Lithium-associated silent thyroiditis: Clinical implications

To the Editor

Lithium has been a well-established treatment for bipolar disorder (BD) for several decades. Two well-known thyroid abnormalities associated with lithium are goitre and hypothyroidism (Barclay et al., 1994). However, hyperthyroidism among lithium-treated patients is relatively rare (Miller and Daniels, 2001). Lithium leads to hyperthyroidism by two mechanisms: direct toxic effect resulting in silent thyroiditis (Miller and Daniels, 2001); and by causing abnormal expansion of the intrathyroidal iodine pool, thereby predisposing to Graves’ disease (Brownlie and Turner, 2011). Silent thyroiditis in patients treated with lithium has an incidence rate of 1.3 cases per 1000 person years, which is 5–43 times higher than the general population (Miller and Daniels, 2001).

We present here a patient with hyperthyroidism while on lithium maintenance therapy and discuss treatment options. Ms B is a 51-year-old woman with an18-year history of BD. She was stable on maintenance lithium carbonate 1350 mg per day and quetiapine 200 mg at night. She was euthyroid until early 2012 when her symptoms started. Her serum lithium level was usually between 0.4 and 0.6 mmol/L throughout those years.

In March 2012, Ms B complained of several months of generalised myalgia, palpitations, hair loss, anxiety, excessive sweating, restlessness, insomnia and hot flushes. On the basis of these symptoms, thyroid function tests were ordered that revealed low thyroid-stimulating hormone 0.02 mIU/L (0.5–4.0), but a normal free thyroxine (T4) 19 pmol/L (10–25). An ultrasound of the neck did not show a thyroid mass. A thyroid nuclear scan showed faint radiotracer uptake with no focal hot nodule. A diagnosis of hyperthyroidism due to silent thyroiditis was made by the endocrinologist.

Ms B was commenced on carbimazole 5 mg twice a day due to severe thyrotoxic symptoms. Lithium was maintained as the patient was concerned about potential relapse of her BD. After 3 months on carbimazole, her thyroid function returned to normal and clinical symptoms of hyperthyroidism resolved. However, subsequent to this period she developed hypothyroidism and carbimazole was ceased. Ms B was commenced on replacement T4 50 μg per day and her thyroid status has been closely monitored since.

Silent thyroiditis associated with lithium use is usually characterised by spontaneously resolving thyrotoxicosis and a very low radioiodine uptake; it is frequently followed by hypothyroidism (Miller and Daniels, 2001). Lithium-induced hyperthyroidism raises interesting management issues in a psychiatrically stable patient with BD. The key decisions would be whether to continue lithium, and when antithyroid medications are warranted.

Literature on the management of lithium-induced hyperthyroidism suggests continuation of lithium therapy given the worsening of thyrotoxicosis on lithium discontinuation (Reus et al., 1979; Thompson and Baylis, 1986). However, two reports (Bernstein and Friedman, 2011; Mizukami et al., 1995) describe lithium cessation in silent thyroiditis due to the potential ongoing toxic effects of lithium on the thyroid gland.

The addition of antithyroid medications depends on the clinical severity of thyrotoxicosis and the aetiology of hyperthyroidism. Graves’ disease is usually protracted, with more severe symptoms than silent thyroiditis, which is more self-limiting in nature (Miller and Daniels, 2001). Thus, antithyroid medications are not usually indicated in silent thyroiditis (Barclay et al., 1994; Brownlie and Turner, 2011) unless the patient suffers severe symptoms of thyrotoxicosis.

On balance, our case report in conjunction with the literature suggests that lithium-associated silent thyroiditis:

• is a transient phenomenon that can be safely managed by continuing lithium for psychiatric indications;

• can occur after several years of treatment;

• may require the addition of antithyroid medications in severe thyrotoxicosis; and

• is often followed by a hypothyroid phase.

On the basis of the issues discussed in this article, clinicians should be aware of this potential long-term adverse effect that can mimic affective states in a patient with BD.