Adjusting CA19-9 values with clinical stage and bilirubin to better predict survival of resectable pancreatic cancer patients: 5-year-follow-up of a single center.

Background: Pancreatic cancer mortality is growing every year, and radical resection is the most essential therapy strategy. It is critical to evaluate the long-term prognosis of individuals receiving radical surgery. CA19-9 is a biomarker for patient recurrence and survival, however obstructive jaundice has a significant impact on this index. Researchers have attempted to modify the index using various modification methods, but the results have been unsatisfactory. In this study, we adjusted CA19-9 values based on clinical stage and bilirubin and found that it provided better prediction than CA19-9 alone in assessing patients. Methods: We analyzed over 5 years follow-up records of patients who underwent radical pancreatic cancer surgery between August 2009 and May 2017 in a single center. We investigated the association of risk factors with overall survival (OS) as well as disease-free survival (DFS) after surgery. Threshold values for high-risk features associated with poor prognosis in resectable pancreatic cancer were determined. The hazard ratios of the indicators were eventually examined under the stratification of patients' clinical stages. Results: A total of 202 patients were involved in the study. The optimum cut-off values for CA19-9 and CA19-9/TB for predicting overall survival were 219.4 (p = 0.0075) and 18.8 (p = 0.0353), respectively. CA19-9>219.4 increased the risk of patient mortality by 1.70 times (95% CI 1.217-2.377, p = 0.002), and tumor poor differentiation raised the risk by 1.66 times (95% CI 1.083-2.553, P = 0.02). Based on clinical stage stratification, we found discrepancies in the predictive efficacy of CA19-9 and CA19-9/TB. CA19-9 was a better predictor in clinical stage 1 (HR = 2.056[CI 95%1.169-3.616], P = 0.012), whereas CA19-9/TB indications were better in stages 2 (HR = 1.650[CI 95%1.023-2.662], P = 0.040) and 3 (HR = 3.989[CI95%1.145-13.896], P = 0.030). Conclusions: CA19-9, CEA, and tumor differentiation are predictors for patients with resectable PDAC. CA19-9 values can be adjusted based on clinical stage and bilirubin levels to better predict overall survival in patients with resectable PDAC. CA19-9>219.4 predicted poor survival in individuals in clinical stage 1, whereas CA19-9/TB>18.8 predicted poor survival for individuals in stages 2 and 3.