Trial Outcomes & Findings for Role of Lisinopril in Preventing the Progression of Non-Alcoholic Fatty Liver Disease, RELIEF-NAFLD Study (NCT NCT04550481)

Role of Lisinopril in Preventing the Progression of Non-Alcoholic Fatty Liver Disease, RELIEF-NAFLD Study

Descriptive statistics will be used to summarize changes and values at each time point. The primary analysis will be performed using a paired t-test to compare the pre- to post-treatment changes in PRO-C3 levels. PRO-C3 levels will be log-transformed to satisfy the normality assumption. If log-transformation does not satisfy the normality assumption, a signed rank test will be used.

Descriptive statistics will be used to summarize changes and values at each time point. The primary analysis will be performed using a paired t-test to compare the pre- to post-treatment changes in PRO-C6 levels. PRO-C6 levels will be log-transformed to satisfy the normality assumption. If log-transformation does not satisfy the normality assumption, a signed rank test will be used.

Measured by CAP, determined with transient elastography. Changes in categorical variables will be analyzed using the Stuart-Maxwell test of marginal homogeneity. Linear regression models will be used to identify baseline characteristics that are associated with biomarker changes. Model selection will be conducted according to the purposeful model selection approach (REF) to a parsimonious set of predictors. Model fit will be assessed using standard regression diagnostics. Changes in categorical outcomes from pre- to post-treatment will also be categorized as improvement vs. no improvement (i.e. no change or worsening in steatosis grade), and logistic regression models will be used to examine association with baseline characteristics.

Measured with magnetic resonance elastography. Changes in categorical variables will be analyzed using the Stuart-Maxwell test of marginal homogeneity. Linear regression models will be used to identify baseline that are associated with biomarker changes. Model selection will be conducted according to the purposeful model selection approach (REF) to a parsimonious set of predictors. Model fit will be assessed using standard regression diagnostics. Changes in categorical outcomes from pre- to post-treatment will also be categorized as improvement vs. no improvement (i.e. no change or worsening in steatosis grade), and logistic regression models will be used to examine association with baseline characteristics.

Measured with transient elastography. Changes in categorical variables will be analyzed using the Stuart-Maxwell test of marginal homogeneity. Linear regression models will be used to identify baseline characteristics that are associated with biomarker changes. Model selection will be conducted according to the purposeful model selection approach (REF) to a parsimonious set of predictors. Model fit will be assessed using standard regression diagnostics. Changes in categorical outcomes from pre- to post-treatment will also be categorized as improvement vs. no improvement (i.e. no change or worsening in steatosis grade), and logistic regression models will be used to examine association with baseline characteristics.

Changes in categorical variables will be analyzed using the Stuart-Maxwell test of marginal homogeneity. Linear regression models will be used to identify baseline characteristics that are associated with biomarker changes. Model selection will be conducted according to the purposeful model selection approach (REF) to a parsimonious set of predictors. Model fit will be assessed using standard regression diagnostics. Changes in categorical outcomes from pre- to post-treatment will also be categorized as improvement vs. no improvement (i.e. no change or worsening in steatosis grade), and logistic regression models will be used to examine association with baseline characteristics. Scores fall into three ranges: \< -1.455 (Low Risk for F0-F2), -1.455 to 0.675 (Indeterminate), and \> 0.675 (High Risk for F3-F4).

Changes in categorical variables will be analyzed using the Stuart-Maxwell test of marginal homogeneity. Linear regression models will be used to identify baseline characteristics that are associated with biomarker changes. Model selection will be conducted according to the purposeful model selection approach (REF) to a parsimonious set of predictors. Model fit will be assessed using standard regression diagnostics. Changes in categorical outcomes from pre- to post-treatment will also be categorized as improvement vs. no improvement (i.e. no change or worsening in steatosis grade), and logistic regression models will be used to examine association with baseline characteristics. \< 1.30: Low likelihood of significant liver fibrosis. 1.30 - 2.67: Indeterminate or intermediate risk. \> 2.67: High likelihood of advanced fibrosis.

Descriptive statistics, including means, 95% confidence intervals, or median and inter-quartile range, will be used to summarize changes as well as biomarker values at each time point. To determine whether changes in these biomarkers occurred following treatment, paired t-test will be used as described for the analysis of the primary endpoint. Biomarker values may be transformed to satisfy the normality assumption. If no suitable transformation can be found, a signed rank test will be used. For the analysis of inflammatory biomarkers, tests will be adjusted for multiple comparisons to control the false discovery rate using the method of Romano.

Changes in categorical variables will be analyzed using the Stuart-Maxwell test of marginal homogeneity. Linear regression models will be used to identify baseline characteristics that are associated with biomarker changes. Model selection will be conducted according to the purposeful model selection approach (REF) to a parsimonious set of predictors. Model fit will be assessed using standard regression diagnostics. Changes in categorical outcomes from pre- to post-treatment will also be categorized as improvement versus (vs.) no improvement (i.e. no change or worsening in steatosis grade), and logistic regression models will be used to examine association with baseline characteristics.

Prognostic Liver Secretome signature (PLSec), was developed as a surrogate biomarker for HCC development by utilizing integrative database/pipeline of protein subcellular localization, extracellular secretion to circulation, and tissue type specificity for HCC. Descriptive statistics, including means, 95% confidence intervals, or median and inter-quartile range (IQR), will be used to summarize changes as well as biomarker values at each time point. To determine whether changes in these biomarkers occurred following treatment, paired t-test will be used as described for the analysis of the primary endpoint. Biomarker values may be transformed to satisfy the normality assumption. The PLSec score ranges from 0 to 8, with a higher score indicating higher risk for HCC.