Abstract
Background and Aims: Pancreatic ductal adenocarcinoma has one of the worst prognosis of all malignancies. This investigated the relationship between the preoperative serum carbohydrate antigen 19-9 and surgical resectability. Methods: A systematic search was performed in three databases (MEDLINE, EMBASE, and Web of Science) to compare the surgical resectability of pancreatic ductal adenocarcinoma in patients with high and low preoperative serum carbohydrate antigen 19-9 values. The receiving operating characteristic curves were constructed and the weighted mean differences for preoperative serum carbohydrate antigen 19-9 levels of resectable and unresectable groups of patients were calculated. The PROSPERO registration number is CRD42019132522. Results: Results showed that there was a significant difference in resectability between the low and high carbohydrate antigen 19-9 groups. Six out of the eight studies utilised receiver operating characteristic curves in order to find the cut-off preoperative carbohydrate antigen 19-9 levels marking unresectability. The overall result from the pooled area under curve values from the receiver operating characteristic curves was 0.794 (CI: 0.694-0.893), showing that the preoperative carbohydrate antigen 19-9 level is a "fair" marker of resectability. The result of the pooled weighted mean differences was 964 U/ml (p < 0.001) showing that there is a significant carbohydrate antigen 19-9 difference between the resectable and unresectable groups. Based on the results of the I-squared test, the result was 87.4%, accounting for "considerable" heterogeneity within the population. Conclusion: Carbohydrate antigen 19-9 is not a reliable marker of unresectability, it should not be used on its own in surgical decision-making.