Evaluation of the severity of nonalcoholic fatty liver disease by the analysis of serum exosomal miRNA expression

Noninvasive modalities exhibit limited diagnostic performance for evaluating the severity of nonalcoholic fatty liver disease (NAFLD). MicroRNAs (miRNAs) serve as useful biomarkers for diagnosing and monitoring disease progression and treatment response. Here, we evaluated whether serum exosomal miRNAs can be used to clinically predict the severity of NAFLD. Exosomal miRNAs were isolated from the sera of 41 NAFLD patients (diagnosed using biopsy) for array profiling. To compare differentially expressed miRNAs, the degree of severity was determined using inflammation, steatosis, ballooning, and NAFLD activity score (NAS). The correlation of miRNAs with clinical and biochemical parameters was analyzed. The correlation between miRNA and mRNA expression was analyzed by comparing our miRNA expression data with publicly available mRNA expression data. Twenty-five, eleven, thirteen, and fourteen miRNAs correlated with inflammation score, steatosis score, ballooning score, and NAS, respectively. Thirty-three significant correlations were observed between twenty-seven miRNAs and six clinical variables (age, ALP, AST, ALT, GGT, and NAS). In fibrosis, 52 and 30 interactions corresponding to high miRNA-low mRNA and low miRNA-high mRNA expression, respectively, were observed. We demonstrated that serum exosomal miRNAs can be used to evaluate NAFLD severity and also delineate potential targets for NAFLD treatment.