Elevated serum ferritin in non-alcoholic fatty liver disease is not predictive of fibrosis

Roberto Trasolini,Ben Cox,Ciaran Galts,Eric M Yoshida,Vladimir Marquez,Eric M. Yoshida

DOI: https://doi.org/10.3138/canlivj-2021-0002

2021-12-23

Canadian Liver Journal

Abstract:Non-Alcoholic fatty liver disease (NAFLD) is common with widely ranging severity. Non-invasive risk scores for risk stratification are recommended but misclassify a significant proportion of patients. In situations where non-invasive risk scores do not provide guidance, referral is typically made to a Hepatologist for transient elastography or liver biopsy. Serum ferritin is elevated in many patients with NAFLD related to dysmetabolic and inflammatory hyperferritinemia. Ferritin is widely available and part of a standard workup for chronic liver disease. Methods: To explore the association of ferritin and risk of fibrosis in NAFLD, we reviewed patients diagnosed with NAFLD at the hepatology clinic of the Vancouver General Hospital between the years of 2015–2018. We collected data on 317 patients retrospectively assessing for a relationship between serum ferritin and elastography score. Results: 224 patients were included in the final analysis. Median ferritin was 145 μg/L (IQR 249). Median liver stiffness was 5.2 kPa with 14.3% of patients having liver stiffness ≥8.7 kPa and 17.4% ≥ 8.0 kPa. ROC curve analysis using a liver stiffness ≥8.0kPa as a cutoff for F2 fibrosis showed an AUROC of 0.54 for serum ferritin levels. At a cut-off of both 300 μg/L and 450 μg/L median liver stiffness did not differ significantly in those with ferritin above the cutoff (ferritin ≥300 μg/L p = 0.099, ferritin ≥450 μg/L p = 0.12). Ferritin was significantly higher in male patients (198 μg/L vs 91 μg/L p = 0.0001). There was a weak linear association between AST and ferritin levels. Conclusion: In this cohort of 224 patients with NAFLD, serum ferritin was not predictive of significant liver fibrosis.