Ophthalmic artery stenosis on three-dimensional rotational angiography: Interrater agreement, prevalence, and risk factors.

Background: There is emerging interest in ophthalmic artery (OA) stenosis angioplasty for the treatment of age-related macular degeneration. Three-dimensional rotational angiography (3DRA) could be used during conventional angiography to determine the presence and severity of OA stenosis. In patients who had undergone 3DRA of the internal carotid artery, we aimed to assess the interrater agreement, prevalence, and risk factors for OA stenosis.

Methods: Consecutive patients from two centers who had undergone conventional angiography with 3DRA of the internal carotid arteries were enrolled in this study. 3DRAs were independently double read for the presence of OA stenosis, as defined as narrowing of the proximal OA of at least 50% when compared to the more distal "normal" OA. Interrater agreement for the evaluation of OA stenosis was assessed with the Cohen's kappa coefficient. Univariate and multivariable logistic regression were used to identify potential predictors of OA stenosis.

Results: Three hundred and two patients (97 men; mean ± SD 57.6 ± 13.4 years) were included in the analysis. Cohen's kappa coefficient (95% CI) was 0.877 (0.798-0.956). OA stenosis was present in 45 patients (14.9%). Multiple logistic regression demonstrated that female sex (odds ratio [OR] = 2.70, 95% confidence interval [CI] 1.18-6.09, p = 0.02) and smoking (OR = 2.11, 95% CI 1.10-4.06, p = 0.03) were significant risk factors for OA stenosis. Age, hypertension, diabetes, coronary artery disease, and subarachnoid hemorrhage were not associated with OA stenosis.

Conclusion: The evaluation of OA stenosis on 3DRA had excellent interrater agreement. OA stenosis was common and was associated with smoking and female sex.