Dissimilar hepatotoxicity profiles of propylthiouracil and methimazole in children

The antithyroid drugs propylthiouracil and methimazole were introduced for clinical use about 60 yr ago and are estimated to be used in more than 6000 children and adolescents per year in the United States. Over the years that these medications have been used, reports of adverse events involving hepatotoxicity have appeared. To date, there has not been a systematic and comparative evaluation of the adverse events associated with antithyroid drug use.

Our objective was to assess safety and hepatotoxicity profiles of propylthiouracil and methimazole by age in the U.S. Food and Drug Administration's Adverse Event Reporting System (AERS).

We used the multi-item gamma-Poisson shrinker (MGPS) data mining algorithm to analyze more than 40 yr of safety data in AERS. MGPS uses a Bayesian model to calculate adjusted observed to expected ratios [empiric Bayes geometric mean (EBGM) values] for every drug-adverse event combination in AERS, focusing on hepatotoxicity events.

MGPS identified higher-than-expected reporting of severe liver injury in pediatric patients treated with propylthiouracil but not with methimazole. Propylthiouracil had a high adjusted reporting ratio for severe liver injury (EBGM 17; 90% confidence interval = 11.5-24.1) in the group less than 17 yr old. The highest EBGM values for methimazole were with mild liver injury in the group 61 yr and older [EBGM 4.8 (3.3-6.8)], which consisted of cholestasis. Vasculitis was also observed for propylthiouracil in children and adolescents, reaching higher EBGM values than hepatotoxicity signals.

MGPS detects higher-than-expected reporting of severe hepatotoxicity and vasculitis in children and adolescents with propylthiouracil but not with methimazole.