Abstract 1982: Genomic analyses of patient-derived xenografts reveal novel molecular subsets of PDAC and their association with clinical features.

Abstract Pancreatic ductal adenocarcinoma (PDAC) is a dismal disease with limited therapeutic options. We have established a large-scale RNA-seq dataset comprised of >120 PDAC patient-derived xenograft (PDX) models and >300 bulk RNA-seq specimens (when counting biological replicates). This cohort includes diverse PDAC clinical manifestations, such as patients with particularly low survival (<8 weeks) that are under-represented in other large-scale efforts (e.g. TCGA). Our analysis identified two transcriptional programs with a highly heterogeneous expression among the basal-like models: a unique partial-Epithelial-to-Mesenchymal program (Basal-Mesenchymal) and an epithelial senescence-associated program (Basal-EpiSen), reminiscent of the pan-cancer epithelial senescence-associated program that we recently described in other epithelial malignancies. Notably, high expression of the Basal- Mesenchymal program is strongly associated with low-survival and with models established from patient's tumors at terminal stage. Next, we compared representative Basal- Mesenchymal and Basal-EpiSen PDAC cell lines across the comprehensive Cancer Cell Line Encyclopedia (CCLE) drug response dataset. Our initial results show differential drug sensitivity and gene dependencies between the Basal- Mesenchymal and Basal-EpiSen subsets, highlighting their potential clinical significance. We show that the Basal-like subtype diverges into two distinct clinically-relevant subsets. Taken together, our work provides an unprecedented resource towards improving PDAC molecular classification, an essential step for PDAC precision medicine. Citation Format: Chani Stossel, Rotem Tal, Rouven Hoefflin, Maria Raitses-Gurevich, Dikla Atias, Yulia Glick Gorman, Gali Altman, Tamar Beller, Talia Golan, Itay Tirosh. Genomic analyses of patient-derived xenografts reveal novel molecular subsets of PDAC and their association with clinical features [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1982.