OVERALL - ABSTRACT Pancreatic ductal adenocarcinoma (PDAC) is the 3rd most common cause of cancer-related mortality in the United States, with the overhwhelming majority of patients presenting advanced stage disease. Invasive neoplasia in the pancreas represents the culmination of a multistep progression that begins with non-invasive precursor lesions, which remain an untapped “window of opportunity” for early detection and cancer interception. Two major histological subtypes of precursor lesions are recognized - the more common non-cystic pathway, represented by pancreatic intarepithelial neoplasia or PanIN lesions, estimated to precede ~90% of PDAC, and the cystic pathway, most commonly represented by intraductal papillary mucinous neoplasms or IPMNs, accounting for the remaining 10%. While members of this team have played a seminal role in characterizing the histology and genetics of PanINs and IPMNs, much remains to be elucidated in terms of the molecular dependencies that sustain early pancreatic neoplasia, and how signaling cues from these early lesions reprogram the “precursor microenvironment” (PME), including “precursor-associated fibroblasts” (PAFs). The goal of our Tri-state Pancreatic Adenocarcinoma TBEL (Tri-PACT) Center - incorporating UT MD Anderson Cancer Center (UTMDACC), University of Michigan (UMich) and Johns Hopkins University (JHU) - is to create a collaborative and integrated U54 center to conduct basic and translational studies in early pancreatic neoplasia. The Tri-PACT Center will be led by Dr. Anirban Maitra (UTMDACC) and Co-PI Dr. Marina Pasca di Magliano (UMich). The title of our Tri-PACT Center proposal is “Tumor Microenvironment Crosstalk Drives Early Lesions in Pancreatic Cancer”, and we are proposing three projects (two basic, one translational), each of which will be supported by a Multiscale Computational Oncology Research Core (M-CORE) and an Administrative Core (AC). Project 1 (basic) will study the functional requirement of a pivotal cytokine, interleukin IL-33, which is induced in the PAF and epithelial compartments of PanINs and IPMNs in response to KRAS and GNAS mutations, respectively, in disease progression and reprogramming of the PME. Project 2 (basic) will study a unique metabolic “synthetic essentiality” centered on mitochondrial quality control created in cystic precursors that harbor loss of RNF43, a E3 ubiquitin ligase lost in ~50% of IPMNs. Notably, the Tri-PACT investigators have developed genetically engineered models (GEMs) of pancreatic preneoplasia that recapitulate the cognate human lesions, and will be extensively leveraged in the two basic projects, with cross-species validation in patient-derived preclinical models. Project 3 (translational) will deploy a unique 3D reconstruction tool (CODA) paired with multi-region sequencing of human precursor lesions to map the evolutionary trajectory of individual precursors at an unprecedented resolution, and correlate subclonal architecture with high dimensional analysis of the immune and PAF composition within the PME. Cumulatively, these projects will enhance our understanding of the drivers of early pancreatic neoplasia, and a seedbed for early detection approaches.

OVERALL – PROJECT NARRATIVE Improved understanding of the biology and heterogeneity of precursor lesions, including the interactions that these lesions harbor with the surrounding microenvironment, are a prerequisite for improving outcomes in pancreatic cancer, the third most common cause of cancer related deaths. The TriState Pancreatic Adenocarcinoma TBEL (Tri-PACT) Center brings together a collaborative group of investigators from three institutions who will use credentialed animal models and patient-derived materials to pursue three highly interdependent projects on basic and translational research in pancreatic cancer precursor lesions.