Abstract B091: Spatial Transcriptomics Unveils Intraductal Papillary Mucinous Neoplasm Heterogeneity: From Novel Clusters to Immune Dynamics

Abstract Intraductal papillary mucinous neoplasms (IPMNs), a common precursor to pancreatic ductal adenocarcinoma, present as a spectrum of benign indolent behavior and progression to invasive carcinoma, which highlights the importance of identifying biomarkers of IPMN progression. This study utilized whole transcriptome spatial transcriptomics to define the molecular characteristics of Epithelial, Immune, and Fibroblast cell populations of the three distinct histological sub-types of IPMN to understand the dynamics between the cell types and identify biomarkers of neoplastic progression. Fifty-two surgically resected IPMN cores from 21 patients with resected IPMNs comprising 24 Gastric, 14 Intestinal, and 14 Pancreaticobiliary with a distribution of low-grade dysplasia (LGD), high-grade dysplasia (HGD), and Invasive carcinoma (IC) were analyzed. Epithelial, Immune, and Fibroblast components were analyzed using the GeoMx spatial profiler. Unsupervised hierarchical clustering of epithelial areas of interest (AOI) identified 3 distinct clusters demonstrating a spectrum of molecular phenotypes in IPMN that were correlated with the three histological subtypes. Comparing the gene expression profile based on the shared genes expressed on all three subtypes, the Intestinal type tends to have an overexpression of genes whereas the Pancreaticobiliary represents a repressed phenotype relative to the Gastric phenotype. Supervised clustering using Moffitt’s classifier to identify the Epithelial-Mesenchymal transition demonstrates the Pancreaticobiliary subtype to be more Basal-like versus the Gastric and Intestinal inclined towards the classical phenotype. Supervised clustering based on the degree of dysplasia within the epithelial compartments of each of these subtypes, demonstrated a spectrum of molecular heterogeneity and identified biomarkers specific to High-Grade dysplasia in the Gastric and Intestinal subtype. Immune AOI deconvolution analysis demonstrated that Tregs were associated with Pancreaticobiliary, Macrophage enrichment associated with the Intestinal phenotype, while there is enrichment of CD8 T-cell and CD4 T-cell gene signatures associated with the Gastric subtype. Fibroblast AOI analyses demonstrated heterogenous myCAF- iCAF signatures between the subtypes. In summary, Spatial whole transcriptome analysis of IPMN identifies molecular signatures in epithelial and immune compartments that differentiate between the three histological subtypes of IPMN and helps understand their progression from Low-grade dysplasia to High-grade dysplasia. Citation Format: Amaya Pankaj, Michael J Raabe, Yuhui Song, Bidish K Patel, Katherine Xu, Joshua R Kocher, Nicholas J Caldwell, Maria L Ganci, Linda T Nieman, Mari Mino-Kenudson, Vikram Deshpande, Nabeel Bardeesy, Carlos Fernandez-Del Castillo, Martin J Aryee, David T Ting, Yasmin G Hernandez-Barco. Spatial Transcriptomics Unveils Intraductal Papillary Mucinous Neoplasm Heterogeneity: From Novel Clusters to Immune Dynamics 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 Suppl₂): Abstract nr B091.