Radiation therapy for ocular choroidal neovascularization (phase I/II study): preliminary report

Choroidal neovascularization (CNV) is a major cause of severe loss of visual acuity in some ocular diseases such as age-related macular degeneration (ARMD) and angioid streaks. Laser photocoagulation has been used to treat patients with subfoveal neovascular lesions with well-demarcated boundaries. However, the treatment method is usually associated with a large decrease in visual acuity. Therefore, indications for this treatment are very limited. Recently, some investigators reported the effect of low dose irradiation on the subretinal neovascular membranes in CNV. We conducted a Phase I/II study to determine the toxicity and efficacy of external photon beam radiotherapy in patients with CNV.

Between April, 1994 and July, 1995, 36 patients with choroidal neovascularization (34 with ARMD and 2 with angioid streaks) were treated with radiation therapy. Treatment planning was performed using a CT simulator that enables real-time treatment planning from multiple CT slices. The clinical target volume that included the macula and optic disc received a dose of 10 Gy/5 fractions/1 week (first 18 eyes) or 20 Gy/10 fractions/2 weeks (last 18 eyes). All eyes were irradiated with a single lateral 6 MV photon beam, angled 10 degrees posteriorly to exclude the ipsilateral lens and the contralateral eye from the radiation field. The ipsilateral lens was irradiated with less than 10% of the total reference dose. The field size averaged 3.0 x 2.5 cm. Records of the 17 eyes with CNV referred to our hospital in 1993, which satisfied the eligibility criteria for this study, were retrospectively analyzed for comparison.

There was no significant acute morbidity. All patients were followed regularly by both ophthalmologists and radiation oncologists. Cataract formation after 1 year of the treatment was observed in one patient who had received a dose of 20 Gy. One patient who had received 20 Gy complained of transient dry-eye sensation 2 months after treatment, but this had disappeared spontaneously by the fourth month. The subjective symptoms, visual acuity, and size of the neovascular membrane were evaluated at 6 and 12 months after treatment in each patient. In the group of patients irradiated with 10 Gy and with 20 Gy, respectively, subjective symptoms improved in five and seven eyes, did not change in seven and four eyes, and deteriorated in six and six eyes at 12 months. Although visual acuity was significantly decreased in the control group, the patients in both irradiated groups did not show such a decrease in visual acuity. The size of the neovascular membrane in the control group progressed significantly. However, the patients in the 20 Gy group showed significant regression of the membrane, although those in the 10 Gy group showed no significant change in size.

This Phase I/II study including a dose escalation study showed that radiation therapy seems to be useful for CNV. The dose of 20 Gy in 10 fractions was more useful in treating neovascular membranes than the dose of 10 Gy in five fractions. These results have encouraged us to start a multicenter randomized prospective study of the treatment of CNV with radiation therapy.