Including Tumor Size Improves Ability of CA 19-9 to Predict Survival in Pancreatic Adenocarcinoma After Resection with Perioperative Chemotherapy

Abstract

Abstract Background and Objective: Five-year survival in pancreatic adenocarcinoma is less than 20%. While previous studies have postulated that a Carbohydrate Antigen 19 − 9 (CA19-9) threshold could predict outcome of resection, the role for CA19-9 in decision-making remains unclear. This study aims to assess whether CA19-9 levels combined with tumor size improve prediction of post-resection survival. Method: A retrospective analysis was conducted on 109 patients with pancreatic adenocarcinoma who underwent perioperative chemotherapy followed by resection. The primary outcome of mortality was, divided into short (< 1year) or prolonged (> 2years). Univariate and multivariable analyses compared the tumor size-adjusted CA19-9 between the outcome groups. Results: Twenty-seven (24.78%) and eighty-two (75.23%) patients were in the short-survival and prolonged-survival groups, respectively. The mean CA19-9 was significantly greater in the short versus prolonged group (P < 0.001). Analyzing CA19-9 level by tumor size, the association of high CA19-9 and short survival was significant for small (≤ 2cm) and large tumor (> 4cm), but not for intermediate-size tumors (2-4cm). Adjusting for preoperative variable did not change this association. Conclusion: CA 19 − 9 in combination with tumor size better identifies patients with prolonged post-resection survival. This prediction is most accurate in patients with either small (≤ 2cms) or large (> 4cms) tumors compared to intermediate size tumors.