Ultrastructural analysis of transjugular biopsies in MASLD reveals core hepatocellular stress and cytoskeletal adaptations.

BACKGROUND AND AIMS: Electron microscopy (EM) has provided key insights into hepatocellular alterations in MASLD, but most studies relied on percutaneous biopsies and primarily focused on selected features such as megamitochondria or autophagic vesicles. We aimed to evaluate the feasibility of ultrastructural analysis in transjugular biopsies and to provide a comprehensive description of hepatocellular changes in patients with MASLD.

METHODS

Patients with MASLD, well characterized by hepatic venous pressure gradient (HVPG) and stratified by fibrosis stage, portal hypertension, and MASH status, were prospectively enrolled. 18G transjugular liver biopsies were processed for EM. Hepatocellular alterations were systematically assessed, including mitochondrial, endoplasmic reticulum (ER), autophagic, cytoskeletal, and nuclear changes.

RESULTS

Eighteen patients were included; 8 (44.4%) had advanced fibrosis (F3-F4), 12 (66.7%) fulfilled diagnostic criteria for MASH, and 7 (38.9%) had portal hypertension (HVPG ≥6 mmHg). Transjugular specimens consistently yielded tissue suitable for high-quality ultrastructural assessment. The most frequent alterations were mitochondrial swelling, ER cisterns dilation, and accumulation of autophagic vesicles, reflecting cellular stress responses linked to hepatocellular injury, inflammation, and fibrogenesis. Cytoskeletal adaptations, specifically reinforcement of the cortical actin cytoskeleton and hypertrophy of adherens junctions, were also observed. These changes were observed across the fibrosis, portal pressure, and MASH subgroups.

CONCLUSIONS

Transjugular liver biopsies are suitable for EM studies in MASLD. The observed hepatocellular alterations appear to reflect core cellular stress responses and were noted consistently throughout the cohort. These findings provide a framework for integrating ultrastructural data into mechanistic and translational research in MASLD.