Letters

Radioiodine and thyroid eye disease

Routine steroid prophylaxis is not yet justified

Authors' reply

Routine steroid prophylaxis is not yet justified

EDITORThe relation between treatment with radioiodine and thyroid eye disease, discussed in Walsh et al's editorial, troubles many endocrinologists and patients.¹ There have been concerns that the use of radioiodine for thyrotoxicosis due to Graves' disease may be associated with a deterioration in ophthalmopathy, raising the question of whether radioiodine is safe for patients with mild ophthalmic Graves' disease. This question has been addressed recently by Bartalena et al, who showed that there is a small but significant risk of deterioration in mild ophthalmopathy after the use of radioiodine and that this risk may be reduced by simultaneous administration of systemic glucocorticoids.²

Walsh et al go further and advocate that high dose prednisone, as used in Bartalena et al's trial, should be used routinely in all patients with mild ophthalmopathy who are to receive radioiodine, to reduce the risk of deterioration in eye disease. Surely this is not yet justified. No account has been taken of the appreciable adverse effects of giving prednisone for three months (typically 30-40 mg/day for the first month and then reducing over the next two months). In Bartalena et al's study under a tenth of patients (7/72) with mild pre-existing ophthalmopathy who received radioiodine had a deterioration that was more than transient and required treatment.

Routine use of glucocorticoids exposes all patients who receive them to important adverse effects, while the benefit is limited to a few. Certain clinical features (for example, mild but active or progressive ophthalmopathy) are likely to mark out those who are at risk. Further studies are needed to examine this and to determine the minimum dose and duration of glucocorticoid treatment that protects against deterioration of eye disease.

At present there is a case for limiting treatment with glucocorticoids to those who have an appreciable symptomatic worsening of ophthalmopathy rather than treating all routinely. Bartalena et al did not go so far as to advocate routine glucocorticoid treatment for all patients with mild ophthalmopathy who receive radioiodine, and with good reason. Clinical trials showing that a treatment is effective are immensely useful but need to be supported by further, balanced evaluation of the risks and benefits of treatment before the original demonstration of efficacy is translated directly into routine clinical practicea message for all clinicians, not just endocrinologists. First do no harm.

James Ahlquist, consultant endocrinologist. 
Endocrine Unit, Southend Hospital SS0 0RY

^a dr.ahlquist{at}southend-hospital.thenhs.com

Authors' reply

EDITORAhlquist suggests that the adverse effects of corticosteroid treatment outweigh the beneficial effect on the course of thyroid eye disease after radioiodine treatment. With regard to patients without pre-existing ophthalmopathy we agree, as the study of Bartalena et al showed a low risk (1%) of severe eye disease developing de novo.¹

Of 72 patients with mild ophthalmopathy at baseline (defined as mild conjunctival oedema and periorbital inflammation) who were not treated with steroids, however, 17 (24%) showed a deterioration in their eye disease after radioiodine treatment. Although in many cases this was transient (lasting two to three months), it is nevertheless likely to have caused distress to those who were affected. Even more importantly, seven (10%) patients had an exacerbation requiring orbital radiotherapy and high dose steroid treatment. Adjuvant steroid treatment at a substantially lower dose reduced the risk of exacerbation of thyroid eye disease to <1%.

We believe that a 24% risk of a short term deterioration in thyroid eye disease and a 10% risk of a more prolonged and severe exacerbation justify the risks of adjuvant, moderated dose corticosteroid treatment. We do not underestimate the problems that some patients experience from prednisolone treatment at a mean dose of 20 mg daily for three months, but similar doses are widely used to treat conditions such as polymyalgia rheumatica and asthma, with few long term adverse effects over this period.

Limiting treatment to patients with mild eye disease (and avoiding radioiodine treatment in patients with moderate, severe, or active eye disease) means that only one patient with Graves' disease in five who are referred for radioiodine will require corticosteroid treatment. At the very least, the 24% risk of exacerbation of thyroid eye disease needs to be fully discussed with the patient. Trials to see if lower steroid doses are effective would be desirable but, in view of the number of patients required, would prove a major undertaking.

We suggest that using appropriate treatment to prevent iatrogenic exacerbation of a disease that is distressing, disfiguring, and difficult to treat is entirely consistent with Ahlquist's philosophy of first do no harm.

John P Walsh, consultant endocrinologist. 
Sir Charles Gairdner Hospital, Nedlands, Western Australia 6009, Australia

Colin M Dayan, consultant senior lecturer. 
University Division of Medicine, Bristol Royal Infirmary, Bristol BS2 8HW