Background and objectives: Several serum based-markers and ratios have been investigated for their prognostic value in pancreatic ductal adenocarcinoma (PDAC). This cohort study aimed to combine these into a novel prognostic scoring system.
Methods: A retrospective cohort study was performed on 145 patients with unresectable histologically-confirmed PDAC. Based on the existing literature the following markers were investigated: neutrophil-lymphocyte ratio (NLR), neutrophil-albumin ratio (NAR), platelet-lymphocyte ratio (PLR), fibrinogen, and Ca19-9. These values were dichotomized about their medians for Kaplan-Meier and Cox regression analysis.
Results: Univariate Cox regression revealed statistically significant prognostic value for: NLR, NAR, PLR, fibrinogen, and Ca19-9. When combining these using Cox regression analysis adjusting for other prognostic indicators, only NAR (hazard ratios [HR] = 3.174, P = 0.022) and Ca19-9 (HR = 2.697, P = 0.031) were independent predictors of survival. Combining NAR and Ca19-9 we split the cohort into three "NARCA" groups: NARCA0 = NAR ≤ 0.13 and Ca19-9 ≤ 770, NARCA1 = either NAR > 0.13 or Ca19-9 >770, NARCA2 = NAR > 0.13 and Ca19-9 > 770. Median survival was 20.5, 9.7 and 4.1 months in NARCA0, 1, and 2 respectively ( P < 0.0005, log-rank test). A separate validation cohort confirmed the prognostic significance of the score ( P = 0.048).
Conclusions: Combining NAR and Ca19-9 into a prognostic score allows stratification of unresectable PDAC patients into groups with significantly different overall survival.
Keywords: CA-19-9 antigen; biomarkers; carcinoma; palliative; pancreatic ductal; tumor.
© 2018 Wiley Periodicals, Inc.