Head-to-head comparison of MASH resolution index versus FAST for noninvasive prediction of resolution of MASH on biopsy

Abstract

BACKGROUND AND AIMS: The MASH Resolution Index (MASHResInd) and FibroScan-aspartate aminotransferase (FAST) score have been reported to noninvasively predict the resolution of metabolic dysfunction-associated steatohepatitis (MASH) on histology. This study aimed to examine head-to-head the diagnostic accuracy of the MASHResInd versus FAST score in detecting MASH resolution. APPROACH AND RESULTS: This prospective study included 160 participants (64% female) with biopsy-proven MASH and stage 2 or stage 3 fibrosis from a randomized, multicenter trial of pegozafermin. Participants received either pegozafermin or placebo and underwent contemporaneous liver biopsies, magnetic resonance imaging-proton density fat fraction, and vibration-controlled transient elastography at 2-time points spaced 24 weeks apart. The primary objective was the detection of MASH resolution with no worsening of fibrosis. Diagnostic models were assessed by the area under the receiver operating characteristic curves (AUC) and compared with the Delong test. The median (IQR) age and body mass index of participants were 56.0 (49.0-62.0) years and 36.5 (32.2-40.4) kg/m 2 , respectively. The MASHResInd outperformed FAST for identifying MASH resolution without worsening fibrosis (AUC: 0.83, 95% CI: 0.75-0.91 vs. 0.65, 95% CI: 0.55-0.75, p =0.001). The positive predictive values for MASH resolution were 29% and 26% for the MASHResInd and FAST using rule-in criteria, respectively. The negative predictive values were 100% and 94% for the MASHResInd and FAST using rule-out criteria, respectively. CONCLUSIONS: The MASHResInd is better than the FAST score in the noninvasive prediction of MASH resolution. These data may have implications for clinical practice and trials.