Efficacy of pharmacologic interventions on magnetic resonance imaging biomarkers in patients with nonalcoholic fatty liver disease: systematic review and network meta-analysis

Background and aim: Several agents are under investigation for nonalcoholic fatty liver disease (NAFLD). We assessed the comparative efficacy of pharmacologic interventions for patients with NAFLD focusing on magnetic resonance imaging (MRI) biomarkers.

Methods: We searched Medline, Embase, and CENTRAL. We included randomized controlled trials of more than 12 weeks of intervention that recruited patients with biopsy-confirmed or MRI-confirmed NAFLD and assessed the efficacy of interventions on liver fat content (LFC) and fibrosis by means of MRI. We performed random-effects frequentist network meta-analyses and assessed confidence in our estimates using the CINeMA (Confidence in Network Meta-Analysis) approach.

Results: We included 47 trials (8583 patients). Versus placebo, thiazolidinediones were the most efficacious for the absolute change in LFC, followed by vitamin E, fibroblast growth factor (FGF) analogs, and glucagon-like peptide-1 receptor agonists (GLP-1 RAs) with mean differences ranging from -7.46% (95% confidence interval [-11.0, -3.9]) to -4.36% (-7.2, -1.5). No differences between drug classes were evident. Patients receiving GLP-1 RAs or glucose-dependent insulinotropic polypeptide (GIP)/GLP-1 RAs were more likely to achieve ≥30% relative reduction in LFC. Among agents, efruxifermin produced the largest reduction in LFC compared to placebo [-13.5% (-18.5, -8.5)], followed by pioglitazone, while being superior to most interventions. The effect of interventions on magnetic resonance elastography assessed fibrosis was small and insignificant. The confidence in our estimates was low to very low.

Conclusions: Several drug classes may reduce LFC in patients with NAFLD without a significant effect on fibrosis; nevertheless, trial duration was small, and confidence in the effect estimates was low.