Fig. 2: Treatment effect on clinical outcomes for the per-protocol population.

A Bar chart of the number of patients having at least 10% reduction in α-SMA expression (primary endpoint), yellow = Yes, blue = No. B Bar chart of the number of patients having at least 20% reduction in liver stiffness measured by transient electrography, yellow = Yes, blue = No. C, D Scatter plots of treatment effect and compliance for each treatment group. Error bands = 95% confidence intervals. Treatment effect was estimated as percentage change in α-SMA expression from baseline to 24 weeks, n = 40. E Dot plot showing percentage change in α-SMA expression from baseline to 24 weeks, n = 40 (p value derived from estimation adjusted for compliance). F Estimated mean difference (95% confidence intervals) of non-invasive tests combined with a forest plot of Cohen’s d to compare effect size. Forest plot of effect sizes and corresponding 95 % confidence intervals for changes in non-invasive secondary endpoints between baseline and 24 weeks, PRO-C3 n = 40, ELF test n = 40, 2D-SWE n = 38, CAP n = 41. All p values are two-sided. A, B Odds ratios (OR) were calculated using unadjusted logistic regression. E The estimate of difference was calculated using linear regression, adjusted for compliance.α-SMA Alpha Smooth Muscle Actin, TE transient elastography, ELF test enhanced liver fibrosis test, PRO-C3 fragment of N-terminal type III collagen, SWE shear wave elastography, CAP controlled attenuation parameter. Source data are provided as a Source Data file.