A Modification of the Brunt System for Scoring Liver Histology of Patients with Non-Alcoholic Fatty Liver Disease

Background: Nonalcoholic steatohepatitis is a common cause of chronic liver disease. It is important to have a uniform and validated method for scoring liver histology in these patients. Therefore, we propose a modification of the Brunt system by scoring the four histologic features separately and reporting their sums as the nonalcoholic steatohepatitis activity index.

Methods: A nonalcoholic steatohepatitis activity index was defined which scored the grade of disease activity between 0 and 12 according to four histologic features: steatosis, hepatocyte ballooning, portal inflammation, and lobular inflammation. Fibrosis was separately scored between 0 and 4. A total of 60 liver specimens collected from nonalcoholic steatohepatitis patients were scored by two pathologists at the time of biopsy and after three months. Liver enzyme levels were also correlated with the histologic score.

Results: The intra-rater agreement (weighted kappa) for various variables of the scoring system was between 0.59 and 0.80 for one pathologist and 0.78 to 0.95 for the other. The repeatability of the nonalcoholic steatohepatitis activity index was excellent with only 4% of the repeated measures out of the acceptable range of agreement. In addition, liver enzyme levels strongly correlated with the nonalcoholic steatohepatitis activity index.

Conclusion: This system provides a fine graduation of liver histologic damage in nonalcoholic steatohepatitis and is well suited for use in clinical trials or natural history studies.