Assessment of portal contribution to liver perfusion by quantitative sequential scintigraphy and Doppler ultrasound in alcoholic cirrhosis. Diagnostic value in the detection of portal hypertension.

To assess the portal contribution to liver perfusion, we carried out quantitative sequential scintigraphy in 110 patients with alcoholic cirrhosis (22 Child-Pugh class A, 39 class B, 49 class C) and 15 normal subjects. Duplex Doppler ultrasound found a type of intrahepatic circulation that made the standard scintigraphic procedure inaccurate in four cases of cirrhosis, which were reevaluated. Portal contribution to liver perfusion was lower in cirrhotics than in normal subjects (48.7 +/- 29% versus 78.4 +/- 6%; p < 0.001). The sensitivity of scintigraphy in detecting portal hypertension, based on portal contribution < or = 66%, was 61.8% (with a 100% specificity) compared with 66.7% for endoscopy (diagnosis based on existence of varices). The overall sensitivity of the two tests together was 86.1%. Portal contribution to liver perfusion was inversely correlated to Child-Pugh score (r = 0.53; p < 0.001), to prothrombin time (r = 0.52; p < 0.001), and to hepatic venous pressure gradient (r = 0.43; p < 0.001) and positively correlated to albuminemia (r = 0.42; p < 0.001). Concurrent alcoholic hepatitis and the existence of large portosystemic collaterals were related to a decrease in portal contribution to liver perfusion. We conclude that quantitative sequential scintigraphy, which shows a direct relationship between portal contribution to liver perfusion, on the one hand, and the amount of portosystemic shunting, the progression of liver disease, and/or acute liver injury, on the other, could serve as a diagnostic test for portal hypertension. The addition of scintigraphy improves the overall sensitivity of endoscopy.