Determinants and outcome of amiodarone-associated thyroid dysfunction

OBJECTIVE: Amiodarone is frequently associated with thyroid dysfunction. Identifying predictors for amiodarone-associated thyroid dysfunction and assessing treatment outcome may aid clinicians in daily practice. METHODS: We included 303 consecutive patients with amiodarone therapy for cardiac arrhythmias (260 with atrial fibrillation and 43 with ventricular arrhythmias). Thyroid function tests were performed every 6 months. RESULTS: Mean age was 63 ± 12 years and 66% was male. After median follow-up of 3·3 (0·1-24) years, 23 (8%) patients developed amiodarone-associated thyrotoxicosis (incidence rate 1·9 per 100 person years) and 18 (6%) hypothyroidism (incidence rate 1·1 per 100 person years). The only predictor for amiodarone-associated thyrotoxicosis was age 1·4 mU/l at baseline HR = 5·1 (95% CI 1·1-22·4), P = 0·03, left ventricular ejection fraction <45% HR = 3·8 (95% CI 1·1-13·3), P = 0·04 and diabetes mellitus at baseline HR = 3·3 (95% CI 1·1-10·3), P = 0·04. Gender was not a predictor for amiodarone-associated thyroid dysfunction. Five out of 12 (42%) patients with thyrotoxicosis exhibited spontaneous normalization of thyroid function on continuation of amiodarone therapy. Mean time to normalization in the total group was 6·2 ± 3·3 months, with no difference between continuing or discontinuing amiodarone (6·6 ± 3·8 vs 5·8 ± 2·8 months, P = 0·5). CONCLUSIONS: During median follow-up of 3·3 years, the incidence of amiodarone-associated thyrotoxicosis was higher compared to hypothyroidism. Only general predictors for amiodarone-associated thyroid dysfunction were observed. Discontinuation of amiodarone did not influence treatment outcome.