Does CA 19-9 Have Prognostic Relevance in Gallbladder Carcinoma (GBC)?

ABSTRACT

BACKGROUND: There is a scarce data on prognostic relevance of carbohydrate antigen (CA 19-9). This retrospective study was undertaken to evaluate its prognostic relevance in different prognostic subsets of gallbladder carcinoma (GBC). MATERIALS AND METHODS: One hundred forty-one patients of GBC treated between January 2012 and December 2014 were the subjects of this retrospective analysis. Baseline CA 19-9 levels of four cohorts of patients extended cholecystectomy (EC), simple cholecystectomy (SC) with residual or recurrent disease, locally advanced disease (LAGBC) and metastatic disease were ascertained. The difference in its median baseline values among above groups was ascertained. The effect of clinicopathological variables, treatment-related variables and CA 19-9 on overall survival (OS) was also evaluated. AUC curve was computed to evaluate its performance. RESULTS: The median baseline levels of CA 19-9 were significantly different [10 units/ml, 24 units/ml, 48 units/ml and 75 units/ml in EC (n = 33), SC (n = 21), LAGBC (n = 38) and metastatic disease (n = 49), respectively, (p value 0.001)]. The median OS was also significantly different [24, 15, 7 and 6 months in EC, SC, LAGBC and metastatic disease, respectively, (p value 0.001)]. Univariate analysis revealed a significant influence of log transformed value of CA 19-9, CA 19-9 levels < or >20 units or 35 units, surgery vs. none and chemoradiation vs. chemotherapy on OS. On multivariate analysis, only treatment-related variables were significant (HR 1.1, 95% CI 1.026-1.19, p = 0.009). AUC curve was 0.63 for all patients and 0.72 for EC group. CONCLUSIONS: The median values of baseline CA 19-9 predict the burden of disease. Raised levels of serum CA 19-9 beyond 20 units/ml should be used for prognostication purposes after EC. A level beyond 35 units has a trend towards prognostication in other prognostic groups and needs to be evaluated in large subset of patients.