We evaluated weight-loss (WL) direct and indirect effects of survodutide, a GCGR/GLP-1R dual agonist, on liver endpoints in a phase 2 trial in people with biopsy-confirmed MASH and F1–F3 fibrosis (NCT04771273).
A total of 295 participants were randomised to once-weekly s.c. survodutide 2.4, 4.8, 6.0 mg or placebo. Participants with fibrosis stage F2/F3 and paired baseline and end-of-treatment biopsy readings (N=170) were included, survodutide arms were pooled. Mediation analysis explores how an intermediate variable explains the pathway between an exposure and an outcome. This analysis assessed the proportion of the effect of survodutide on liver endpoints attributable to a direct/indirect effect mediated by WL. Liver endpoints assessed at Week 48 included resolution of MASH without worsening of fibrosis, improvement in fibrosis without worsening of MASH, absolute change in ELF™ score, relative change in PRO-C3, FAST score, LSM (FibroScan), and MRI-PDFF. Model included treatment and percentage change in bodyweight, with baseline bodyweight, type 2 diabetes status, and fibrosis stage as covariates. Total treatment effect (TE) or total odds ratios (OR), percentages mediated (proportion mediated by WL [indirect effect of changes in bodyweight in relation to the total effect]) and 95% CI are presented.
Liver endpoints that were highly mediated by WL were resolution of MASH without worsening of fibrosis (response rate: 62.9 vs 13.0% for placebo; OR: 14.70), or changes in MRI-PDFF (TE: −50.3 [−60.1, −40.5]). Endpoints for which WL had a lower attribution were ELF™ (TE: -0.64 [−0.83, −0.45]), improvement in fibrosis without worsening of MASH (response rate: 52.6 vs 25.9% for placebo; OR: 3.68), LSM (TE: −38.78 [−49.08, −28.49]), FAST (TE: −65.16 [−79.13, −51.18]), and PRO-C3 (TE: −28.63 [−36.38, −20.88]).
Survodutide had a direct (WL-independent) effect on liver endpoints related to improvement in inflammation and fibrosis via direct glucagon agonism in the liver.