Abstract 6092: Three-dimensional multi-omic analysis of early invasion of human pancreatic ductal adenocarcinoma from IPMN

Abstract

Abstract Background: Pancreatic ductal adenocarcinoma (PDAC) is a deadly cancer, with only 11% of patients surviving beyond 5 years. Intraductal papillary mucinous neoplasm (IPMN), a precancerous lesion and the most common type of neoplastic pancreatic cyst, presents a critical opportunity for cancer interception, but the drivers of malignant progression in IPMN are still largely unknown. Although the immunosuppressive microenvironment of PDAC has been well characterized, the timing of the development of these immune alterations in IPMNs is not well studied and could provide a rational foundation for cancer immuno-prevention. Moreover, the inception of invasive PDAC and its spread into the microenvironment from IPMN is complex, multifactorial, and, due to its unique anatomy, challenging to analyze with traditional 2-dimensional visualization. We sought to capture the precise point of transition from IPMN to invasive carcinoma utilizing CODA, a supervised deep learning tool for three-dimensional (3D) reconstruction of serially sectioned human tissue, in order to quantitatively assess the molecular and cellular alterations associated with malignant progression. Method: Formalin-fixed paraffin-embedded (FFPE) tissue blocks with PDAC arising from IPMN were serially sectioned, every third slide was stained with H&E, and digitized at 20x magnification. Following pathologist-guided annotations on a subset of H&E slides, CODA generated 3D models of each tissue block, including automated annotation of 9 pancreatic tissue components. Using these annotations, we identified the transition from IPMN to invasive carcinoma in each sample and selected regions for multi-omic profiling based on quantitative features of the cellular microenvironment. Multi-omic profiling of IPMN, transition zone, and PDAC included laser capture microdissection followed by whole exome sequencing to identify somatic DNA alterations, as well as spatial transcriptomics to identify alterations in gene expression in neoplastic cells and spatial proteomics to identify cellular alterations in the surrounding microenvironment. Results: CODA generated accurate annotation on all the H&E slides from each case and robustly identified regions of interest for multi-omic profiling. Spatial proteomic profiling using imaging mass cytometry revealed a decrease in T cell density as IPMNs transition into PDAC. Sub- clustering of the T cell compartment identified decreases in multiple T cell subsets, including activated cytotoxic T cells and helper T cells, in the transition zone compared to IPMN. Conclusion: In this study, we integrated multi-omic profiling with CODA-generated high-resolution 3D tissue maps to identify molecular and cellular drivers of malignant progression in human PDAC. Citation Format: Shalini Datta, Sarah M. Shin, Jessie Kanacharoen, Michael Johannes Pflüger, Katsuya Hirose, André Forjaz, Sarah Graham, Pei-Hsun Wu, Ralph H. Hruban, Denis Wirtz, Won Jin Ho, Ashley L. Kiemen, Laura D. Wood. Three-dimensional multi-omic analysis of early invasion of human pancreatic ductal adenocarcinoma from IPMN [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 6092.