Pancreatic cancer ranks as the third leading cause of cancer-related deaths in the United States, with a dismal 5-year survival rate of only 13%. This poor prognosis is largely due to the challenge of early detection, which is crucial for effective treatment. Given the low prevalence of pancreatic ductal adenocarcinoma (PDAC), widespread screening is impractical and economically unfeasible. We present a novel, multiplex blood-based assay aimed to improve early detection for PDAC in high-risk populations.

Utilizing comprehensive proteomic data, we established a mass spectrometry-based assay that simultaneously detects a panel of five protein biomarkers (APOA4, SERPING1, LRG1, C3, ORM1) in whole plasma. Machine learning was applied to construct a composite biomarker readout, and a “direct sampling” approach was employed to establish a reference threshold for PDAC detection. Characterized through Receiver Operating Characteristic (ROC) analysis with case-control cohorts from multiple medical centers, the assay’s efficacy was further evaluated in high-risk groups.

The assay achieved an area under the curve (AUC) of 0.87 and an overall classification accuracy of 75% in distinguishing early-stage PDAC cases from controls. Its efficacy was further assessed in high-risk groups, including adults aged 50 and older with new-onset diabetes, individuals with familial risks, and patients with chronic pancreatitis. Across all high-risk groups, the assay demonstrated a positive predictive value (PPV) exceeding 18% and a diagnostic odds ratio (DOR) above 67 - sufficient to identify potential PDAC cases for further clinical evaluation.

This straightforward multiplex assay enables direct detection of early-stage PDAC from whole plasma without the need for immunodepletion, enrichment, or prefractionation, demonstrating high efficacy and strong clinical applicability. It significantly outperforms the FDA-approved CA19-9 in distinguishing early-stage PDAC from control samples and in detecting PDAC among high-risk populations. The improved performance is attributable to the diverse biological functions and pathways represented by the selected biomarkers, which are actively involved in PDAC development. Although the assay is not predictive in nature, its diagnostic strength stems from the functional relevance of these biomarkers to disease-associated processes. Overall, the novel assay demonstrates robust efficacy in identifying PDAC in high-risk groups.