Abstract Pancreatic ductal adenocarcinoma (PDAC) arises from precursor lesions, the most common of which are Pancreatic Intraepithelial Neoplasia (PanIN). PanINs are microscopic and cannot be detected in live individuals; as such, PanIN characterization has occurred prevalently in the context of cancer-bearing pancreata (adjacent normal) rather than in healthy tissue. Recently, through a unique collaboration with Gift of Life Michigan, we showed that PanIN prevalence is over 60% across age groups in donor healthy tissues, with high prevalence even in young individuals. Given PanIN prevalence and the rarity of pancreatic cancer, we hypothesized that the majority of PanIN is not fated to progress to malignancy. We postulated that a combination of neoplastic cell-intrinsic factors and components of the microenvironment controls progression to malignancy. We utilized spatial transcriptomics using the Visium platform (n = 14) and scRNASeq (n = 45) to dissect gene expression profiles and characterize spatial domains in donor healthy, adjacent-normal, and tumor tissues. We developed a spatial domain identification workflow that integrates spatial transcriptomics with the scRNASeq atlas, leveraging cell-type deconvolution, spatially informed clustering and pathology annotation. This approach allowed us to study the co-evolution of epithelial and stromal cells over the progression of pancreatic cancer. We were able to delineate distinct epithelial spatial domains—including Acinar, Ductal, and PanIN structures—in both donor and tumor tissues, as well as tumor-specific regions such as glandular tumor and poorly differentiated tumor. Additionally, we identified key stromal niches, including immune-rich regions such as tertiary lymphoid structures (TLS), plasma cell clusters, and macrophage-enriched zones, alongside non-immune fibrotic domains. We began by analyzing the epithelial domains and observed that while all acinar and ductal domains formed transcriptionally distinct clusters, neoplastic cells formed a continuum from sporadic healthy donor-associated PanINs to well- and poorly-differentiated tumors. We identified key genes that exhibit expression changes throughout the progression trajectory. Secondly, we analyzed characteristics of the stroma surrounding sporadic lesions, well-differentiated, and poorly-differentiated tumors. We observed key differences in fibroblasts between healthy donor-PanIN and tumorigenic regions, as fibroblasts surrounding donor-associated PanIN lacked known cancer-associated fibroblast (CAF) markers such as Alpha-Smooth Muscle Actin (α-SMA). We also observed distinct differences in immune cells, where there were more abundant plasma and CD8+ T-cells surrounding donor-associated PanIN, while tumorigenic regions showed macrophage enrichment. We then performed immunostaining and confocal imaging to validate the defined markers and specific cell populations. Our work defines the PanIN microenvironment as distinct from that of PDAC and maps the co-evolution of cancer cells and stroma over the course of carcinogenesis. Citation Format: Ahmed M. Elhossiny, Jude Okoye, Padma Kadiyala, Alexander Bray, Jamie Mills, Hannah Watkoske, Yaqing Zhang, Jiaqi Shi, Arvind Rao, Elana J. Fertig, Eileen S. Carpetner, Timothy Frankel, Marina Pasca di Magliano. Spatially Resolved Insights into epithelial-stromal co-evolution in pancreatic ductal adenocarcinoma abstract. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research—Emerging Science Driving Transformative Solutions; Boston, MA; 2025 Sep 28-Oct 1; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2025;85 (18Suppl₃): Abstract nr A052.
Building similarity graph...
Analyzing shared references across papers
Loading...
Elhossiny et al. (Sun,) studied this question.
www.synapsesocial.com/papers/68da58dcc1728099cfd113ae — DOI: https://doi.org/10.1158/1538-7445.pancreatic25-a052
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:
Ahmed M. Elhossiny
Jude Ogechukwu Okoye
Padma Kadiyala
Cancer Research
University of Michigan
University of Maryland, Baltimore
Building similarity graph...
Analyzing shared references across papers
Loading...