Intravitreal bevacizumab for neovascular glaucoma in uveal melanoma treated by proton beam therapy.

BACKGROUND: To evaluate the efficacy of bevacizumab on reduction of the enucleation rate and control of intraocular pressure (IOP) in neovascular glaucoma (NVG)-complicating proton beam therapy for UM and to identify the determinants of the efficacy of bevacizumab.

METHODS: Retrospective comparative study of patients with rubeosis following proton therapy for uveal melanoma. Patients were divided into two groups: a bevacizumab group and a control group which comprised two subgroups: panretinal photocoagulation (PRP)/cryotherapy and observation subgroups. Bevacizumab was administered by three intravitreal injections at 1-month intervals. A second series of injections was administered when necessary. Data concerning IOP and the secondary enucleation rate were collected and compared between the two groups. Univariate and multivariate analyses were performed to determine predictive factors of response to bevacizumab.

RESULTS: A total of 169 patients who developed rubeosis following proton therapy between 2006 and 2016 were included: 44 patients in the bevacizumab group and 125 in the control group (38 in the PRP/cryotherapy subgroup and 87 in the observation subgroup). The two groups presented the same baseline characteristics apart from hypertension, retro-equatorial site, and proximity of the optic disk, which were more frequent in the control group, while initial retinal detachment and larger tumor volume were more frequent in the bevacizumab group. After a mean follow-up of 31 months, IOP was less than 21 mmHg in 54.54% of patients after IVB versus 72.7% before treatment (p = 0.06). Statistical analysis did not reveal any statistically significant reduction of the enucleation rate in the bevacizumab group compared to the observational group, whereas the PRP/cryotherapy group showed better eye retention rate (p = 0.15). No enucleation was performed when IOP was < 21 mmHg before IVB. Multivariate analysis identified initial IOP < 21 mmHg and UM situated away from the macula as predictive factors of good response to bevacizumab.

CONCLUSION: Despite the improvement of IOP level, intravitreal bevacizumab (IVB) did not reduce the overall enucleation rate in NVG following proton beam therapy. Nevertheless, this treatment was effective in the early phases of NVG or as preventive treatment. PRP remains a valid treatment for NVG.