BACKGROUND Life expectancy of pancreatic ductal adenocarcinoma (PDAC) patients is usually short and selection of the most appropriate treatment is crucial. The aim of this study was to investigate the usefulness of serum CA 19-9 as a surrogate marker under no impress excluding other factors affecting CA 19-9 level other than tumor itself.

METHODS We recruited 1,446 patients with PDACs and patients with Lewis antigen both negative or obstructive jaundice were excluded to eliminate the false effects on CA 19-9 level. The clinicopathologic factors were reviewed including initial and post-treatment CA 19-9, and statistical analysis was done to evaluate the association of clinicopathologic factors with overall survival (OS).

RESULTS The total of 944 patients was enrolled, and205 patients (22%) underwent operation with curative intention and 541 patients (57%) received chemotherapy and/or radiotherapy. The median CA 19-9 levels of initial and post-treatment were 670 IU/ml and 147 IU/ml respectively. The prognostic factors affecting OS were performance status, AJCC stage and post-treatment CA 19-9 level in multivariate analysis. Subgroup analysis was done for the patients who underwent R0 and R1 resection, and patients with normalized post-operative CA 19-9 (≤37 IU/mL) had significantly longer OS and DFS regardless of initial CA 19-9 level; 32 vs. 18 months, P<0.001, 16 vs. 9 months, P = 0.004 respectively.

CONCLUSIONS Post-treatment CA 19-9 and normalized post-operative CA 19-9 (R0 and R1 resected tumors) were independent factors associated with OS and DFS, however, initial CA 19-9 level was not statistically significant in multivariate analysis.