Development, Validation, and Application of a Scoring Model for Non-alcoholic Steatohepatitis.

Abstract

The aim of this study was to explore risk factors of NASH and then develop a non-invasive scoring model in Chinese patients with obesity. A scoring system was then applied to assess the effect of sleeve gastrectomy on NASH. A total of 243 patients with obesity were included and divided into NASH group and non-NASH group according to the pathological results of liver biopsy. Logistic regression was used to determine risk factors of NASH. A scoring model was derived by risk factors of NASH. Then, postoperative follow-up was performed in 70 patients. Among the 243 patients, 118 (48.56%) patients showed NASH. Multivariate logistic regression identified aspartate aminotransferase (AST) (>21.50 IU/L), high-density lipoprotein cholesterol (HDL-C) (<1.155mmol/L), and homeostasis model assessment (HOMA-IR) (>9.368) as independent risk factors of NASH. The model included above risk factors showed a negative predictive value (NPV) of 70.38% in the low-risk category and a positive predictive value (PPV) of 85.71% in the high-risk category, with the area under the receiver operator curve (AUROC) of 0.737. Bariatric surgery resulted in a sharp decline in AST and HOMA-IR and a significant increase of HDL-C. The points of scoring model were improved at 6 months after surgery. A non-invasive scoring model was derived by the risk factors of NASH included AST, HDL-C, and HOMA-IR and applied to the postoperative follow-up. After sleeve gastrectomy, the above risk factors and points of scoring model were significantly improved.