Current issues of ACP Journal Club are published in Annals of Internal Medicine


Ophthalmopathy was less likely to develop or worsen with methimazole treatment or surgery for Graves hyperthyroidism

ACP J Club. 1992 Nov-Dec;117:84. doi:10.7326/ACPJC-1992-117-3-084

Related Content in the Archives
Ophthalmopathy after treatment for Graves hyperthyroidism

Source Citation

Tallstedt L, Lundell G, Tørring O, et al. Occurrence of ophthalmopathy after treatment for Graves' hyperthyroidism. N Engl J Med. 1992 Jun 25;326:1733-8.



To compare treatment with methimazole, subtotal thyroidectomy, or iodine-131 (131I) on the development and exacerbation of ophthalmopathy in patients with Graves hyperthyroidism.


Randomized controlled trial with a minimum follow-up of 2 years.


University hospital in Sweden.


168 adults (140 women) with hyperthyroid Graves disease. Randomization was by age (35 to 55 y, n = 114; 20 to 34 y, n = 54).


In the older group, 38 patients received methimazole, 37 had surgery, and 39 received 131I therapy. In the younger group, 27 patients received methimazole, and 27 had surgery. Methimazole treatment was 10 mg 4 times per day for 18 months; thyroxine, 0.1 to 0.3 mg per day, added after 3 to 5 weeks to prevent hypothyroidism; and either propranolol, 60 to 240 mg per day, or metoprolol, 50 to 300 mg per day, for several weeks. Surgery was bilateral subtotal thyroidectomy. Thyroxine, 0.1 to 0.3 mg/d, was administered to achieve euthyroidism. The single-dose 131I therapy, based on thyroid size, 24-hour 131I uptake, and estimated effective isotope half-life, was intended to deliver 120 Gy and was repeated as necessary. Thyroxine was given as needed.

Main outcome measures

An ophthalmologist masked to treatment status examined all patients before treatment. Those without ophthalmopathy were examined 1 year later or as indicated; those with ophthalmopathy were examined monthly until they improved or stabilized. An ophthalmopathy-index score was calculated using visual acuity, tonometry, Hertel exophthalmometry, slit-lamp examination, and extraocular muscle function.

Main results

16% of patients in the older group and 7% in the younger group had pretreatment ophthalmopathy. New ophthalmopathy developed in 22 patients (13%) and worsened in 8 (5%). The relative risk (RR) for development or worsening of ophthalmopathy for patients receiving 131I was 3.2 (95% CI 1.1 to 8.8); for older patients who had surgery, 1.5 (CI 0.5 to 5.0); for younger patients who had surgery, 0.8 (CI 0.2 to 3.0); and for both groups assigned to methimazole, 1.0. The probability of developing or worsening ophthalmopathy increased with pretreatment serum triiodothyronine levels.


Ophthalmopathy was less likely to develop or worsen in patients who were treated with methimazole or surgery for hyperthyroid Graves disease than in patients who received iodine-131.

Sources of funding: Swedish Medical Research Council; Gustav V Jubilee Foundation; Foundation of the Karolinska Institute.

For article reprint: Dr. L. Tallstedt, Department of Ophthalmology, Huddinge University Hospital, S-141 86, Huddinge, Sweden.


The study by Tallstedt and colleagues could lead to restricted use of 131I in some patients. Such a restriction is not justified at present for 3 reasons. First, it is impossible to assess ophthalmopathy adequately without orbital computed tomography (CT) scan or ultrasound. Indeed, when CT scans are done on patients with Graves hyperthyroidism with absent or minimal signs, > 90% are shown to have eye-muscle volume increase. Before making definitive conclusions about the effects of antithyroid therapy, a similar study must be done with CT scan or ultrasound assessments. Second, transient inflammatory changes that might be labeled ophthalmopathy occur in Graves hyperthyroidism but are not always associated with true, progressive ophthalmopathy. Patients should, therefore, be studied for longer periods to establish a definitive diagnosis of ophthalmopathy or worsening eye disease. Indeed, the post-131 release of thyroid antigen and the resulting autoimmune response would be transient because the thyroid mass is eventually decreased. Third, after any form of therapy for hyperthyroidism, eye disease improves, making it almost impossible to assess the natural history of the ophthalmopathy.

The rationale for the possible worsening of eye disease with 131I and improvement after thyroidectomy is that levels of thyroid and eye-muscle cross-reactive antibodies might decrease when the antigen mass is surgically removed or that levels might transiently increase after 131I therapy. The authors did not discuss this or offer an alternative explanation for the worsening of eye disease. Although this study was generally well done, the findings of apparent worsening of existing eye disease or development of new eye disease after 131I therapy should not influence the way we treat patients with Graves disease.

Jack Wall, MD
Allegheny-Singer Research InstitutePittsburgh, Pennsylvania, USA