Prognostic value of non-invasive fibrosis assessment scores in predicting mortality among individuals with metabolic dysfunction-associated steatotic liver disease.

BACKGROUND: To evaluate the prognostic value of nine non-invasive fibrosis scores in predicting all-cause and cardio-cerebrovascular disease (CCD) mortality among individuals with metabolic dysfunction-associated steatotic liver disease (MASLD). METHODS: This study included 4,377 U.S. adults with MASLD, identified using the NHANES 1999–2018 fasting subsample, with follow-up through December 2019. The nine fibrosis scores evaluated were the non-alcoholic fatty liver disease (NAFLD) fibrosis score, Fibrosis-4 (FIB-4), BARD score, aspartate transaminase-to-platelet ratio index (APRI), Forns score, hepatic steatosis index, NAFLD liver fat score, Steatosis-associated Fibrosis Estimator (SAFE) score, and metabolic dysfunction–associated fibrosis 5 (MAF-5). Kaplan-Meier survival curves, Cox proportional hazards models, and random survival forest (RSF) models were used to assess associations and predictive performance of these scores on mortality outcomes. RESULTS: The study cohort had a median age of 52.7 years and was 48.6% male. Over a median follow-up of 8.92 years, 868 deaths occurred, including 289 from CCD. Higher quartiles of fibrosis scores, especially the SAFE score, were significantly associated with elevated mortality risk among individuals with MASLD. Specifically, participants in the highest SAFE score quartile had a 4.4-fold higher risk of all-cause mortality (HR = 4.39, 95% CI: 2.94–6.54) and a 3.9-fold higher risk of CCD mortality (HR = 3.91, 95% CI: 1.69–9.03) compared to those in the lowest quartile. RSF analysis ranked the SAFE score as the most important predictor among the nine fibrosis scores. CONCLUSION: The SAFE score was the most robust predictor of both all-cause and CCD mortality, highlighting its prognostic utility in MASLD.