Thrombosis is a common and highly morbid complication of cancer. Tumor type is among the best currently available predictors of cancer-associated thrombosis and pancreatic ductal adenocarcinoma (PDAC) is among the most highly thrombogenic cancers. Our underlying premise is that factors elaborated by the tumor itself drive thrombus formation in cancer. From this assertion arise specific hypotheses that form the basis of our project. (1) Evaluation of plasma proteins from patients with cancer can accurately predict cancer-associated thrombosis. To evaluate this hypothesis, we will perform a high throughput proteomic analysis of large cohorts of cancer patients comparing plasma proteins in patients who subsequently develop clots compared to those who remain clot free. This project utilizes novel proximity extension assay technologies. (2) If the tumor drives clot formation in cancer, then the sporadic nature of cancer thrombosis could be explained by the heterogeneity of tumors even within a single type of cancer. We predict that PDACs derived from different patients will show variation in their ability to elaborate prothrombotic factors in keeping with their molecular subclassification. To evaluate this prediction, we will use patient-derived organoids, which retain the genetic and phenotypic signatures of the primary tumor in order to perform in-depth analysis of individual tumors relative to prothrombotic potential in vitro and in vivo. (3) The fact that thrombosis is more common in aggressive, advanced stage cancers indicates that tumor progression enhances the thrombogenicity of tumors. We will evaluate this premise at the molecular level by testing the hypothesis that activation of the unfolded protein response (UPR) contributes to thrombus formation in PDAC. Specifically, we will determine whether UPR signaling in PDAC leads to the elaboration of prothrombotic factors such as pancreatic-specific protein disulfide isomerase, tissue factor, and prothrombotic microparticles. These studies leverage the unique resource of the BIDMC Pancreatic Disease Registry and Biorepository, which includes a >150 patient-derived xenograft models. These PDAC models are fully curated with associated patient history, transciptome and proteome data, and correlated plasma samples. The proposed studies are of substantial clinical significance, since they will discover and validate biomarkers using large clinical cohorts with the goal of identifying which patients will benefit most from aggressive thromboprophyhlaxis. These experiments will also enhance our fundamental understanding of how pathways leading to cancer progression such as the UPR contribute to the prothrombotic phenotype of PDAC. The utility of interfering with protein disulfide isomerase (PDI) as a mechanistic link between cancer progression and thrombosis will be evaluated using samples from our phase II/III trial of evaluating a small-molecule inhibitor of PDI activity in advanced cancer patients. Thus, these studies have important and immediate implications for both basic knowledge and treatment of cancer- associated thrombosis.
Blood clots are a common complication of cancer and are associated with significant suffering and increased likelihood of death. Our studies will use cutting-edge proteomic technology to identify markers of cancer- associated thrombosis allowing caregivers to identify cancer patients at greatest risk of thrombosis. These studies will also evaluate novel culture techniques and thrombotic pathways to identify factors that predispose pancreatic cancer patients to blood clots.
