Parapapillary Choroidal Microvasculature Predicts Diabetic Retinopathy Progression and Diabetic Macular Edema Development: A Three-Year Prospective Study

Purpose: To determine the predictive value of the microcirculation of the optic nerve head by swept-source optical coherence tomography angiography for identifying individuals with high risk of diabetic retinopathy (DR) progression and diabetic macular edema (DME) development.

Design: Prospective observational cohort study.

Methods: A total of 946 patients (1879 eyes) with type 2 diabetes mellitus were recruited who had no DR or mild nonproliferative DR at baseline, and no DME. All subjects underwent 3 × 3 mm swept-source optical coherence tomography angiography centered on the optic nerve head to generate angiograms in 4 layers: radial peripapillary plexus, superficial retinal capillary plexus (SCP), deep retinal capillary plexus, and choriocapillaris (CC). The CC flow deficit percentage (CC FD%), vessel density (VD), and perfusion density (PD) were quantified.

Results: During the 3 consecutive years of follow-up, 312 eyes (16.60%) experienced DR progression and 115 eyes (6.12%) developed DME. The DR progression was related to a lower VD of the SCP (relative risk per standard deviation decrease, 95% confidence interval): 1.30, 1.14-1.48; P < .001), a lower PD of the SCP (1.41, 1.24-1.60; P < .001), a lower VD of the radial peripapillary plexus (1.23, 1.08-1.40; P = .002), and an elevated CC FD% (1.62, 1.40-1.88; P < .001). The DME occurrence was associated with a lower VD of SCP (1.35, 1.09-1.66; P = .005), a lower PD of SCP (1.29, 1.05-1.59; P = .016), and a higher CC FD% (1.29, 1.03-1.61; P < .001). The CC FD% significantly improved the predictive power, with the increase of the C-statistic for DR progression and DME occurrence by 3.83% (P = .002) and 5.24% (P < .001), respectively.

Conclusions: This study provides the first longitudinal evidence suggesting that peripapillary CC FD% can improve the prediction of DR progression and DME development beyond traditional risk factors.