Predictors of enucleation in Coats disease: analysis of 259 eyes of 259 patients at a single center.

PURPOSE

To evaluate the effects of clinical features associated with enucleation in eyes with Coats disease.

METHODS

The medical records of all patients with Coats disease at the Ocular Oncology, Wills Eye Hospital from November 1, 1973, to July 31, 2018, were reviewed retrospectively. The clinical features pertaining to need for ultimate enucleation and time to enucleation were compared.

RESULTS

The records of 351 eyes were reviewed, of which 259 had follow-up at our center and 32 (12%) were managed with enucleation. Reasons for enucleation included neovascular glaucoma (n = 24 [75%]), possible tumor (6 [19%]), and phthisis bulbi (2 [6%]). Compared to nonenucleated eyes, enucleated eyes had more extensive clock hour involvement of telangiectasia (P < 0.001), light bulb aneurysms (P < 0.001), exudation (P < 0.001), and subretinal fluid (P < 0.001). On adjusted analysis by binomial logistic regression, variables predictive of enucleation included presence of iris neovascularization (P = 0.01), ultrasonographic retinal detachment (P = 0.004), open-funnel retinal detachment (P = 0.04), closed-funnel retinal detachment (P = 0.01), ultrasonographic elevation of subretinal fluid by millimeters (P = 0.001), and angiographic extent of light bulb aneurysms by clock hours (P = 0.02). By Kaplan-Meier analysis of 4-year cumulative risk of enucleation, risk factors for enucleation included presence of iris neovascularization (hazard ratio [HR] 31.0; P < 0.001), ultrasonographic retinal detachment (HR 56.2; P < 0.001), open-funnel retinal detachment (HR 2.7; P = 0.01), and closed-funnel retinal detachment (HR 4.5; P < 0.001).

CONCLUSIONS

Clinical features that predict risk of and time to enucleation in eyes with Coats disease include iris neovascularization, ultrasonographic presence and millimeter-elevation of retinal detachment, and angiographic extent of light bulb aneurysms.