OBJECTIVEIt is important to use noninvasive methods to differentiate liver fibrosis and liver cirrhosis. A prospective study was conducted to evaluate the validity of ultrasonography (US) in evaluating the severity of liver fibrosis in patients with chronic viral hepatitis in reference to the pathologic diagnosis of their liver biopsy specimens.

METHODSThe liver fibrosis status of 324 chronic viral hepatitis patients was evaluated by both needle biopsy and US. Histologically their liver fibrosis was graded as S0-S4, and the inflammatory reaction in the liver was graded as G1-G4. The US examination included qualitative description of the liver surface and liver parenchyma, and the quantitative parameters were vascular diameters, blood flow volume and spleen size.

RESULTSUS qualitative description of the liver surface and liver parenchyma was correlated to the severity of fibrosis and the degree of the inflammation seen in the liver biopsies. An analysis of US quantitative parameters showed that a cut-off value of 12.1 cm for the length of spleen had a sensitivity of 60.0%, and specificity of 75.3% in detecting early liver fibrosis. For other quantitative parameters, the cut-off values were 8mm for the diameter of the splenic vein, 30.5 cm/sec for maximal blood flow velocity in the portal vein and 12 mm in diameter of the main portal vein. The diagnostic sensitivities for these parameters were 60.0%, 78.6% and 76.7%; the diagnostic specificities were 78.1%, 66.9% and 44.6% respectively.

CONCLUSIONEarly cirrhosis can be detected by US, and the sonographic results were well paralleled with their pathologic diagnoses made by liver biopsies. Individual US parameter has limited sensitivity and specificity in diagnosing early cirrhosis. In clinical practice a combination of 2-3 parameters could be used to detect or exclude severe liver fibrosis.