Background & Aims: Non alcoholic steatohepatitis (NASH) has no approved pharmacological therapy. Insulin sensitisers such as thiazolidinediones ameliorate insulin resistance and are a potential therapeutic option. We performed a systematic review and meta-analysis of the effect of thiazolidinediones on histological and biochemical variables in NASH. Methods: Two reviewers searched Medline, Embase, Cochrane Central, international meeting abstracts, reference lists, and contacted experts. Inclusion criteria were randomized trials of people with NASH receiving thiazolidinediones, compared with placebo or other treatments. Methodological quality was assessed in domains suggested by the Cochrane Collaboration. The primary outcome was histological improvement (fibrosis, steatosis, inflammation, hepatocellular ballooning, and NAS score). Secondary outcomes included change in alanine transaminase, insulin resistance, body mass index, weight, and adverse events. Meta-analysis used random effects with dichotomous outcomes as relative risk (RR) and continuous outcomes as mean difference (MD), both with 95% confidence intervals (CI). Results: Of seven randomized trials (n = 489) with histological outcomes, four were placebo controlled (n = 355). Methodological quality was variable although better for placebo controlled studies. Treated participants showed improvement in fibrosis (RR 1.38, CI 1.01-1.89), steatosis (RR 2.03, CI 1.57-2.62), inflammation (RR 1.71, CI 1.32-2.21), and hepatocellular ballooning (RR 1.62, CI 1.15-2.28). Treatment increased weight by an average of 4.4 kg (CI 2.6-5.2 kg). Adverse event reporting was inconsistent and only one trial assessed quality of life. Conclusions: Thiazolidinediones modestly improve histological variables including fibrosis and hepatocellular ballooning, but at the cost of significant weight gain. Trials of longer duration and reporting of patient oriented outcomes would be informative.
- Metabolic liver disease