Successful Treatment of Amiodarone-Induced Thyrotoxicosis
Background— Amiodarone-induced thyrotoxicosis (AIT) is a difficult management problem about which there are little published data. We examined whether continuing amiodarone or differentiating AIT into 2 subtypes affected outcome.
Methods and Results— The type and duration of antithyroid treatment and response were recorded in a consecutive series of 28 cases. Comparisons were made between those in whom amiodarone either was continued or stopped and between those with either possible type 1 or type 2 AIT. Of the 28 cases, 5 had spontaneous resolution of AIT; 23 received carbimazole (CBZ) alone as first-line therapy. Eleven achieved long-term euthyroidism off CBZ or on a maintenance dose. Five became hypothyroid and required long-term thyroxine. Five relapsed after stopping CBZ treatment and were rendered euthyroid with either long-term CBZ (n=3) or radioiodine (n=2). Four were intolerant of CBZ and received propylthiouracil (PTU), with good effect in 3. One was resistant to thionamide alone (CBZ then PTU) and responded to adjunctive steroids. No difference in presentation or outcome was noted between those in whom amiodarone was continued or stopped or between possible type 1 or type 2 AIT.
Conclusions— Continuing amiodarone has no adverse influence on response to treatment of AIT. First-line therapy with a thionamide alone is appropriate in iodine-replete areas, thus avoiding potential complications of other drugs. Differentiating between 2 possible types of AIT does not influence management or outcome.
Received December 11, 2001; revision received January 23, 2002; accepted January 24, 2002.
Amiodarone treatment results in a large iodine load and affects thyroid status by decreasing peripheral deiodination of thyroxine (T4) to tri-iodothyronine (T3), leading to an increase in serum T4 and decrease in T3.1,2⇓ Serum thyrotropin (TSH) levels increase in the early phase of treatment (1 to 3 months) and typically return to normal thereafter.3 These changes are found in euthyroid subjects.
Amiodarone also can induce thyroid dysfunction, with the relative proportion of patients developing thyrotoxicosis or hypothyroidism dependent on dietary iodine content. In iodine-replete areas, such as the United Kingdom and United States, about 3% become thyrotoxic,4 with a higher prevalence in iodine-deficient areas.5 Development of thyrotoxicosis in patients taking amiodarone is associated with significant morbidity.6 Withdrawal of amiodarone often is undesirable because it may provoke life-threatening arrhythmias and may worsen cardiovascular manifestations caused by thyrotoxicosis. Even if withdrawal is possible, the half-life of the drug (≈50 days) means that it influences thyroid function for months. This makes amiodarone-induced thyrotoxicosis (AIT) a difficult condition to manage, especially because data on optimal treatment are limited as the result of a lack of controlled trials.
The pathogenesis of AIT is poorly understood, but current opinion suggests that there are 2 forms: type 1 and type 2.4,7⇓ Type 1 AIT occurs in subjects with an abnormal thyroid (goiter or latent autoimmune disease), with the iodine load triggering autonomous thyroid hormone production. Type 2 develops in subjects who have an apparently normal gland7 and may reflect thyroid hormone release from a destructive thyroiditis. Some suggest that type 1 be treated with thionamides combined with potassium perchlorate to deplete intrathyroidal iodine stores, and that type 2 be treated with high-dose glucocorticoids8; however, both perchlorate and glucocorticoids are associated with significant side effects.
We performed a retrospective study of all patients with AIT seen in our thyroid clinic during the last decade. We defined AIT as a new finding of suppressed serum TSH with raised free T4 and free T3 levels during amiodarone treatment.1 All received the thionamide carbimazole (CBZ) alone (starting dose, 20 to 40 mg/d). Propylthiouracil (PTU) was prescribed if CBZ was not tolerated. Our policy of using thionamides as first-line therapy was based on our previously reported outcome in a series of 5 patients.9 The decision about the continuation of amiodarone was made after consideration of the original indication and the availability of an alternative antiarrhythmic agent.
Cardiovascular diagnoses and other drug therapies, symptoms before diagnosis of AIT, biochemical severity, management, and outcome for each patient were evaluated. We subdivided AIT patients into 2 possible subtypes, type 1 and type 2, on the basis of simple clinical and immunologic features.1,4,7⇓⇓ We defined type 1 by the presence of a nodular or diffuse goiter or other features of Graves disease on clinical examination or by the presence of antithyroid peroxidase autoantibodies (found in 3 subjects). Cases without these features were called type 2.
Serum free T3 and T4 and serum TSH were measured using the Bayer ACS 180 and Bayer Adria Centaur System (normal ranges: 3.5 to 6.5 pmol/L, 9 to 20 pmol/L, and 0.4 to 5.5 mU/L, respectively). Thyroid peroxidase antibodies were measured with an agglutination method (Serodia-AMC kit). An antibody titer of 1:400 or more was identified as positive.
Statistical analysis was performed with the SPSS 10.0 package and Mann-Whitney test for comparisons between 2 groups. Variables are expressed as median with interquartile range.
A total of 28 patients (median age 64.1 years [interquartile range, 53 to 72 years]; 4 women, 24 men) had a biochemical diagnosis of AIT. Arrhythmias requiring amiodarone treatment and underlying cardiac diagnosis are shown in the Table. The median cumulative amiodarone dose before onset of AIT was 136.5 g (73 to 837 g), and the daily dose was 200 mg (200 to 200 mg); symptom onset occurred after a median of 24.2 months (7 to 87 months) after amiodarone commencement.
The most common presenting symptoms were weight loss (in 61%) and worsening palpitation (39%). The median serum free T4 concentration at diagnosis of AIT was 48.3 pmol/L (41 to 121 pmol/L), and the mean serum free T3 concentration was 8.2 pmol/L (7 to 35 pmol/L); TSH was undetectable (<0.1 mU/L) in all.
Patient Management and Outcome
Five patients with AIT resolved spontaneously over a median period of 3.3 months (3.0 to 4.7 months). These patients were noted at their first clinic visit to have normal or improving thyroid function tests, despite not receiving treatment, and were observed; all achieved euthyroidism. In 4 of these patients, amiodarone was continued (Figure).
Twenty-three patients began treatment with CBZ alone. The median time to euthyroidism (defined as normal free T4 and free T3 concentrations) was 4.7 months (3 to 7 months). Eleven patients achieved sustained biochemical euthyroidism with CBZ in doses titrated according to free T4; 5 patients (continuing amiodarone) received maintenance CBZ therapy and remained euthyroid, and 5 patients stopped CBZ after a median duration of 7.2 months (6 to 12 months) and remained euthyroid off therapy. One patient underwent total thyroidectomy 11 months after diagnosis of AIT for follicular thyroid carcinoma after being rendered euthyroid with CBZ.
Three patients became hypothyroid on CBZ and remained so after withdrawal; they were treated with maintenance T4. Four (17%) of the 23 patients were intolerant of CBZ and received PTU; one of these responded poorly to PTU, displaying resistance to thionamide treatment. Steroid therapy was added after 3 months thionamide therapy, with good effect. Of the remaining 3 who received PTU, one remained euthyroid off treatment and 2 became hypothyroid off PTU and required T4.
Five patients (22%) had relapse of their AIT after treatment with CBZ alone (median duration of treatment, 9.8 months [6 to 18 months]); 3 responded to a further course of CBZ alone and were euthyroid off CBZ. Two received radioiodine with good effect 13 and 34 months after initial diagnosis of AIT, amiodarone having been stopped 5 and 36 months earlier, respectively.
Amiodarone: Stopped Versus Continued
Amiodarone was continued in 16 patients (14 men, 2 women), most of whom had ventricular tachycardia, and amiodarone was stopped in 12 patients (10 men, 2 women), most of whom had supraventricular dysrhythmias (Table). The total dose of CBZ required to induce euthyroidism was no different if amiodarone was continued (2.0 g [0 to 9 g]) versus stopped (3.3 g [1 to 7 g]) (P=not significant). The rate of improvement in thyroid function tests was likewise no different (serum free T4 at 6 weeks: 25.6 pmol/L in patients in whom amiodarone was continued [16 to 38 pmol/L] versus 22.2 pmol/L in those in whom it was stopped [14 to 35 pmol/L]; serum free T4 at 12 weeks: 18.4 pmol/L in patients in whom amiodarone was continued [14 to 28 pmol/L] versus 14.8 pmol/L in those in whom it was stopped [13 to 18 pmol/L]) (P=not significant). Two (13%) of those who continued amiodarone had relapse of AIT, compared with 3 (25%) who stopped. Spontaneous euthyroidism resulted in 3 (19%) in whom amiodarone was continued, compared with 2 (20%) in whom it was stopped (P=not significant).
Differentiation of Types 1 and 2 AIT
Of the 28 subjects, 14 initially were classified as possible type 1 AIT and 14 as type 2. Women had more type 1 AIT than type 2 (29% versus 0%, P<0.05). Age, symptom duration, and time from commencing amiodarone to onset revealed no differences between types 1 and 2 (60.0 versus 65.7 years, 2.8 versus 1.9 months, 12.1 versus 24.0 months, respectively; P=not significant). The total cumulative amiodarone dosage and cumulative CBZ dosage to achieve euthyroidism also were not significantly different. Only the 12-week free T4 in type 2 AIT was lower than in type 1 (14.6 pmol/L [11 to 16 pmol/L] versus 18.4 pmol/L [15 to 35 pmol/L]) (P=0.02). Because those who became euthyroid spontaneously and those who became hypothyroid after thionamide therapy alone might retrospectively be considered to have AIT type 2, we repeated comparisons, including those 4 patients in the type 2 group originally classified as type 1. Again we found no differences, except for serum free T4 at 12 weeks, which was lower in type 2 (14.9 pmol/L [12 to 18 pmol/L] versus 20.4 pmol/L [16 to 37 pmol/L]) (P=0.03).
AIT presents a therapeutic challenge, and reports of different treatments are unsatisfactory because of patient heterogeneity (especially iodine intake) and small numbers of patients.10,11⇓ We collected data from one of the largest groups with AIT described so far. We found that the majority of patients treated with CBZ or PTU required no additional therapy, regardless of whether amiodarone was stopped, dose and duration of thionamide therapy being similar. This is in keeping with other data suggesting that thionamides are effective while amiodarone is continued.9,12⇓ We suggest that the decision to discontinue amiodarone should be made on cardiological grounds because successful antithyroid treatment does not depend on stopping amiodarone.
We found that there was a female preponderance in those classified as type 1 AIT, perhaps reflecting a female preponderance for autoimmune thyroid disease. No other features were different between the 2 subtypes, suggesting that the clinical presentations of AIT are similar, irrespective of the underlying mechanism. The 12-week serum free T4 was significantly lower in type 2 AIT, perhaps reflecting the self-limiting nature of this illness. No differences in overall outcome between the 2 subtypes were found, suggesting that irrespective of classification, first-line therapy with CBZ is appropriate.
The present findings conflict with the only reported prospective study of AIT treatment, which comprised 24 patients treated in Italy.8 The authors concluded that distinction of AIT is essential for management and suggested that type 1 be treated with both methimazole and potassium perchlorate and type 2 be treated with glucocorticoids. This discrepancy may reflect differences in iodine intake in the 2 areas of study, with the United Kingdom being an iodine-replete area in contrast to Italy.
In conclusion, management of AIT remains a clinical challenge, occurring in a group of patients with underlying cardiovascular abnormalities. Our retrospective study suggests that treatment with a thionamide alone as first-line therapy for AIT, at least in iodine-replete areas such as the United Kingdom and United States, is appropriate, thus avoiding potential side effects of medications such as perchlorate and glucocorticoids. Further investigation of these adjunctive therapies should, however, be considered to determine if they shorten time to euthyroidism—an important consideration in subjects with dysrhythmias. Stopping or continuing amiodarone therapy and differentiating between 2 types of AIT does not influence clinical outcome.
Dr Osman is supported by a British Heart Foundation Fellowship. We thank Jacquie Daykin, Research Nurse, for her contribution.
All authors contributed to study design, data analysis, and writing.
- ↵Newman CM, Price A, Davies DW, et al. Amiodarone and the thyroid: a practical guide to the management of thyroid dysfunction induced by amiodarone therapy. Heart. 1998; 79: 121–127.
- ↵Martino E, Safran M, Aghini-Lombardi F, et al. Environmental iodine intake and thyroid dysfunction during chronic amiodarone therapy. Ann Intern Med. 1984; 101: 28–34.
- ↵Davies PH, Franklyn JA, Sheppard MC. Treatment of amiodarone induced thyrotoxicosis with carbimazole alone and continuation of amiodarone. BMJ. 1992; 305: 224–225.
- ↵Reichert L, de Rooy H. Treatment of amiodarone induced hyperthyroidism with potassium perchlorate and methimazole during amiodarone treatment. BMJ. 1989; 298: 1547–1548.