Mouse models of non-alcoholic steatohepatitis: A reflection on recent literature.

Non-alcoholic steatohepatitis (NASH) is strongly associated with overnutrition, insulin resistance, and predisposition to type 2 diabetes. To critically analyze the translational significance of currently used animal models of NASH, we reviewed articles published during the last 3 years that studied NASH pathogenesis using mouse models. Among 146 articles, 34 (23%) used models in which overnutrition was reported, and 36 (25%) demonstrated insulin resistance, with or without glucose intolerance. Half the articles contained no information on whether mice exhibited overnutrition or insulin resistance. While 75 papers (52%) reported > 2-fold increase of serum/plasma alanine aminotransferase (ALT) compared with controls, ALT levels were near normal or not reported in 48%. Liver pathology was assessed by a pathologist with an interest in liver pathology in 53% of articles published in gastroenterology/hepatology journals, versus 43-44% in other journals. While there appears to be a trend to use models that are potentially relevant to the pathogenesis of human NASH, journals currently publish data on mouse models in which overnutrition and insulin resistance do not occur, without ALT increase or appropriate analysis of NASH pathology. We recommend that investigators, reviewers, and journal editors carefully consider the validity of NASH models in current use and that moves are made to reach a consensus on what the minimal criteria should be.