MRI multicomponent relaxometry for liver fibrosis, inflammation, and steatosis in nonalcoholic steatohepatitis: A...

To validate contrast-free MRI multicomponent relaxometry (MCR) for noninvasive quantification of fibrosis, inflammation, steatosis, and iron in biopsy-proven metabolic dysfunction-associated steatohepatitis (MASH), formerly nonalcoholic steatohepatitis (NASH).

In this prospective single-center study, adults with biopsy-confirmed MASH underwent MRI within 6 months of biopsy; healthy volunteers served as controls. The 10-minute protocol comprised multi-echo gradient-echo (for proton density fat fraction, PDFF, and R2*_w) and multi-echo spin-echo (for MCR). Histopathology served as the reference (fibrosis, ballooning, lobular inflammation, steatosis, iron) read blinded to imaging. The primary biomarker was extracellular non-vascular water fraction (ECNV_wf).

Spearman correlations (rs); receiver operating characteristic (ROC) curve analysis with area under the ROC (AUROC); bootstrap validation (1000 replicates); interobserver agreement by intraclass correlation coefficient (ICC).

122 participants: 106 MASH patients (68 women, 38 men; mean age 50.5 ± 11.4 years) and 16 controls (7 women, 9 men; 33.6 ± 6.2 years). ECNV_wf correlated with fibrosis (rs = 0.80; p < 0.001) and detected any fibrosis (F1-F4 vs F0) with AUROC 0.94 (threshold 14.65%; specificity 100%; sensitivity 85%; bootstrap 95% CI: 14.2-14.7%; ordinal AUROC: 0.87). ECNV_wf correlated with ballooning (rs = 0.65; p < 0.001; AUROC 0.92) and inflammation (rs = 0.52; p < 0.001; AUROC 0.94). PDFF correlated with steatosis (rs = 0.72; AUROC 0.96), and R2*_w with iron (rs = 0.52; AUROC 0.97). Interobserver agreement was excellent (ICC: ECNV_wf 0.973; R2*_w 0.979; PDFF 0.994).

Contrast-free MCR enables accurate detection of fibrosis and comprehensive characterization of inflammation, fat, and iron in MASH, warranting further multicenter studies.