OBJECTIVE: To evaluate the value of biliary carcino-embryonic antigen (CEA) in the differential diagnosis of malignant and benign hepatopancreatobiliary disease. PATIENTS: One hundred patients were prospectively studied. Benign diseases were present in 39% of the patients while 61% had malignant diseases. METHODS: Samples of serum were taken from all patients just before endoscopic retrograde cholangiopancreatography (ERCP) and samples of biliary CEA were obtained during ERCP. RESULTS: The sensitivity of serum CEA and carbohydrate antigen 19-9 (CA 19-9) in detecting malignancy were 50% and 92%, respectively, while the respective specificities were 95% and 72%. The mean biliary CEA level of the benign group was significantly different from that of the malignant group (35.7 +/- 8.7 ng/ml vs 268 +/- 85.5 ng/ml), but there was considerable overlap between the two groups. With a cut-off level of 20 ng/ml, the sensitivity and specificity were 84% and 64% respectively. The mean bilirubinaemia value was significantly higher in malignant disease than in benign disease (57.4 +/- 13.9 mumol/l vs 235 +/- 19.8 mumol/l). Multidimensional analysis indicated that only bilirubinaemia (P < 109-3)) was independently predictive of malignant disease. CONCLUSION: Biliary CEA assessment seems useless in distinguished between benign and malignant causes of cholestasis.
OBJECTIVE: To evaluate the value of biliary carcino-embryonic antigen (CEA) in the differential diagnosis of malignant and benign hepatopancreatobiliary disease. PATIENTS: One hundred patients were prospectively studied. Benign diseases were present in 39% of the patients while 61% had malignant diseases. METHODS: Samples of serum were taken from all patients just before endoscopic retrograde cholangiopancreatography (ERCP) and samples of biliary CEA were obtained during ERCP. RESULTS: The sensitivity of serum CEA and carbohydrate antigen 19-9 (CA 19-9) in detecting malignancy were 50% and 92%, respectively, while the respective specificities were 95% and 72%. The mean biliary CEA level of the benign group was significantly different from that of the malignant group (35.7 +/- 8.7 ng/ml vs 268 +/- 85.5 ng/ml), but there was considerable overlap between the two groups. With a cut-off level of 20 ng/ml, the sensitivity and specificity were 84% and 64% respectively. The mean bilirubinaemia value was significantly higher in malignant disease than in benign disease (57.4 +/- 13.9 mumol/l vs 235 +/- 19.8 mumol/l). Multidimensional analysis indicated that only bilirubinaemia (P < 109-3)) was independently predictive of malignant disease. CONCLUSION: Biliary CEA assessment seems useless in distinguished between benign and malignant causes of cholestasis.