A validated method for quantifying macrovesicular hepatic steatosis in chronic hepatitis C.

Hepatic steatosis is increasingly seen as an important prognostic factor in chronic hepatitis C infection (HCV). The commonly used semiquantitative method of measuring steatosis is based on a study that excluded patients with HCV. Several potentially useful methods of quantifying steatosis using computer-assisted morphometric analysis have been proposed, but none has been validated against a proposed gold standard other than the method they were intended to replace. We present a novel method and propose a gold standard based on manual measurements. The manual method is time consuming but shows little interobserver error, and the mean value of 3 observations by separate investigators is proposed as the gold standard. The computer-assisted method is fast, with a single interactive step that shows minimal interobserver variation. It accurately identifies biopsies with <1% steatosis (7 of 7) and predicts the gold standard value for biopsies with > 1% steatosis with narrow CIs (geometric mean ratio 0.85 with 95% CIs 0.77-0.95). This novel method of computer-assisted morphometric analysis is fast, reliable, and suitable for future research in HCV steatosis. It may be used to reanalyze previous studies. The semiquantitative method of assessing steatosis remains appropriate for clinical purposes.