Validity of monoamine oxidase in serum for diagnosis of liver cirrhosis: estimation of predictive values, sensitivities and specificities.

The validity of monoamine oxidase (EC 1.4.3.4) activity in serum for the diagnosis of fibroproliferative liver disorders was assessed by measuring the specificity, sensitivity and positive and negative predictive values of the enzyme. Enzyme activity was measured in sera of 567 patients including those with biochemically and/or histologically verified non-fibrotic liver diseases (n = 64), liver fibrosis (n = 45), and liver cirrhosis (n = 51). The fraction of liver cirrhotic subjects with pathologically elevated monoamine oxidase activity (greater than 630 U/1) was 0.61, whereas only 0.16 of the cases with liver fibrosis and 0.11 of those with non-fibrotic liver diseases had abnormally high enzyme activities. Among the various categories of diseases tested, significantly increased enzyme activities were confined to liver cirrhosis and chronic active hepatitis. For liver cirrhosis, sensitivities and specificities were calculated as functions of various cut-off (critical) values of monoamine oxidase activity in serum, and with respect to a reference population of healthy men and non-cirrhotic subjects. The predictive value of the positive test result (enzyme activity higher than 720 U/1) at a prevalence of liver cirrhosis of maximum 0.033 (estimated incidence of chronic liver diseases in West Germany) is 0.68 if tested against healthy persons and less than 0.30 if tested against patients with non-liver cirrhotic diseases. It is concluded that monoamine oxidase is probably not helpful in the early diagnosis of fibroproliferative liver dysfunctions but may provide a parameter of complications of cirrhosis, e.g. portal-systemic collateral circulation (portal hypertension).