We’re very excited to bring you a Two-Part Special Episode of Dialogues on AI Digital Pathology!
Listen to our guests, KOL liver pathologists as they discuss the recent FDA Webcast on NASH drug development and utilization of AI digital pathology as potential clinical trial endpoints.
Here’s Part 1, where the KOL liver pathologists will discuss key takeaways from the recent FDA Webcast on NASH Drug Development, and respond to questions sent in by Pharma companies and clinical investigators on potential clinical trial strategies using AI digital pathology following the Webcast.
Prof. Pierre Bedossa,
Professor of Pathology at the University of Paris; Consultant in Liver Pathology at LIVERPAT, Paris, France.
Prof. Zachary Goodman,
Director of Hepatic Pathology Consultation and Research, Inova Fairfax Hospital, Falls Church, Virginia, USA.
Prof. David Kleiner,
Senior Research Physician and Chief of Post-mortem Section, Laboratory of Pathology, Center for Cancer Research,National Cancer Institute, Bethesda, Maryland, USA.
Dr. Nikolai Naoumov, Adviser for Clinical Research and Drug Development in Liver Diseases, Novartis Pharma AG.
Key Takeaways from the KOLs in this Special Episode (Part 1)
- It was suggested that the liver biopsy should undergo a quality check for adequate amount of tissue with a core length of 2cm. Digital-read of stained glass slides would be possible only with a high level of quality control and rigour in tissue preparation and staining.
- For liver histology adjudication in clinical trials, it was suggested that two pathologists will be adequate to reach an agreement without a need for a third pathologist. In the event of a disagreement, a consensus among the two pathologists could be achieved through a mutual discussion.
- AI digital pathology (AI-DP) could serve its purpose on two fronts during NASH clinical trials – to train younger pathologists in scoring/staging biopsies and aid expert pathologists in assessing borderline cases.
- AI-DP could be useful as secondary endpoints in NASH clinical trials to augment semi-quantitative read. Especially for quantitative parameters that cannot be seen or detected by the human eye, AI-DP provides additional data possibly into the biology of response or regression.
- For AI-DP to be considered as primary endpoints, more data is needed to show the magnitude of effect change that correlates to clinical outcomes. The data should come from longitudinal studies evaluating the natural history of NASH vs placebo.