Low- versus high-dose radiotherapy for Graves' ophthalmopathy: a randomized, single blind trial

Orbital radiotherapy (Rx) is a commonly used treatment for Graves' ophthalmopathy (GO), but controlled clinical trials evaluating different Rx doses and application forms have not been performed. In euthyroid patients with moderately severe GO, we randomly compared the efficacy and tolerability of three Rx protocols. Orbital Rx (telecobalt) was administered either in 20 divided fractions of 1 Gray (Gy) weekly over 20 weeks (group A) or in 10 fractions of 1 Gy (B) and 2 Gy (C) daily over 2 weeks. Before and 24 weeks after starting Rx, ophthalmic investigation and magnetic resonance imaging were performed. Response to therapy, defined as a significant amelioration of three objective parameters, was noted in 12 A (67%), 13 B (59%), and 12 C (55%) subjects (C vs. A, P = 0.007). Ophthalmic symptoms and signs regressed most in group A; changes in lid fissure width were -1.5, -0.5, and 0 mm in the A, B, and C groups, respectively (A vs. C, P = 0.005), whereas changes in intraocular pressure (upgaze) were -3, +1, and -1.5 mm Hg, respectively (A vs. B, P = 0.002). The median decreases in proptosis were -2 mm (A, P = 0.0001), -1.5 mm (B, P = 0.02), and -1 mm (C, P = 0.007; A vs. C, P = 0.0380. Visual acuity (+0.15; P = 0.02) and eye muscle motility (bulbar elevation, 30 degrees vs. 37 degrees, P = 0.03, A vs. C, P = 0.0020; abduction, 45 vs. 49 degrees, P = 0.02; A vs. C, P = 0.017) improved in group A only. A significant change in all rectus muscle areas was noted in 14 A (78%), 12 B (55%), and 9 C (41%) subjects (C vs. A, P = 0.002). A decrease in the NOSPECS classes was observed in 12 A (67%), 13 B (59%), and 13 C (59%) patients (A vs. B/C, P = 0.01). Rx-induced conjunctivitis was not observed in group A, but was seen in 4 B (18%) and 8 C (36%) subjects (C vs. A, P = 0.003). At 24 weeks, satisfaction rates were 67%, 59%, and 55% in the A, B, and C groups, respectively (C vs. A, P = 0.008). Thus, in patients with moderately severe GO, similar response rates were observed for low and high Rx doses, but the 1 Gy/week protocol was more effective and better tolerated than the short arm regimens.