Evaluation of fasting and postprandial serum bile acid concentrations in hepatobiliary diseases
Cheng, T.C.; Chen, C.F.; Chou, P.T.; Sheen, T.Y.
Journal of the Formosan Medical Association 86(4): 413-421
1987
ISSN/ISBN: 0371-7682 PMID: 3598551 Document Number: 291658
The value of serum bile acid in the diagnosis of hepatobiliary disease has been investigated. An enzymatic colorometric method with an overall coefficient of variation of .+-. 9% in the control range was used. Serum was obtained aftr 8 hour fasting and after a meal from 49 healthy subjects, 28 patients without hepatobiliary diseases and 145 patients with hepatobiliary diseases. The upper limits for normal fasting and postprandial serum bile acid were 9 .mu.mol/L and 11 .mu.mol/L, respectively. In healthy subjects, the serum bile acid showed normal distribution. The fasting serum bile acid compared with postprandial serum bile acid was statistically significant in healthy subjects, as well as in patients without hepatobiliary diseases, acute hepatitis, cirrhosis of liver, or hepatocellular carcinoma. The largest number of cases with elevated serum bile acid was observed in chronic active hepatitis (10/10) and decompensated cirrhosis of the liver (9/9). The lowest was observed in patients with chronic persistent hepatitis (8/17, 7/17). A fasting serum bile acid at 9 .mu.mol/liter was related with a 95% probability of hepatobiliary disease. By increasing the cut off level of serum bile acid, there was a corresponding rise of predictive value for the positive test and specificity. On the other hand, there was a corresponding decrease of predictive value for a negative test and sensitivity. At 9 .mu.mol/liter, the predictive value for a positive test was 95.3%, and for the specificity, 93.5%; the predictive value for a negative test was 62.6%, and for the sensitivity, 70.3%. A postprandial serum bile acid level at 11 .mu.mol/liter showed the same results as in fasting. The difference between the fasting and postprandial serum bile acid levels were significantly smaller in the healthy subjects and patients without hepatobiliary disease than in patients with hepatobiliary disease except in metastatic tumor of the liver. The difference was significantly larger in acute hepatitis than in chronic hepatitis, compensated cirrhosis than chronic hepatitis, compensated cirrhosis of the liver, and cholelithiasis. A smaller difference was noted in chronic hepatitis and cholelithiasis than in cirrhosis of liver and hepatocellular carcinoma. The relationship of total hepatobiliary disease to both fasting and postprandial serum bile acid was significantly positive. In the diagnosis of hepatobiliary disease, alanine aminotransferase showed the highest sensitivity and predictive values for the negative test, and aspartate aminotransferase showed the highest specificity and predictive value of a positive test. Total protein and albumin were characterized by a poor sensitivity and predictive value for a negative test. The predictive values for a positive test and specificity of fasting and postprandial serum bile acid showed no statistically significant difference from those of conventional liver function tests. The predictive values for a negative test and sensitivity of serum bile acid were higher than those of total protein and albumin, but lower than those of aminotransferase. There was a significant difference between the fasting serum bile acid concentrations before and after hepatic coma in 11 cases of hepatocellular carcinoma and 6 cases of cirrhosis of the liver (by Wilcoxon T test). This indicated that increased fasting serum bile acid levels after onset of hepatic coma is related to deterioration of liver function. In conclusion, this study shows that both fasting and postprandial serum bile acid have the same sensitivity, specificity, predictive value of a positive and a negative test in a variety of hepatobiliary disease. Furthermore, detection of serum aminotransferase is a better screening test than serum bile acid.