Detection of pancreatic cancer using serum protein profiling

AbstractBackgroundCurrently, no suitable biomarkers for the early detection of pancreatic cancer (PC) are available. Proteins present in the serum could reflect a state of the disease. In this study, these profiles as a diagnostic marker for PC were evaluated.MethodsSerum samples were obtained from PC patients (n = 50 calibration set, n = 39 validation set) and healthy volunteers (n = 110 and n = 75 respectively) according to a uniform standardized collection and processing protocol. For peptide and protein isolation, automated solid‐phase extraction (SPE) with Weak Cation Exchange (WCX) magnetic beads (MB) was performed using a 96‐channel liquid handling platform. Protein profiles were obtained by mass spectrometry (MS) and evaluated by linear discriminant analysis with double cross‐validation.ResultsA discriminating profile for PC has been identified, with a sensitivity of 78% and a specificity of 89% in the calibration set with an area under the curve (AUC) of 90%. These results were validated with a sensitivity of 74% and a specificity of 91% (AUC 90%).ConclusionSerum profiles of healthy controls and PC can be discrimated between. Further research is warranted to evaluate specificity and whether this biosignature can be used for early detection in a high risk population