Abstract
BACKGROUND AND AIMS: Magnetic resonance (MR)-based techniques are increasingly being used in early-phase trials of metabolic dysfunction-associated steatohepatitis (MASH) as a noninvasive alternative to liver histology for assessing drug efficacy. Understanding the placebo effect in MR-based endpoints is essential for interpreting results and designing future trials. We conducted a systematic review and meta-analysis to quantify the placebo response and identify its modifiers in MR-based endpoints. METHODS: We systematically searched PubMed and the Cochrane Library through December 13, 2024 for phase ≥2 randomized controlled trials (RCTs) comparing pharmacologic agents with placebo in adults with MASH that provided MR-based assessments. Pooled placebo response rates were estimated using generalized linear mixed models with Clopper-Pearson intervals. RESULTS: We included 21 RCTs comprising 800 placebo-treated patients. A clinically meaningful relative reduction in liver fat (≥30% relative reduction in magnetic resonance imaging [MRI] proton density fat fraction [MRI-PDFF]) was observed in 19.21% (95% CI: 16.23 to 22.59; I(2) = 7.2%) of placebo patients. The pooled placebo response rates for liver fat normalization (≥5% absolute reduction in MRI-PDFF) and MRI-PDFF super-response (≥50% relative reduction in MRI-PDFF) were 17.19% (95% CI: 11.70 to 24.56; I(2) = 27.9%) and 6.14% (95% CI: 2.63 to 13.69; I(2) = 12.9%), respectively. Reductions in liver stiffness (≥15% relative reduction by magnetic resonance elastography) and fibro-inflammation (corrected T1 response by multiparametric MRI) occurred in 26.09% (95% CI: 18.14 to 35.98; I(2) = 15.1%) and 13.43% (95% CI: 7.14 to 23.85; I(2) = 0%), respectively. CONCLUSIONS: Placebo-treated patients with MASH show significant improvements across MR-based endpoints with low heterogeneity. These data support the potential reproducibility of MR-based endpoints in assessing treatment efficacy in MASH and should be considered in future trial design.