Abstract

Background: Following reductions in steatosis and body weight with glucagon-like peptide-1 receptor agonist treatment for nonalcoholic steatohepatitis (NASH), decreases in liver volume can cause collagen condensation resulting in an over estimation of fibrosis burden when measured by pathologist-reported histology evaluation. Thus, improved methods to objectively evaluate histological changes are needed. Here, digital quantification of the collagen proportionate area (CPA) was compared in the total biopsy area and non-steatotic liver tissue (fat-free CPA) following semaglutide treatment for NASH.

Methods: This was a post-hoc exploratory analysis of a phase 2 randomized trial of subcutaneous semaglutide 0.1, 0.2, or 0.4 mg once daily versus placebo (NCT02970942). Patients had biopsy-confirmed NASH and fibrosis stage F1–F3. Liver biopsies were obtained up to 21 weeks before screening or at baseline and at week 72; digitized biopsy slides were evaluated. CPA was quantified by measuring collagen deposition as a proportion of either: 1) the total biopsy area (standard CPA) or 2) the non-steatotic biopsy area (i.e. normalized for fat: fat-free CPA = collagen area / [biopsy area – steatosis area]). Changes from baseline to week 72 were analyzed by analysis of covariance for both methods.

Results: Digitized slides were available for 249 patients for the standard CPA analysis, and 246 patients for the fat-free CPA analysis. A dose-dependent semaglutide treatment effect was seen with both methods (Figure). Standard CPA was numerically reduced with semaglutide 0.4 mg vs placebo (estimated treatment difference [ETD]: –1.68 [95% confidence interval: –4.62, 1.26]; P=0.26). For fat-free CPA, the semaglutide 0.4 mg ETD increased to –2.99 (95% confidence interval: –6.39, 0.41), and the P-value approached statistical significance (P=0.08). An enhanced reduction of CPA was seen across all semaglutide doses when measured by fat-free versus standard CPA (Figure).

Conclusion: When measuring CPA before and after semaglutide treatment, the removal of the confounding effect of the fat area results in numerically greater improvements in fibrosis. The fat-free adjustment analysis for CPA increases the accuracy of fibrosis resolution assessment when using drugs with strong anti-steatogenic effects.