Validity and reliability of the nonalcoholic fatty liver diseases activity score (NAS) in Korean NAFLD patients and its correlation with clinical factors

BACKGROUND/AIMS: Nonalcoholic steatohepatitis (NASH) is commonly diagnosed using the semi-quantitative grading and staging system proposed by Brunt et al. in 1999. The Pathology Committee of the NASH established the nonalcoholic fatty liver diseases (NAFLD) activity score (NAS) in 2005. The aim of this study was to elucidate the validity and reliability of the NAS in Korean NAFLD patients. METHODS: Fifty-six patients on whom sonography-guided liver biopsy for well-defined NAFLD was performed between 1999 and 2007 were identified retrospectively. Two pathologists evaluated each biopsy sample. NAFLD was evaluated using both the grading system developed by Brunt et al. and the NAS. Each pathologist was blinded to the patients' clinical data and scored independently. We evaluated the body mass index (BMI), liver enzymes, lipid profile, peripheral insulin resistance, leptin, insulin/c-peptide ratio, ferritin, and fasting blood glucose. RESULTS: The patients were aged 32.1+/-12.5 years (mean+/-SD) and comprised 44 males (78.6%). Patients with different grades at the two grading systems had mild steatosis or ballooning changes with fibrosis, and 36.6% of them were borderline cases (NAS of 3 or 4). The interobserver agreement on diagnostic category was 0.748 (P<0.001) for the NAS (using weighted kappa statistics). Elevated fasting glucose, ALT, and triglyceride were associated with the NAS. CONCLUSIONS: The simple and reproducible NAS was found to be a useful pathologic grading system in Korean NAFLD patients. However, the proportion of borderline cases based on the NAS was high. The "wait and see" strategy is necessary for evaluating the long-term prognosis.