Spatial analysis of Intraductal Papillary Mucinous Neoplasms reveals secretory cell-enriched neighborhoods
Cephas, A. T.; Jarvis, B.; Gell, K.; Taranto, C. P.; Batardiere, M.; Sapon-Cousineau, S.; Dean, E. D.; Singhi, A. D.; Tan, M. C. B.; Trinh, V. Q.; DelGiorno, K. E.
Show abstract
Pancreatic ductal adenocarcinoma (PDAC) is currently the third leading cause of cancer-related deaths in the United States. Intraductal papillary mucinous neoplasms (IPMNs) are neoplastic lesions of ductal origin that seed 10-25% of PDAC. There are currently no markers that distinguish between IPMN that will remain benign and those that will progress to cancer. A heterogenous population of secretory cells, including chemosensory tuft cells and hormone-expressing enteroendocrine cells (EECs), form during metaplasia and neoplastic progression in the pancreas, but the relevance of these populations as it relates to IPMN progression is not well characterized. Here, we performed spatial transcriptomics as well as multiplex immunostaining and spatial statistics on surgically resected IPMN from 60 patients to characterize these populations in all subtypes (gastric foveolar, intestinal, pancreatobiliary) and grades (low-grade, high-grade, invasive). We found that POU2F3+ tuft-like cells, CHGA+ EECs, and a subset of pancreatic endocrine cells ([a] and {gamma} cells) were present in all types of IPMN. Further, serotonin-expressing enterochromaffin cells made up the bulk of EECs in low-grade disease. Enterochromaffin, tuft-like, and glucagon-expressing alpha cells were not evenly distributed and instead were significantly enriched in a spatial manner, which is overlooked using conventional whole tissue quantification approaches. Tuft-like cell clusters were enriched with monocytes and resident memory T cells and anti-correlated to activated fibroblasts (myCAFs, iCAFs). Overall, these secretory cell clusters may reflect clonal expansion resulting in formation of distinct stromal niches with unknown consequences for disease progression.
Matching journals
The top 9 journals account for 50% of the predicted probability mass.