We’re very happy to bring you Part 4 that will wrap up the Special Episode podcast of Dialogues on AI Digital Pathology!
In this Part 4 podcast, our guests – KOL Hepatologists and Pharma Thought Leaders in NASH – continue their discussions in response to questions by Pharma companies and clinical investigators. Find out their views on how AI Digital Pathology (AI-DP) can effectively impact NASH drug development, preclinical studies and clinical practice with granular histopathological data.
Prof Stephen Harrison,
Medical Director, Pinnacle Clinical Research (USA)
Prof Vlad Ratziu,
Professor of Hepatology, Sorbonne University and Pitié-Salpêtrière Hospital (France)
Pharma Thought Leaders in NASH
Dr. Judith Ertle,
Senior Clinical Program Lead NASH, Boehringer Ingelheim (Germany)
Dr. Karin Conde-Knape,
Senior Vice President Global Drug Discovery, Novo Nordisk (Denmark)
Absent with Apologies
Ms. Kitty Yale,
EVP and Chief Development Officer, Akero Therapeutics (USA)
Dr. Nikolai Naoumov, Senior Adviser for Clinical Research and Drug Development in Liver Diseases,Novartis Pharma AG (Switzerland)
Key Takeaways from this Special Episode (Part 4)
Pathologists will still take a center stage role when it comes to diagnosis in clinical practice and could use AI Digital Pathology (AI-DP) as an aid for NASH diagnosis to help gain a more robust and definitive interpretation of the lesions. In clinical trials, and in assessing treatment response in general, AI-DP readouts will be dominant, and will rapidly become part of trial protocol for endpoints, provided that the methodology fulfils the regulatory requirements and is validated. For example, what magnitude of effect with AI-DP equates to outcome measures; what % change of qFibrosis will translate into clinical benefit, etc.
- As AI-DP utilizes unique parameters that are different from traditional histological assessments to describe, for example, fibrosis or ballooned cells, there will be a need for more understanding on how these measurable parameters are correlated with the traditional method and how much of these changes are relevant. Using AI-DP in clinical trials would allow the testing of this association as DP fully quantitative readouts are overlaid with the categorical, semiquantitative stages/scores.
- Clinical trials involving patients with NASH cirrhosis could be of particular benefit – firstly, the use of AI-DP will provide greater granularity of fibrosis dynamics and changes in collagen fibers as a result of intervention, which cannot be seen by conventional microscopy; secondly, there would be an opportunity for linking digital measures with hard endpoints and clinical outcomes in the cirrhotic population. Thus, clinical trials with cirrhotic patients may provide a unique opportunity for AI-DP, more rapidly than in pre-cirrhotic trials with F2/F3 patients.
- AI-DP would be very valuable in Phase II clinical trials where the subject size is usually small. Having the additional quantitative assessment tool alongside the pathologist would enhance sensitivity and provide more confidence in the observed changes or reduction in fibrosis between the treatment groups or treatment arms with a placebo. This would be particularly important for the pharma companies to better plan for dose selection, evaluate efficacy, and make a decision for their clinical development especially in preparation for Phase III.
- AI-DP will work in parallel with the pathologist’s reading for eligibility and assessing treatment response as this is in accordance with the regulatory guidance. Although the current workflow involves the pathologist as the decision maker, and the AI-DP as an additional tool to provide more granularity and quantitative readouts on the data, in future, it is possible to have an initial AI-DP readout, followed by a QC reading by a pathologist in real-time.
- There is also a great potential for AI-DP to be incorporated during the screening period of trials. Thus, increasing the probability of patients’ continued enrolment and minimizing inertia, placebo responses as well as inter and intra observer variabilities. The use of AI-DP however, should not prolong the current turnaround/screening period.
- Using AI-DP in assessing NASH biopsies will give us greater sensitivity and accurate consistent assessment of disease activity and stage which will benefit patients and all stakeholders involved in NASH management. AI-DP will also be useful to study disease activity in preclinical models, to standardize what is being seen in preclinical models in terms of the effect of different therapeutic drugs. This will provide a much-improved assessment of changes on, for instance, balloon cells, and fibrosis, which would translate significantly to drug development.