Pancreatic adenocarcinoma (PDAC) is known to have a poor prognosis partly because of lack of effective biomarkers. In the test set, we investigated dysbindin (DTNBP1) as a potential biomarker for PDAC by comparing preoperative and postoperative serum mass spectrometry (MS) proteomic profilings. Of the included 50 PDAC patients, 42 (positivity of 84.0%) detected a lower MS peak in postoperative serums than preoperative ones which was then identified as dysbindin. In the verification set, receiver operating characteristics (ROC) were used to assess diagnostic efficiency. 550 participants were included in the verification set [250 with PDAC, 80 with benign biliary obstruction (BBO), 70 with chronic pancreatitis (CP) and 150 healthy donors (HD)]. Dysbindin was increased in PDAC patient sera than in all controls. ROC curves revealed the optimum diagnostic cutoff for dysbindin was 699.16 pg/ml [area under curve (AUC) 0.849 (95% CI 0.812-0.885), sensitivity 81.9% and specificity 84.7%]. Raised concentration of dysbindin in sera could differentiate PDAC from BBO, CP and HD. Moreover, dysbindin maintained its diagnostic accuracy for PDAC patients who were CA19-9 negative [AUC 0.875 (95% CI 0.804-0.945), sensitivity 83.0%, specificity 89.0%] and for patients with benign biliary obstruction [AUC 0.849 (95% CI 0.803-0.894), sensitivity 82.3%, specificity 84.0%].Our discovery of dysbindin may complement measurement of CA19-9 in the diagnosis of PDAC and help to discriminate PDAC from other pancreatic diseases or begin biliary obstruction.
Keywords: CA19-9; ROC; biomarker; dysbindin; pancreatic ductal adenocarcinoma.
© 2016 UICC.