HistoIndex‘s Second Harmonic Generation (SHG) stain-free digital pathology platform has shown marked reduction in liver fibrosis (overall, reduction in perisinusoidal fibrosis, as well as regression of fibrous septa) from baseline to end-of-treatment at 48 weeks, in patients with NASH participating in a Phase 2 FLIGHT-FXR study. The data from the analysis further illustrates the significance of the AI-based digital pathology platform in revealing fibrosis regression in patients treated with Tropifexor (TXR) and its relation to steatosis (liver fat) reduction.
The results of this study were presented at EASL’s International Liver Congress 2021 (ILC 2021) entitled, “Digital Pathology with Artificial Intelligence Analyses Reveal New Dynamics of Treatment-induced Fibrosis Regression in Nonalcoholic Steatohepatitis“. The AI-based digital pathology platform by HistoIndex scanned and analyzed unstained sections from 198 paired liver biopsies obtained from 99 patients F2–F3 stage NASH fibrosis.
Fibrosis Regression and Relation to Steatosis Reduction
In contrast to pathologist’s conventional staging, the AI-based analysis showed fibrosis regression in a greater proportion of TXR-treated patients compared with placebo especially in the higher dose of TXR 200 µg. Alongside the fibrosis improvements, the SHG fully quantitative assessment showed that regression of fibrous septa or fine collagen fibers within the septa region, from baseline to end-of treatment at 48 weeks, was present in 67% of TXR 200 µg-treated patients, much greater when compared with 28% of patients receiving placebo.
Noteworthy observations were made from a concomitant measurement (co-localization analysis) of fibrosis changes around fat vacuoles in the same area of the liver lobule. The AI-based co-localized analysis indicated that TXR-treated patients with reduced steatosis showed a marked fibrosis reduction of 70-83% across the perisinusoidal area, including all three zones of the liver lobule, as compared to 47% – 54% fibrosis reduction in placebo patients. This is a novel finding – enabled by AI digital pathology – that reinforces the ability of SHG stain-free digital pathology platform to fully quantify changes in fibrosis along with other NASH characteristics in various regions of the liver biopsy on a continuous scale, which is critical in assessing post-treatment responses.
Professor Arun Sanyal, Professor of the Internal Medicine Department, Division of Gastroenterology, Hepatology and Nutrition at Virginia Commonwealth University, and Lead Investigator of the study, commented by saying, “The current approaches for histological evaluation of NASH after initiation of therapy is restrictive by a limited dynamic range, making it difficult to assess changes in fibrosis in phase 2 studies, which are typically of relatively short duration unless there is a substantial change in fibrosis burden and distribution. AI-based approaches using SHG provide a method to assess changes in fibrosis with greater sensitivity in a standardized and reproducible way to inform decision-making in drug development”.
“As various characteristic features of hepatic fibrosis are relevant prognostic determinants of NASH clinical outcomes, it is encouraging to see our SHG digital pathology platform unlock meaningful histological evidence that is pertinent to the FLIGHT-FXR study in highlighting dose-dependent treatment responses. This encouraging data spurs HistoIndex to continue supporting new and on-going pharmaceutical efforts with its AI-based stain-free quantitative tool, as patients with NASH wait in hope for a treatment,” says Dr. Dean Tai, Chief Scientific Officer of HistoIndex.
The presentation, P#889 – Digital Pathology with Artificial Intelligence Analyses Reveal New Dynamics of Treatment-induced Fibrosis Regression in Nonalcoholic Steatohepatitis, is available to registered delegates of the International Liver Congress 2021 (ILC 2021), Virtual Congress.