Background: Serum carbohydrate antigen 19-9 (CA 19-9) has been shown to be a sensitive and specific serum marker for pancreatic cancer. Little has been published about correlations between baseline CA 19-9 level or changes to CA 19-9 level and median overall survival (mOS). Its impact on monitoring treatment efficacy remains under discussion, however.
Methods: CA 19-9 serum level was measured in 181 consecutive patients with advanced pancreatic cancer (APC) being treated with gemcitabine-based first-line chemotherapy. We separated the patients into several groups depending on baseline CA 19-9 levels and the CA 19-9 response after 6-8 weeks of treatment. Evaluations were made using SPSS 19.9.
Results: Median baseline CA 19-9 level was 1,493 U/ml (range 40-1,043,301). Patients with baseline CA 19-9 ≤1,000 U/ml had a mOS of 14.9 months (95% CI: 11.36:18.44), whereas patients with CA 19-9>1,000 U/ml had a mOS of 7.4 months [(95% CI: 5.93:8.87) p < 0.001, HR 2.12]. With regard to the change in CA 19-9 after 6-8 weeks of treatment: patients with increased CA 19-9 levels had a mOS of 8.1 months, those with stabilized CA 19-9 levels 11.6 months, and those with decreased CA 19-9 levels 11.1 months (p < 0.019).
Conclusion: CA 19-9 levels can separate patients with differing mortality risks at baseline. Patients with stabilization or high response of CA 19-9 after 6-8 weeks of treatment had no significant differences in survival rates, whereas patients with increased CA 19-9 had significantly lower survival rates, indicating an early treatment failure.
Keywords: CA 19-9; pancreatic neoplasms; prediction; response; therapy control.