Send us a text

The first three days of the 2022 program focused on a range of issues, with specific emphasis on non-invasive tests (NITs) and their role at different stages in diagnosis and treatment. This conversation From the Vault touches on two issues: use of AI to better understand the meaning of liver volume and elements of morphology and, relatedly, how much more sophisticated a view we take of NAFLD and NASH than we did even five years ago.

The issues around AI and liver morphology arise from Jörn Schattenberg’s comment that consistent with Stephen Harrison’s “KISS” (Keep It Simple, Stupid!) principle, researchers are starting to explore the meaning of changes in liver volume. Ultimately, Jörn notes, pairing these kinds of measures to AI-supported histopathology can yield tremendous benefits. Roger comments on a breakfast he attended that morning that suggested that AI and NITs each provide different, important information on individual liver health: NITs can address collagen burden but not structure, while AI can identify changes in structure but not link them to the impact on the patient. Zobair ends this part of the conversation by noting that companies are starting to use AI in these ways.

The second part of this conversation stems from Roger’s observation that we know far more about the disease than we did 3-4 years ago. He goes on to describe how the environment is more collaborative and open-minded than it might have been if, in fact, we saw drug approvals at that time. Zobair takes this observation to a different plane, noting that 5-10 years ago, “some very important experts” believed that we all understood the etiology of Fatty Liver disease, so why spend more time? He goes on, “we could not have been more wrong,” observing that what we have learned about progression and regression in placebo arms suggests a far more complex disease than something that progresses linearly, or even constantly in one direction. He goes on to add that the multiple drug trial failures is that targeting drugs to a single solution based on animal models is likely to fail because this is a multiple mode of action disease. His third point: the “source” of the disease is visceral obesity and insulin resistance, which all viable solutions must address. This identifies two targets for treatment, while simultaneously demonstrating that therapy will be chronic, lifelong and with behavioral elements. The rest of the conversation addresses the challenges with shaping this kind of lifelong, multi-target therapy in the US today.