A prospective, randomized, double-blind, placebo-controlled study of orbital radiotherapy for Graves' ophthalmopathy

Context: Although widely used for more than 85 years, the efficacy of radiotherapy for Graves' ophthalmopathy (GO) has not been established convincingly. Objective: To evaluate the efficacy of radiotherapy for GO. Design: Prospective, randomized, internally controlled, double-blind clinical trial in a tertiary care academic medical center. Participants: The patients were ethnically diverse males and females over age 30 seen in a referral practice. The patients had moderate, symptomatic Graves' ophthalmopathy (mean clinical activity score, 6.2) but no optic neuropathy, diabetes, recent steroid treatment, previous decompression, or muscle surgery. Forty-two of 53 consecutive patients were enrolled after giving informed consent and fulfilling study entry criteria. Eleven eligible patients declined to participate because of inconvenience, desire for alternative therapy, or concern about radiation. Intervention: One randomly selected orbit was treated with 20 Gy of external beam therapy; sham therapy was given to the other side. Six months later, the therapies were reversed. Main Outcome Measures: Every 3 months for 1 year, we measured the volume of extraocular muscle and fat, proptosis, range of extraocular muscle motion, area of diplopia fields, and lid fissure width. Effective treatment for GO will modify one or more of these parameters. Results: No clinically or statistically significant difference between the treated and untreated orbit was observed in any of the main outcome measures at 6 months. At 12 months, muscle volume and proptosis improved slightly more in the orbit that was treated first. Conclusions: In this group of patients, representative of those for whom radiotherapy is frequently recommended, we were unable to demonstrate any beneficial therapeutic effect. The slight improvement noted in both orbits at 12 months may be the result of natural remission or of radiotherapy, but the changes are of marginal clinical significance.