASi sensitivity, whereas CDKN2A, SMAD4, or PDAC subtype showed no predictive value. scRNA-seq revealed marked lineage plasticity in vitro with dynamic shifts among classical, basal-like, and mesenchymal states displaying variable KRAS dependency. Growth factor-rich media modulated oncogenic signaling and reduced KRASi and MEKi sensitivity. Resistant models converged toward mesenchymal identity with activation of epithelial-mesenchymal transition (EMT), extracellular matrix (ECM) remodeling, and IFN-γ signaling, accompanied by reduced KRAS dependency with partial reversibility after drug withdrawal. WES identified a broad spectrum of resistance-associated changes, including KRAS or MYC amplification, ABCC1 mutation, alterations in RTK, MAPK, PI3K/mTOR genes, WNT/β-catenin pathway genes, as well as co-occurring alterations in genes involved in ECM, epigenetic, and metabolic processes. An ongoing combination screen has identified sensitivities to YAP-TEAD, SRC, FGFR/FAK, and dual-mTOR inhibition in resistant lines. Conclusions: The genomic complexity and transcriptional plasticity of PDAC jointly drive heterogeneous responses to KRASi. Resistance reflects convergence of genetic events and adaptive processes, with EMT-driven lineage plasticity as a hallmark. Integration of organoid-based platforms with multi-omic profiling uncovers tumor-intrinsic resistance mechanisms and informs strategies to achieve durable KRAS pathway suppression. Citation Format: Nicolas Lecomte, Elias-Ramzey Karnoub, Emily M. Kang, Florencia Velez-Cortes, Lori Kim, Jerry P. Melchor, Christine A. Iacobuzio-Donahue. Leveraging patient-derived organoids and matched 2D models to elucidate resistance mechanisms to KRAS inhibition in pancreas 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 B015.