Prospective head-to-head comparison of non-invasive scores for diagnosis of fibrotic MASH in patients with type 2 diabetes

Background Non-invasive scores have been proposed to identify patients with fibrotic, metabolic dysfunction associated steatohepatitis (MASH), who are at the highest risk of progression to complications of cirrhosis and may benefit from pharmacologic treatments. However, data in type 2 diabetes (T2DM) patients are lacking. The aim of this multicenter prospective study was to compare head-to-head FAST (FibroScan-aspartate aminotransferase [AST]), MAST (magnetic resonance imaging [MRI]-AST), MEFIB (magnetic resonance elastography [MRE] plus FIB-4), and FNI (fibrotic NASH index) for detecting fibrotic MASH in T2DM patients. Methods 330 T2DM outpatients with biopsy-proven metabolic dysfunction associated steatotic liver disease (MASLD) from the QUID-NASH study (NCT03634098), who underwent FibroScan, MRI-PDFF and MRE at the time of liver biopsy (LB) were studied. The main outcome was fibrotic MASH, defined as NAS ≥ 4 (with at least one point each) and fibrosis stage ≥ 2 (centrally reviewed). Results 245 patients (median age 59 years, male 65%, BMI 31 kg/m2; fibrotic MASH 39%) had all data available for scores comparison. FAST and MAST had similar accuracy (AUROCs 0.81 vs. 0.79, p = 0.41) but outperformed FNI (0.74; p = 0.01) and MEFIB (0.68; p < 0.0001). When using original cutoffs, MAST outperformed FAST, MEFIB and FNI when comparing the percentage of correctly classified patients, in whom LB would be avoided (69% vs. 48%, 46%, 39%, respectively; p < 0.001). When using cutoffs specific to our population, FAST outperformed FNI and MAST (56% vs. 40%, and 38%, respectively; p < 0.001). Conclusion Our findings show that FAST, MAST, MEFIB and FNI are accurate non-invasive tools to identify T2DM patients with fibrotic MASH in secondary/tertiary diabetes clinics. Cutoffs adapted to T2DM population should be considered. Trial registration number NCT03634098.