Brachytherapy of Choroidal Melanomas Less Than 10 mm in Largest Basal Diameter: Comparison of 10-mm and 15-mm Ruthenium Plaques.

ABSTRACT

PURPOSE: To compare tumor control, vision, and complications between patients with a choroidal melanoma of <10 mm in largest basal diameter (LBD) irradiated with 10-mm or 15-mm ruthenium plaques. DESIGN: Retrospective, comparative case series. PARTICIPANTS: One hundred sixty-four consecutive patients with a choroidal melanoma of <10 mm in LBD, 76 and 88 treated with the 10-mm and 15-mm plaque, respectively, from 1998-2014 in a national ocular oncology service. METHODS: Diagnosis was based on growth or high-risk characteristics. The apical dose was 100 to 120 Gy aiming to deliver ≥250 Gy to the sclera. Plaque positioning was modeled retrospectively. An increase of ≥0.3 mm in thickness and ≥0.5 mm in LBD indicated local recurrence. Outcomes were compared with cumulative incidence analysis and Cox regression. Median follow-up time for patients still alive was 8.4 years. MAIN OUTCOME MEASURES: Recurrence rate, low vision, blindness, radiation maculopathy, and optic neuropathy. RESULTS: Melanomas treated with the 10-mm plaque were smaller (median thickness, 1.9 mm vs. 2.6 mm; LBD, 7.1 mm vs. 8.6 mm) and located closer to foveola (median, 2.0 mm vs. 2.8 mm) than those treated with the 15-mm plaque (P < 0.001). The 2 plaques provided a safety margin in 43% versus 40% eyes, provided no safety margin to guard foveola in 17% versus 33%, and did not entirely cover tumor mainly close to the disc in 32% versus 18% of eyes, respectively (P = 0.052). The incidence of a local recurrence was comparable (13% vs. 15% at 10 years; P = 0.31) and associated with plaque positioning (hazard ratio [HR], 2.81 for no safety margin; P = 0.041). At 5 years, the incidence of low vision was 14% versus 24%, and that of blindness was 3% versus 6%. Distance to the foveola was associated with loss of both levels of vision (HR, 0.65 per 1 mm vs. 0.68 per 1 mm; P ≤ 0.001 vs. P = 0.004). The incidence of radiation maculopathy was comparable (19% vs. 18% at 5 years), whereas that of optic neuropathy tended to be higher with the 15-mm plaque (2% vs. 9%; P = 0.054). CONCLUSIONS: The 10-mm ruthenium plaque contributes to better visual preservation, particularly with tumors close to fovea, without increase in local recurrence rate, and may therefore be preferable to the 15-mm plaque.