LAUSR

Introduction Pancreatic ductal adenocarcinoma (PDAC) is the most common form of pancreatic cancer and is one of the deadliest cancer in the U.S., with a 5-year survival rate of just 11%. It is expected that ~62,000 new cases will be diagnosed in the US in 2023 and >70% of those will be at the regional or distant state, when curative resection is no longer an option. It is imperative that tools enabling early disease detection (stages I and II) are developed to improve patient survival outcomes. Here, we report the results of a study on the utility of extracellular vesicles (EVs) for early PDAC detection. Methods Our case-control training set was comprised of 75 pathologically confirmed PDAC cases (Stage I = 31, Stage II = 44) and 640 controls (including 11 subjects with pancreatitis) for EV biomarker and algorithm selection via machine learning. EVs from plasma were isolated using a proprietary technology, then EV-bound proteins of interest were analyzed via an immunoassay. A second independently collected cohort of 20 confirmed PDAC cases (Stage I = 10, Stage II = 10) and 27 controls (including 9 patients newly diagnosed with diabetes) was used for validation. Results The machine learning analysis using EV-isolated proteins generated a signature that utilized 8 different biomarkers, with an AUC of 0.989 (95% CI: 0.980 - 0.998) in the training set and an AUC of 0.987 (CI: 0.964 - 1.000) in the validation set. For the validation set, the sensitivity was 95% (CI: 76.4% - 99.1%) and the specificity was 92.6% (CI: 76.6% - 97.9%), with 10 out of 10 stage I cases and 9 out of 10 stage II cases correctly classified. Conclusions The ability to alter the future outcomes of PDAC patients will be predicated on development of a new generation of early-detection biomarkers and technologies. The use of EV protein signatures, a new class of cell-free biomarker, permits detection at high sensitivity and specificity, as demonstrated for the independently collected validation cohort. Studies aimed to assess this test for high-risk patients (family history of PDAC, known germline mutations, precursor lesions, hereditary pancreatitis, new onset diabetes will continue towards the establishment of real-world evidence. Citation Format: Juan P. Hinestrosa, Harmeet Dhani, Gregor Schroeder, Jean M. Lewis, Heath I. Balcer, Razelle Kurzrock, Dove Keith, Rosalie Sears, Paul Billings. Pancreatic ductal adenocarcinoma (PDAC) early detection. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 6515.