MRI techniques shown to improve cirrhosis diagnosis

Magnetic resonance–based imaging techniques can accurately assess both fibrosis and steatosis in patients with advanced liver disease, according to a new study of liver transplant recipients. The approach could offer a non-invasive alternative to biopsy, which is often difficult in this population.

In the retrospective analysis, published in the Journal of Gastroenterology and Hepatology, investigators from Korea evaluated 187 adults with a median age of 57 years who underwent liver transplantation between 2019 and 2023 and had pretransplant MR elastography (MRE) and MRI–proton density fat fraction (MRI-PDFF) imaging. Whole liver tissue analysis was used as the reference standard, allowing detailed evaluation of both fibrosis and steatosis.

Nearly all patients (97%) had cirrhosis, most commonly due to hepatitis B or alcohol-related liver disease, and the median Model for End-Stage Liver Disease (MELD) score was 11.

MRE showed strong performance for staging fibrosis. It identified advanced fibrosis (F4) with an area under the curve (AUC) of 0.92 and distinguished severe cirrhosis (Laennec stage F4c) with an AUC of 0.85. Liver stiffness measurements increased progressively with fibrosis severity, rising from about 3 kPa in earlier stages to more than 8 kPa in advanced cirrhosis.

MRI-PDFF also demonstrated good diagnostic accuracy for steatosis, with an AUC of 0.83. The optimal threshold for detecting steatosis was 2.8%, lower than previously reported values, a difference the authors attributed to the effects of advanced fibrosis on fat quantification.

MRE findings were also associated with clinical measures of disease severity. Patients classified as having cirrhosis or severe cirrhosis by MRE had higher MELD scores, more frequent complications of portal hypertension — including ascites and varices — worse laboratory parameters, and greater intraoperative transfusion requirements. However, MRE-defined cirrhosis severity was not independently associated with one-year graft survival.

According to the authors, the findings indicate that MRE can provide detailed, noninvasive staging even in patients with advanced disease. They noted that the ability to distinguish degrees of cirrhosis severity, rather than simply identifying its presence, may offer additional clinical insight. MRI-PDFF remained useful for fat quantification, although its performance may be affected by fibrosis-related signal changes and iron deposition.

The results show that imaging-based assessment could reduce reliance on biopsy, support risk stratification, and complement established scoring systems such as MELD and Child–Pugh by capturing structural aspects of disease not reflected in laboratory values.

The authors noted several limitations of the analysis, including that the study population consisted entirely of transplant candidates, many with relatively low MELD scores and a high prevalence of hepatocellular carcinoma, which may limit generalizability. The retrospective, single-center design also warrants confirmation in broader populations. In addition, cutoff values may vary by disease etiology, few patients underwent transient elastography for comparison, and MRI-PDFF accuracy may be reduced in advanced fibrosis.

The investigators called for further research to validate MRE thresholds in non-transplant populations, standardize cutoff values across etiologies, integrate imaging into clinical decision-making frameworks, and refine MRI-based fat quantification in cirrhosis.

They reported having no disclosures.