Abstract
BACKGROUND: Pancreatic ductal adenocarcinoma (PDAC) often recurs early after radical resection, and such early recurrence (ER) is associated with a poor prognosis. Predicting ER is useful for determining the optimal treatment. METHODS: One hundred fifty-three patients who underwent pancreatectomy for PDAC were divided into an ER group (n = 54) and non-ER group (n = 99). Clinicopathological factors were compared between the groups, and the predictors of ER and prognosis after PDAC resection were examined. RESULTS: The ER group had a higher platelet count, higher platelet-to-lymphocyte ratio (PLR), higher preoperative CA19-9 concentration, higher SPan-1 concentration, larger tumor diameter, and more lymph node metastasis. The receiver operating characteristic (ROC) curve analysis identified cut-off values for PLR, carbohydrate antigen 19-9 (CA19-9), SPan-1, and tumor diameter. In the multivariate analysis, a high PLR, high CA19-9, and tumor diameter of >3.1 cm were independent predictors of ER after resection (all p < 0.05). When the parameter exceeded the cut-off level, 1 point was given, and the total score of the three factors was defined as the ER prediction score. Next, our new ER prediction model using PLR, CA19-9 and tumor diameter (Logit(p) = 1.6 + 1.2 × high PLR + 0.7 × high CA19-9 + 0.5 × tumor diameter > 3.1cm) distinguished ER with an area under the curve of 0.763, a sensitivity of 85.2%, and a specificity of 55.6%. CONCLUSIONS: ER after resection of PDAC can be predicted by calculation of a score using the preoperative serum CA19-9 concentration, PLR, and tumor diameter.