Abstract
CONTEXT: Serum thyrotropin and thyroid hormone (TH) levels are routine markers of thyroid function. However, their diagnostic performance is limited under special conditions, such as in amiodarone-induced hyperthyroidism (AIH). Such cases would require the assessment of tissue TH action, which is currently unfeasible. OBJECTIVE: Development of an approach that determines how well serum parameters are reflected in tissue TH action of patients. METHODS: TH-responsive marker genes were identified from human hair follicles (HFs) with next-generation sequencing, validated by quantitative polymerase chain reaction. A classification model was built with these markers to assess tissue TH action and was deployed on amiodarone-treated patients. The impact of amiodarone on tissue TH action was also studied in thyroid hormone action indicator (THAI) mice. RESULTS: The classification model was validated and shown to predict tissue TH status of subjects with good performance. Serum- and HF-based TH statuses were concordant in hypothyroid and euthyroid amiodarone-treated patients. In contrast, amiodarone decreased the coincidence of serum-based and HF-based TH statuses in patients with hyperthyroidism, indicating that AIH is not unequivocally associated with tissue hyperthyroidism. This was confirmed in the THAI model, where amiodarone prevented tissue hyperthyroidism in THAI mice despite high serum free thyroxine. CONCLUSION: We developed a minimally invasive approach using HF markers to assess tissue TH economy that could complement routine diagnostics in controversial cases. We observed that a substantial proportion of patients with AIH do not develop tissue hyperthyroidism, indicating that amiodarone protects tissues from thyrotoxicosis. Assessing tissue TH action in patients with AIH may be warranted for treatment decisions.