Network Meta-Analysis Ranks Best Drugs for Reducing Liver Fat in MASH

Metabolic dysfunction-associated steatohepatitis (MASH) is a progressive inflammatory liver disease driving rising rates of cirrhosis, hepatocellular carcinoma, and liver-related mortality. As numerous pharmacologic agents enter clinical trials simultaneously, clinicians and trial designers urgently need comparative data on which agents most effectively reduce hepatic fat—a validated surrogate for treatment response—without relying on invasive liver biopsy.

This study pooled data from 39 Phase II–IV randomized controlled trials (3,311 participants, mean age ~53 years) identified through systematic searches of MEDLINE and Embase through December 2023. Investigators used Bayesian network meta-analysis to estimate the relative efficacy of 41 distinct interventions—including monotherapies and combination regimens—in reducing liver fat as measured by MRI proton-density-fat fraction (MRI-PDFF). Primary outcome was absolute MRI-PDFF change; secondary outcome was achieving a ≥30% relative decline, a threshold associated with MASH histologic resolution. Surface under the cumulative ranking curve (SUCRA) scores ranked each treatment's probability of being best.

At the 24-week primary timepoint, aldafermin (SUCRA: 83.65) and pegozafermin (SUCRA: 83.46)—both FGF21 analogs—led rankings for absolute MRI-PDFF reduction, followed by pioglitazone (SUCRA: 71.67). Aldafermin reduced MRI-PDFF by approximately 6.64 percentage points versus placebo, and pegozafermin by 6.60 points. Notably, colesevelam and sitagliptin were associated with statistically higher MRI-PDFF compared to placebo, suggesting potential harm or futility. At 12 weeks, efruxifermin (another FGF21 analog) ranked highest (SUCRA: 97.89) with a remarkable 14.4 percentage-point reduction. At 48 weeks, the cilofexor-firsocostat combination ranked first (SUCRA: 82.65). For the secondary endpoint of ≥30% MRI-PDFF decline at 24 weeks, efinopegdutide (GLP-1/GIP dual agonist, SUCRA: 67.02), semaglutide combined with firsocostat (SUCRA: 62.43), and pegbelfermin (SUCRA: 61.68) led rankings.

These results highlight FGF21 analogs and incretin-based therapies as consistently top-ranked agents across timepoints and endpoints. The superiority of certain combinations over monotherapies—particularly semaglutide plus firsocostat—supports the hypothesis that multi-mechanistic regimens targeting both lipid metabolism and inflammation may yield additive benefits. The authors note these data can directly inform sample size calculations and treatment arm selection for future phase III trials.

Several caveats temper interpretation. The network is built largely on Phase II trials with limited sample sizes, and most drug comparisons relied on a single study, precluding formal publication bias assessment. Differences in baseline patient characteristics, disease severity, and trial duration across studies introduce clinical heterogeneity. Additionally, MRI-PDFF reduction, while validated, is a surrogate endpoint; confirmatory evidence linking these rankings to hard clinical outcomes such as fibrosis regression or hepatic decompensation remains needed.