MRI-based proton density fat fraction combined with adipose tissue for identifying metabolic dysfunction-associated steatohepatitis

To develop a non-invasive diagnostic model combining magnetic resonance imaging (MRI)-derived proton density fat fraction (PDFF) and adipose tissue for identifying metabolic dysfunction-associated steatohepatitis (MASH).

A total of 197 consecutive patients with metabolic dysfunction-associated steatotic liver disease (MASLD) who underwent liver biopsy, MRI-PDFF, and transient elastography (TE) were retrospectively enrolled. Independent predictors of MASH were identified using univariate and multivariate logistic regression analyses. A combined model was constructed and internally validated via 1000 bootstrap resamples. The diagnostic performance was compared with that of established non-invasive tests (NITs).

Multivariable logistic regression analyses demonstrated that the PDFF and total adipose tissue index (TATI) were independent risk factors for MASH. The MRI-based PDFF-adipose tissue (MR-PAT) model achieved an area under the receiver operating characteristic curve (AUC) of 0.866 for MASH identification. Bootstrap internal validation yielded an AUC of 0.870. The performance of the models was consistent across clinically relevant subgroups. The diagnostic performance of the MR-PAT model for MASH identification was significantly better than that of all the established NITs (all p < 0.05).

The MR-PAT model, combining MRI-derived PDFF and TATI, provides an accurate and robust non-invasive tool for identifying MASH. It outperforms widely recognized NITs and holds promise for risk stratification in patients with MASLD.

Question MASH is a critical histological endpoint of MASLD, but reliable non-invasive tools for its assessment remain inadequate. Findings The MR-PAT model integrating MRI-based PDFF and adipose tissue quantification demonstrates robust and excellent diagnostic performance for identifying MASH. Critical Relevance Statement The MR-PAT model leverages hepatic steatosis and the adipose tissue burden to elucidate the pathophysiological mechanisms driving steatohepatitis. It can serve as an accurate, non-invasive tool for identifying MASH and holds promise for risk stratification in patients with MASLD.