Clin Transl Sci. 2026 Oct;19(10):e70716. doi: 10.1111/cts.70716.
ABSTRACT
Metabolic dysfunction-associated steatotic liver disease (MASLD) is a major cause of liver-related morbidity and mortality worldwide. Although liver biopsy remains the gold standard for assessing fibrosis and predicting outcomes, it is invasive, costly and associated with complications. Therefore, a non-invasive assessment of fibrosis stage or outcome prediction are urgently needed. In a clinical cohort of 87 patients with biopsy-confirmed MASLD, we developed a random forest pipeline to determine potential novel biomarkers of interest for the non-invasive assessment of advanced fibrosis (F3-F4). We investigated the top-ranked Boruta-selected biomarker, epidermal growth factor-like 7 (EGFL7), as a non-invasive test for advanced fibrosis, using Youden's index to establish a threshold. We performed a Kaplan Meier survival analysis and Cox proportional-hazards analysis of EGFL7 in a MASLD subpopulation of the UK Biobank to assess its predictive value for all-cause and liver-related mortality. This study discovered EGFL7 as a promising candidate for fibrosis classification, achieving 83% accuracy and an AUROC of 0.77. Its diagnostic performance was comparable to FIB-4. Furthermore, our analysis of a MASLD subpopulation of the UK Biobank revealed that elevated plasma EGFL7 levels were associated with all-cause mortality (p < 0.0001) and liver-related mortality (p < 0.0001). Cox proportional-hazards analysis revealed patients with a plasma EGFL7 level above 0.12 had a 51% (HR = 1.51, CI = [1.05-2.17], p = 0.026) higher mortality hazard than those with EGFL7 plasma value below 0.12. Our results suggest that circulating levels of plasma EGFL7 may have utility in the diagnosis and prognosis of patients with MASLD, although these finding do not yet establish EGFL7 as a clinically actionable test.
PMID:42768894 | DOI:10.1111/cts.70716