Poster
Retrospective AI-based Measurement of NASH Histology (AIM-NASH) Analysis of Biopsies From Phase 2 Study of Resmetirom Confirms Significant Treatment-induced Changes in Histologic Features of Nonalcoholic Steatohepatitis
EASL 2022
Study Background
• Biopsy-based endpoints are recommended for evaluating efficacy in nonalcoholic steatohepatitis (NASH) clinical trials due to slow progression of the disease. However, to date, no effective therapies have been identified.
• Manual histologic scoring of NASH features (lobular inflammation, ballooning, steatosis, fibrosis) is the current standard for assessing biopsies; however, manual scoring is subjective and prone to intra- and inter-rater variability, leading to only moderate-to-fair scoring reproducibility. Furthermore, manual scoring lacks the sensitivity to detect small histologic changes that may indicate treatment-induced clinical improvements.
• We hypothesize that improving the methods by which changes in NASH histology are measured may improve clinical trial outcomes. AI-assisted or -powered pathology may enhance the reproducibility and sensitivity of NASH histologic feature assessment and allow analyses of additional exploratory features with known critical relevance, such as portal inflammation.
• In the current retrospective analysis, we demonstrate the utility of an AI-based tool (AIM-NASH) for scoring NASH histology of liver biopsies from a Phase 2 trial of resmetirom (MGL-3196), a liver-targeted, oral, once-daily selective thyroid hormone receptor-β agonist in development for treatment of NASH with significant fibrosis.
• Manual histologic scoring of NASH features (lobular inflammation, ballooning, steatosis, fibrosis) is the current standard for assessing biopsies; however, manual scoring is subjective and prone to intra- and inter-rater variability, leading to only moderate-to-fair scoring reproducibility. Furthermore, manual scoring lacks the sensitivity to detect small histologic changes that may indicate treatment-induced clinical improvements.
• We hypothesize that improving the methods by which changes in NASH histology are measured may improve clinical trial outcomes. AI-assisted or -powered pathology may enhance the reproducibility and sensitivity of NASH histologic feature assessment and allow analyses of additional exploratory features with known critical relevance, such as portal inflammation.
• In the current retrospective analysis, we demonstrate the utility of an AI-based tool (AIM-NASH) for scoring NASH histology of liver biopsies from a Phase 2 trial of resmetirom (MGL-3196), a liver-targeted, oral, once-daily selective thyroid hormone receptor-β agonist in development for treatment of NASH with significant fibrosis.
Authors
Harrison et al.