Primary biliary cirrhosis: proposal for a new simple histological scoring system

Background & aims: A simple and reproducible evaluation of non diagnostic histological lesions related to prognosis remains crucial in primary biliary cirrhosis (PBC). Presently there is no satisfactory simple scoring system analysing them reliably. We elaborated a semi-quantitative scoring system that assesses fibrosis, lymphocytic interface hepatitis (LIH) and ductopenia, separately. This study was aimed to evaluate its intra/interobserver reproducibility and its correlation with the main biochemical data.

Methods: Liver biopsies from 33 consecutive newly diagnosed PBC patients were independently analysed by five liver pathologists. Fibrosis was classified into five stages (portal/periportal fibrosis/few septa/numerous septa/cirrhosis) and LIH into four grades. The bile duct ratio (BDR), i.e. ratio of the number of portal tracts with ducts to total number of portal tracts, Ludwig's and Scheuer's stages were evaluated. Intra and interobserver agreements were assessed. Histological results were correlated to the biochemical data.

Results: Most patients had an early disease on clinical and biological parameters. The biopsies measured 23 mm on average (range 12 - 40 mm). Intraobserver reproducibility was substantial for fibrosis (κ = 0.68), LIH (κ = 0.69) and BDR (ICC = 0.69). Interobserver agreement for fibrosis was fair with the 5-class system (κ = 0.36), moderate with a 4-class system (κ = 0.56). moderate for LIH (κ = 0.59) and BDR (ICC = 0.50). Ludwig's and Scheuer's staging showed a fair interobserver agreement (κ = 0.32, κ = 0.31 respectively). Our system showed better correlations with biochemistry than Ludwig's and Scheuer's systems did.

Conclusions: This simple scoring system, assessing fibrosis, LIH and BDR separately, has a substantial intraobserver and a moderate interobserver reproducibility. Its prognostic relevance has to be evaluated.