Multivariate analysis of six serum tumor markers (CEA, CA 50, CA 242, TPA, TPS, TATI) and conventional laboratory tests in the diagnosis of hepatopancreatobiliary malignancy.

A prospective study of 277 patients with benign (n = 212) and malignant (n = 65) hepatopancreatobiliary disease was carried out to evaluate the value of six serum tumour markers (CEA, CA 50, CA 242, TPA, TPS, TATI) and 16 conventional laboratory tests in the distinction between benign and malignant diseases. In univariate analysis, all tumour marker tests except TATI showed significantly (p < 0.001) higher serum values in the patients with malignant disease than in the patients with benign disease. Among the conventional laboratory tests serum bilirubin, alkaline phosphatase and leucine aminopeptidase showed significantly. (p < 0.001) higher values in the patients with malignant disease, whereas serum protein and amylase levels were significantly (p < 0.01) higher in the patients with benign disease. In a multivariate analysis, serum bilirubin (p < 0.001), antithrombin III (p < 0.01) and blood hemoglobin (p < 0.05) were the most significant independent predictors of hepatopancreatobiliary malignancy. To sum up the contributions of the best tests a diagnostic score (DS) was developed. The sensitivity of DS in detecting malignancy was 73% with a specificity of 82% and an efficiency of 79%. In conclusion, our results speak against the use of multiple tumour marker tests, and rather suggest the use of a relatively limited amount of conventional laboratory tests in the distinction between benign and malignant hepatopancreatobiliary disease.