Incidence and Risk Factors for Secondary Glaucoma in Eyes with Uveal Melanoma.

Purpose: To estimate incidence of and analyze risk factors for developing secondary glaucoma in eyes with uveal melanoma before and after diagnosis.

Design: A cross-sectional, population-based cohort study.

Participants: Seven hundred eighty-one patients (median age, 64 years; range, 14-93) consecutively diagnosed with uveal melanoma from 1997 to 2012 in a national ocular oncology service, 708 (91%) of whom received ruthenium (50%) or iodine (50%) brachytherapy.

Methods: Patient, tumor, treatment, and follow-up data were collected prospectively. Frequency and associations of melanoma-related glaucoma at tumor diagnosis were assessed. Incidence of developing secondary glaucoma after diagnosis was estimated by Kaplan-Meier analysis. Independent risk factors were modeled using Cox regression.

Main Outcome Measures: Melanoma-related glaucoma and related risk factors.

Results: Forty-five patients (5.8%; 95% confidence interval [CI], 4.2-7.6) had tumor-related secondary glaucoma at diagnosis, 34 (76%) from a narrow-to-closed angle (25 had direct angle invasion) and 10 (22%) from anterior neovascularization. Synchronous metastases were common in patients with initial secondary glaucoma (11% vs. 1.2% with incident glaucoma, P = 0.005). Patients with secondary glaucoma were often male (58% vs. 48% without glaucoma; P = 0.010) and had larger tumors (median thickness, 9.1 vs. 4.0 mm; P < 0.001) involving the ciliary body (43% vs. 21%; P < 0.001) with retinal detachment (53% vs. 30%; P < 0.001). One hundred and sixty-eight patients 165 of which were treated with brachytherapy developed incident tumor- or treatment-related secondary glaucoma a median of 1.7 years (range, 0.1-13.6) after tumor diagnosis. Cumulative proportion of developing secondary glaucoma was 23% (95% CI, 20-27) at 5 years. The most common mechanism was neovascularization in 119 patients (71%; 95% CI, 63-78). By multivariable regression, initial retinal detachment 3 to 4 quadrants (hazard ratio [HR], 2.18; P < 0.001), initial intraocular pressure 17 mmHg or higher (HR, 1.64; P = 0.01), and tumor thickness predicted incident secondary glaucoma.

Conclusions: Secondary glaucoma at initial uveal melanoma diagnosis predicts high risk of synchronous metastases. Although anterior neovascularization is the most common mechanism for secondary glaucoma after diagnosis, other mechanisms such as angle narrowing and anterior chamber hemorrhage are not infrequent. Initial retinal detachment and intraocular pressure with tumor thickness could inform interim assessments of intraocular pressure and neovascularization.