To the Editor: Rapid-cycling bipolar disorder is notoriously difficult to treat (1). Hypothyroidism has been associated with rapid cycling in patients with bipolar disorder (2). High doses of levothyroxine (supranormal doses that elevate serum thyroxine levels above normal) have been reported effective in the treatment of refractory rapid cycling in preliminary studies (3). In this case a patient with severe treatment-resistant rapid-cycling bipolar disorder was successfully treated with high doses of levothyroxine, as demonstrated by an unplanned on-off-on trial.
Ms. A was a 36-year-old married woman who was disabled by bipolar disorder. A rapid-cycling pattern began in her teens and included hypomanias and depressions characterized by low-energy auditory hallucinations and suicidality. The rapid cycling was unrelated to the effects of antidepressant medications. After her initial hospitalization in 1991 following a suicide attempt, she required multiple hospitalizations despite treatment with fluoxetine, lithium, various antipsychotics, valproic acid, and carbamazepine. There was no history of primary thyroid disease, although Ms. A was taking liothyronine when she was referred to me in 1994.
Her rapid cycling, which occurred every few days to every few weeks, continued despite a treatment regimen of 200 mg/day of clozapine, 1000 mg b.i.d. of valproic acid in therapeutic doses, and 0.15 mg/day of levothyroxine. The results of tests of her thyroid function were normal.
After Ms. A’s levothyroxine dose was increased to 0.25 mg/day in 1995 in an effort to treat her rapid cycling, she had a marked decrease in the frequency and severity of her cyclic episodes. She was proud that she "set her record" for staying out of the hospital. Olanzapine and verapamil therapy were tried, but there was little effect. She was maintained in a mildly hyperthyroid state; she had an increased resting heart rate of 120 bpm and a mild tremor. In 1998 Ms. A had a slightly elevated free thyroxine index of 3.9 ng/dl and a decreased thyroid-stimulating hormone (TSH) level of 0.02 mU/ml.
In March 1999 her internist insisted that her levothyroxine dose be decreased to 0.1 mg/day because of concern that her hyperthyroid condition was interfering with management of her newly diagnosed non-insulin-dependent diabetes mellitus and placing her at risk for osteoporosis. Within several weeks Ms. A developed severe psychotic depression and suicidality, requiring her first hospitalization in 4 years. Her TSH level was elevated, at 10.05 mU/ml, and her free thyroxine index was normal at 2.7 ng/dl. She cycled between depression and hypomania every few days in the hospital. With the approval of a consulting endocrinologist and the agreement of her internist, her dose of levothyroxine was again increased to 0.25 mg/day. Ms. A began to improve within a week and was discharged from the hospital after 14 days.
For the last 7 months she has remained in a nearly euthymic state, with perhaps mild euphoria and much improved functioning, except for a brief period of recurrent cycling that remitted relatively quickly after we increased her levothyroxine dose to 0.3 mg/day. Ms. A continues to be mildly hyperthyroidal; she has a high-normal free thyroxine level of 1.5 ng/dl and a decreased TSH level of less than 0.1 mU/ml.
This case represents, in a sense, an accident of nature that allowed a test in a clinical setting of the efficacy of high doses of levothyroxine for rapid-cycling bipolar disorder in an on-off-on design. The patient’s clinical course was clearly consistent with the hypothesis that such treatment is efficacious. Risks such as tachycardia, tremor, diaphoresis, and possibly osteoporosis must be considered. But for carefully selected patients with refractory rapid cycling who are closely monitored, the potential benefits of high doses of levothyroxine may outweigh the risks (4).