In Situ Multi-Modal Characterization of Pancreatic Ductal Adenocarcinoma Reveals Tumor Cell Identity As a Defining Factor of the Surrounding Microenvironment

Abstract Pancreatic Ductal Adenocarcinoma (PDAC) is heterogeneous with low tumor purity, prominent microenvironment and complex architecture, which preclude identification of shared tumor intrinsic biology within and across patients. We overcame these challenges by applying complementary spatial omics approaches – providing necessary resolution and context – to primary untreated PDAC tumors from 39 patients capturing 341,949 low-bulk and 531,718 single-cell spatial transcriptomes. Of these, 59,403 low-bulk profiles and 205,665 single-cells represented tumor cells. We leveraged this data to build on prior reports of tumor cell subtypes in PDAC. We identify 5 distinct malignant subtypes, excluding ductal-like and intraepithelial neoplasia cells, that span the classical-to-basal spectrum. One tumor subtype is highly proliferative and enriched for the DNA synthesis-condensation-mitosis transcriptional network suggesting this subset disproportionately contribute to tumor growth. Strikingly, the spectrum of tumor cell subtypes was present within all patients, suggesting that subtype-based treatment stratification may need refinement. We find that each subtype has its own distinct regulators and histology. The classical subtype has extensive cell intrinsic regulation, higher cell density and lower extracellular matrix content. In contrast, the basal subtype phenotype results from a combination of cell intrinsic and tumor microenvironment (TME) interactions, wherein the basal subtype has a lower cell density, is surrounded by activated stroma and dense collagen matrix, and is poised for hypoxic adaptation. We derive the composition of cellular neighborhoods providing an explanation for the complex architecture seen in PDAC. Tumor cell neighborhoods stratified based on which side of the classical-to-basal spectrum they fell. For example, cells towards the classical-side of the spectrum had more homogenous tumor neighborhoods that generally lacked intermediate and basal tumor cells. In contrast, cells towards the basal-side of the spectrum were likely to contain classical neighbors. More broadly, neighborhoods along this continuum also contained different proportions of immune and stromal cells, cell densities, collagen deposition, and TME adaptation. These observations support a model where tumor subtypes exist within disparate niches. Lastly, our atlas unifies the catalog of previously reported stromal cells (fibroblasts and stellate cells versus myofibroblastic and inflammatory cancer associated fibroblasts (CAF)) and describes a novel interferon stimulated gene (ISG) resistance signature (ISG.RS) fibroblast. We map the location of these various stromal cells relative to tumor subtypes. For example, myofibroblast CAFs and ISG.RS fibroblasts are enriched near the basal tumor subtype whereas inflammatory CAF abundance increases with distance away from tumor cells. In sum, our atlas provides a lens for stratifying tumor complexity in a cancer cell centric manner and provides a path for testing therapeutic hypothesis in the right cell subtype and context. Citation Format: Brian Rabe, Anna Lyubetskaya, Andrew Kavran, Yulong Bai, Andrew Fisher, Hannah Pliner, Alba Font-Tello, Anne Lewin, Ruifeng Hu, Alexandre P Alloy, Yelena Cheng, Chao Dai, Yunfan Fan, Constance Brett, Todd Brett, Lauren Giampapa, Soeren Stahlschmidt, Fayaz Seifuddin, Steven Vasquez Grinnell, Daniel Carrera, Carlos Rios, Tom Lila, Pradeep Kar, Abhishek Shukla, Rachael Bashford Rogers, Lara Heij, Mike Mason, Ryan Golhar, Isaac Neuhaus, Enas Abu Shah, Shivan Sivakumar, Jimena Trillo Tinoco, Benjamin J Chen, Konstantinos J Mavrakis, Eugene Drokhlyansky. In situ multi-modal characterization of pancreatic ductal adenocarcinoma reveals tumor cell identity as a defining factor of the surrounding microenvironment [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research; 2024 Sep 15-18; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl_2):Abstract nr B081.