Independent blinded validation of a novel blood-based early pancreatic cancer detection test.

4166 Background: Pancreatic ductal adenocarcinoma (PDAC) frequently presents at late stages, resulting in increased mortality. Early detection is crucial for enhancing survival rates; however, available screening methods are limited. This gap underscores the critical need for advancements in early PDAC detection. To address this, Wholomics developed a proprietary early detection assay, leveraging molecular signatures found in the blood, and has further validated its efficacy through a comprehensive blinded validation study. Methods: Wholomics conducted retrospective analysis of serum samples from the Amsterdam UMC Liquid Biopsy Center and the Bank of Biological Material of the Masaryk Memorial Cancer Institute. A training cohort included 290 control subjects, 12 patients with a confirmed diagnosis of chronic pancreatitis, and 71 patients with a histopathologically confirmed diagnosis of PDAC (20 stage I, 30 stage II, 13 stage III, 8 stage IV). A blinded validation cohort comprised samples from 20 control subjects, 24 patients with non-cancerous conditions (5 with benign tumors, 10 with IPMNs, 9 with chronic pancreatitis), and 60 patients with histologically confirmed PDAC. Peripheral blood samples of cancer patients were drawn before surgery or treatment. Wholomics used its proprietary NMR-based technology to quantitatively analyze serum samples. Disease-specific molecular signatures were identified and used to develop a computational biomarker specific to PDAC. A comparative analysis via 10-fold cross-validation and a blinded validation were conducted. The accuracy, specificity and sensitivity are reported. Results: In this study, a dual-cohort design, comprising both a training and a blinded validation cohort, was utilized to assess the capability of the Wholomics technology in detecting cancer-specific molecular signatures. In the training cohort, a 94.6% accuracy was achieved in distinguishing PDAC patients from non-cancer controls. A specificity of 95.7% and sensitivity of 90.7% were reached. A similar performance was observed in the blinded validation cohort, where the blood-based test accurately differentiated PDAC patients from controls with an accuracy of 92.5%. In more detail, a specificity of 95% (noting a single false positive) and a sensitivity of 91.7% were achieved. Conclusions: The collective findings of this research indicate that utilizing disease-specific molecular markers in peripheral blood offers a highly accurate method for differentiating between PDAC patients, healthy subjects, and those with pre-cancerous diseases. Should future studies across broader cohorts corroborate these results, this testing approach has potential for consideration as a screening mechanism for PDAC, aiming to facilitate early detection and consequently enhance patient prognoses.