Incidence of radiation retinopathy after high-dosage single-fraction gamma knife radiosurgery for choroidal melanoma.

OBJECTIVE: To investigate the incidence and clinical findings of radiation retinopathy after single-fraction high-dose gamma knife radiosurgery for choroidal melanoma. DESIGN: Retrospective noncomparative interventional case series. PARTICIPANTS: Thirty-two patients with choroidal melanoma. METHODS: Review of charts, color fundus photographs, and fluorescein angiograms of 32 choroidal melanoma patients after radiosurgery. All patients were treated with the Leksell gamma knife in one fraction with a marginal dose between 40 and 80 Gy (median, 50 Gy) and were followed for at least 24 months (or until enucleation because of complications secondary to radiation). MAIN OUTCOME MEASURES: Any clinical feature of radiation retinopathy and neovascular glaucoma. RESULTS: During a mean follow-up of 38 months (range, 6-81 months) we found radiation retinopathy in 84% of our patients. The most common findings in these patients were intraretinal hemorrhages with an incidence of 70%, macular edema and capillary nonperfusion in 63%, and hard exudates in 52% of the patients. Less common were microaneurysms in 30% and retinal neovascularization in 22%. The time of onset of the various radiation-associated retinal findings ranged between 1 and 22 months. Forty-seven percent of all patients developed neovascular glaucoma. In our study there was no correlation between radiation dosage applied and clinical findings. CONCLUSIONS: Single-fraction high-dose Leksell gamma knife radiosurgery of choroidal melanomas with a median marginal dose of 50 Gy is highly associated with early radiation retinopathy and with neovascular glaucoma.

High-compared with low-dose radiosurgery for uveal melanomas.

OBJECT: The authors compared the results of gamma knife radiosurgery in patients with uveal melanoma who underwent high-dose (treated from 1992-1995) and low-dose irradiation (treated from 1996-2002). METHODS: Thirty-one patients with uveal melanomas were treated with a mean margin dose of 52.1 Gy (high dose) and 33 with a mean dose of 41.5 Gy (low dose), and results were compared between groups. The technical procedure was the same in each group except for radiation dose. In the low-dose group, complete tumor regression (scar formation) occurred in 12% and in the high-dose group in 26%. Partial regression (reduction of the tumor prominence between 50 and 80%) occurred in 81% of the low-dose group and in 58% of the high-dose group. Neovascular glaucoma as a severe complication developed in 9% of the low-dose group and in 48% of the high-dose group. CONCLUSIONS: Reduction of the margin dose from 52.1 to 41.5 Gy appears to achieve the same rate of tumor regression but is associated with a lower rate of severe side effects such as neovascular glaucoma. The follow-up period in the low-dose group, however, was much shorter.