Abstract
INTRODUCTION: The aims of the study were to determine the effect of retinal zone placement on nonperfusion index (NPI) measurements using ultrawide-field fluorescein angiography (FA) in eyes with diabetic retinopathy (DR) and assess its association with disease progression. METHODS: The study analyzed retinal periphery zones based on FA findings in 18 eyes without DR to standardize anatomical boundaries. NPI measurements were then conducted on a separate set of 285 eyes with DR. Eyes were grouped by NPI levels, and Cox regression was used to examine the association between nonperfusion in the mid-peripheral and far peripheral retina and progression to proliferative diabetic retinopathy (PDR). RESULTS: The retinal zones defined by FA (FA-defined zones) were similar in size to a standard 15-mm radius circle centered on the optic disk (14.98 ± 1.40 mm). However, when compared to circles centered on the fovea, the fovea-centered circles were displaced temporally by an average of 3.94 mm, leading to a significant overestimation of the anatomic nasal far periphery (79.10 ± 13.60 mm2 vs. 36.91 ± 10.52 mm2, p < 0.001). Among the eyes with at least 1 year of follow-up (N = 121, 42.5%), those with medium and high mid-peripheral NPI had a significantly higher risk of progression to PDR (HR, 3.44; 95% CI, 1.36-8.70; p = 0.009) when using FA-defined zones. No significant association was found when using fovea-centered zones (HR, 1.39; 95% CI, 0.56-3.46; p = 0.476). CONCLUSION: While the diameter of the retinal measurement zones did not differ significantly, the location of the zone center had a substantial impact on NPI measurements. Using FA-defined zones centered on the optic disk led to lower mid-peripheral and higher far peripheral NPI measurements, providing a more accurate prediction of progression to PDR compared to fovea-centered zone.