AI‐based digital pathology provides newer insights into lifestyle intervention‐induced fibrosis regression in MASLD: An exploratory study

Abstract Background and Aims Lifestyle intervention is the mainstay of therapy for metabolic dysfunction‐associated steatohepatitis (MASH), and liver fibrosis is a key consequence of MASH that predicts adverse clinical outcomes. The placebo response plays a pivotal role in the outcome of MASH clinical trials. Second harmonic generation/two‐photon excitation fluorescence (SHG/TPEF) microscopy with artificial intelligence analyses can provide an automated quantitative assessment of fibrosis features on a continuous scale called qFibrosis. In this exploratory study, we used this approach to gain insight into the effect of lifestyle intervention‐induced fibrosis changes in MASH. Methods We examined unstained sections from paired liver biopsies (baseline and end‐of‐intervention) from MASH individuals who had received either routine lifestyle intervention (RLI) ( n = 35) or strengthened lifestyle intervention (SLI) ( n = 17). We quantified liver fibrosis with qFibrosis in the portal tract, periportal, transitional, pericentral, and central vein regions. Results About 20% (7/35) and 65% (11/17) of patients had fibrosis regression in the RLI and SLI groups, respectively. Liver fibrosis tended towards no change or regression after each lifestyle intervention, and this phenomenon was more prominent in the SLI group. SLI‐induced liver fibrosis regression was concentrated in the periportal region. Conclusion Using digital pathology, we could detect a more pronounced fibrosis regression with SLI, mainly in the periportal region. With changes in fibrosis area in the periportal region, we could differentiate RLI and SLI patients in the placebo group in the MASH clinical trial. Digital pathology provides new insight into lifestyle‐induced fibrosis regression and placebo responses, which is not captured by conventional histological staging.