Performance of MAST, FAST, and MEFIB in predicting metabolic dysfunction‐associated steatohepatitis

Abstract Background and Aim To identify individuals with metabolic dysfunction‐associated steatohepatitis (MASH) or “at‐risk” MASH among patients with metabolic dysfunction‐associated steatotic liver disease (MASLD), three noninvasive models are available with satisfactory efficiency, which include magnetic resonance imaging [MRI]‐ AST (MAST), FibroScan‐AST (FAST score), and magnetic resonance elastography [MRE] plus FIB‐4 (MEFIB). We aimed to evaluate the most accurate approach for diagnosing MASH or “at‐risk” MASH. Methods We included 108 biopsy‐proven MASLD patients who underwent simultaneous assessment of MRE, MRI proton density fat fraction (MRI‐PDFF), and FibroScan scans. Compared with the histological diagnosis, we analyzed the AUC of each model and assessed the accuracy. Results Our study cohort consisted of 64.8% of MASH and 25.9% of “at‐risk” MASH. When analyzing the performance of each model for the diagnostic accuracy of MASH, we found that the AUC [95% CI] of MAST was comparable to FAST (0.803 [0.719–0.886] vs 0.799 [0.707–0.891], P = 0.930) and better than MEFIB (0.671 [0.571–0.772], P = 0.005). Similar findings were observed in the “at‐risk” MASH patients. The AUCs [95% CI] for MAST, FAST, and MEFIB were 0.810 [0.719–0.900], 0.782 [0.689–0.874], and 0.729 [0.619–0.838], respectively. The models of MAST and FAST had comparable AUCs ( P = 0.347), which were statistically significantly higher than that of MEFIB ( P = 0.041). Additionally, the cutoffs for diagnosis of MASH were lower than “at‐risk” MASH. Conclusion MAST and FAST performed better than MEFIB in diagnosing “at‐risk” MASH and MASH using lower cutoff values. Our findings provided evidence for selecting the most accurate noninvasive model to identify patients with MASH or at‐risk MASH.