T Yodoshi et al. Pediatrics: 2021, 147 (4) e2020009829; DOI: https://doi.org/10.1542/peds.2020-009829. Alternative Etiologies of Liver Disease in Children With Suspected NAFLD
METHODS: Multicenter, retrospective cohort study of patients aged ≤18 years with overweight and obesity and evidence of elevated serum aminotransferases and/or hepatic steatosis on imaging, referred for suspected NAFLD to Cincinnati Children’s Hospital Medical Center (2009–2017) or Yale New Haven Children’s Hospital (2012–2017). Testing was performed to exclude the following: autoimmune hepatitis (AIH), Wilson disease, viral hepatitis (B and C), thyroid dysfunction, celiac disease, α-1 antitrypsin deficiency, and hemochromatosis
RESULTS: A total of 900 children with overweight and obesity (63% boys, 26% Hispanic ethnicity) were referred, with a median age of 13 years (range: 2–18). Most had severe obesity (n = 666; 76%) with a median BMI z score of 2.45 (interquartile range [IQR]: 2.2–2.7). Median alanine aminotransferase level at presentation was 64 U/L (IQR: 42–95). A clinically indicated liver biopsy was performed in 358 children (40%) at a median of 6 months (IQR: 1–14) post initial visit; of those, 46% had confirmed nonalcoholic steatohepatitis. Positive autoantibodies were observed in 13% of the cohort, but none met criteria for AIH. Only 19 (2%) were found to have other causes of liver disease, with no cases of viral hepatitis or Wilson disease detected.
Specific diseases included thyroid dysfunction in 11 (1.2%), celiac disease in 3 (0.4%), A1AT deficiency in 3 (0.4%), and non-Hodgkin’s lymphoma in 1. A prior study had indicated that AIH was the second-most common etiology (Aliment Pharmacol Ther 2013; 38: 1267-77); this study indicated a much higher rate of alternative diagnosis (24%) in children undergoing a liver biopsy for suspected NAFLD (see related blogs: Screening for NAFLD and Concise Review: Fatty Liver in Pediatrics).
My take: The yield of extensive testing in children with suspected fatty liver disease is very low. I suspect that a cost-effective analysis would indicate a much more limited role for further liver evaluations.
Related blog posts: