Liver Viscosity Imaging and Disease Severity in Alcohol-Associated Hepatitis: A Prospective Cross-Sectional Pilot Study

Background: Alcohol-associated hepatitis (AH) is an acute inflammatory phenotype of alcohol-associated liver disease associated with substantial short-term mortality. Established clinical scores quantify disease severity but do not directly characterize liver tissue mechanics. Ultrasound-derived viscosity-related measurements may provide complementary, non-invasive information on viscoelastic tissue behavior, although their biological determinants remain incompletely defined. Objective: We aimed to evaluate the association between liver viscosity and AH through the Sound Touch Visco-elastography Imaging (STVi) parameter and inflammatory, biochemical, and MDF-defined severity measures, and to explore its discriminatory performance. Methods: This prospective, single-center, cross-sectional pilot study included 62 patients with clinically diagnosed AH and 30 healthy volunteers. Participants underwent multiparametric ultrasound assessment using the Resona A20 platform with STVi, liver stiffness measurement, and attenuation imaging. Results: The viscosity-related parameter was higher in patients with AH than in healthy volunteers and was associated with liver stiffness, inflammatory markers, and MDF-defined severity. The exploratory AH-versus-healthy comparison yielded an AUC of 0.873 (95% CI, 0.797–0.934). Within the AH cohort, 28 patients had severe AH (MDF ≥ 32) and 34 had non-severe AH. For MDF-defined severe AH, the AUC was 0.863 (bootstrap 95% CI, 0.756–0.955). At the Youden-derived cutoff of 2.8 Pa·s, sensitivity was 78.6% and specificity was 94.1%. Conclusions: STVi provides a model-derived, platform-dependent viscosity-related parameter associated with inflammatory measures and clinical severity in AH. These preliminary findings suggest that liver viscosity imaging may represent a valuable new avenue for the non-invasive assessment of AH. However, standardized acquisition, longitudinal assessment, histological correlation, and external validation are required before clinical implementation.