Shotgun Label-Free Proteomic Analysis for Identification of a Potential Diagnostic Biomarker for Pancreatic Cancer.

Background: Pancreatic cancer (PC) is one of the cancers with poor prognosis. Although carbohydrate antigen (CA) 19-9 is currently the most specific and sensitive biomarker for the diagnosis of PC and, therefore, the most frequently used, approximately 34% of PC patients may have a Lewis antigen-negative phenotype and secrete little, if any, CA19-9. Therefore, effective alternative diagnostic methods for PC is required. Methods: In this study, we examined whether differentially expressed proteins in the blood of PC patients could be identified by global shotgun proteomics analysis using an in vitro cell line sample. Results: We identified 142 candidate proteins with differential expression in PC cells. A semiquantitative method and bioinformatic analysis led to a total of 14 candidate proteins that could potentially be detected in blood. Validation studies revealed that the expression of fibulin-1 was lower in PC cells than in normal pancreatic ductal cells. Moreover, in vivo fibulin-1 expression was significantly lower in serum from PC patients than in healthy individuals. Conclusions: These findings suggest that lower serum levels of fibulin-1 may be a novel biomarker for the detection of PC.