Diagnostic efficacy of serum alkaline phosphatase and γ-glutamyltransferase in dogs with histologically confirmed hepatobiliary disease: 270 cases (1980-1990)

Abstract

Summary The diagnostic efficacy of serum alkaline phosphatase (alp) and γ-glutamy I transferase (ggt) activities was examined, using the records of 270 dogs initially suspected of having hepatobiliary disease on the basis of history, findings on physical examination, results of baseline screening tests, or any combination of these data. Histologic examination of hepatic tissue was performed in each dog. Sixty-three dogs did not have histologic evidence of hepatobiliary disease and served as the control group. On the basis of diagnosis, dogs were assigned to 1 of 8 groups: dogs with cirrhosis (n = 34), steroid hepatopathy (n = 16), hepatic neoplasia (primary and secondary, n = 36), chronic hepatitis (n = 14), chronic passive congestion (n = 5), hepatic necrosis (n = 17), portosystemic vascular anomaly (n = 35), and cholestasis (extrahepatic bile-duct obstruction and intrahepatic cholestasis, n = 50). Of the 207 dogs with hepatobiliary disease, 29 (14%) had normal alp and ggt activities, 31 (15%) had normal alp activity, and 112 (54%) had normal ggt activity. Of the 63 control dogs, 29 (46%) had normal serum alp and ggt activities, 32 had normal alp activity (alp specificity, 51%), and 55 had normal ggt activity (ggt specificity, 87%). The specificity of alp and ggt in parallel (positive result = result of either test abnormal) was 46%, and in series (positive result = results of both tests abnormal) was 91%. The highest median activities of alp developed in dogs with cholestasis, steroid hepatopathy, chronic hepatitis, and hepatic necrosis. The highest median activities of ggt developed in dogs with steroid hepatopathy, cholestasis, and hepatic necrosis. In most dogs with hepatobiliary disease, the magnitude of variation (measured value/highest normal value) for alp was greater than that for ggt. The sensitivity of alp exceeded that of ggt in all groups but the passive congestion group. The high sensitivity of alp increased the sensitivity of testing in parallel, compared with the sensitivity of testing in series. The positive predictive value (+ pv) was greater for ggt than for alp in all groups except dogs with portosytemic vascular anomaly. The + pv of testing in series was greater than the + pv associated with single tests or testing in parallel in all groups except dogs with passive congestion and with portosystemic vascular anomaly. Single tests or testing in parallel provided a greater negative predictive value (− pv) than did testing in series in most groups, although the differences were small in individual groups. The exception was the steroid hepatopathy group, in which testing in series provided the greatest − pv. The poor specificity of serum alp activity in the detection of hepatobiliary disease in dogs was confirmed by our data. Serum ggt activity was less influenced by nonhepatic disease processes or enzymeinducing drugs, compared with serum alp activity. The diagnostic value of alp and ggt was greater when tests were evaluated in series.