Comparison of simulated transient elastography sample volumes and global measurements on MRE and MRI-PDFF in patients with suspected MASLD.

BACKGROUND

Hepatic fibrosis can be staged by biopsy or non-invasive stiffness measurements. However, the extent to which liver stiffness varies throughout the liver, and the impact of regions of interest (ROIs) selection, remains unclear. FibroScan measures stiffness in an anatomically constrained ROI. We compared liver stiffness and fat quantification from a simulated FibroScan-sized ROI on Magnetic Resonance Elastography (MRE) with global liver stiffness obtained from standard whole-liver MRE.

METHODS

This retrospective study included adults who underwent MRE and Magnetic Resonance Imaging-Proton Density Fat Fraction (MRI-PDFF) as part of evaluation for suspected metabolic dysfunction-associated steatotic liver disease (MASLD). Patients with other chronic liver diseases were excluded. No liver biopsies were performed. Using MRE stiffness maps and MRI-PDFF images, a 4 × 1-cm ROI was placed in segment VIII to simulate FibroScan-sized sampling volume. Reference standards were global MRE and global MRI-PDFF.

RESULTS

Among 118 patients, 73 had non-elevated stiffness on whole-liver MRE. The simulated ROI correlated with whole-liver MRE (r = 0.72, p < 0.0001) and MRI-PDFF (r = 0.96, p < 0.0001). Specificity, negative predictive value, and accuracy for cirrhosis-equivalent stiffness were 95.0%, 96.9%, and 93.2%, respectively, and for steatosis grade 3 were 99.0%, 93.8%, and 97.5%. Sensitivity was 83% for cirrhosis-equivalent stiffness and 52% for intermediate stiffness ranges.

CONCLUSION

A simulated FibroScan-sized ROI on MRE demonstrated good agreement with whole-liver MRE but showed reduced sensitivity in intermediate stiffness ranges; in uncertain or discordant cases, whole-liver MRE may provide a more comprehensive stiffness assessment.