Diabetic Choroidopathy: Choroidal Vascular Density and Volume in Diabetic Retinopathy With Swept-Source Optical Coherence Tomography

Purpose To compare choroidal vascular density (CVD) and volume (CVV) in diabetic eyes and controls, using en face swept-source optical coherence tomography (SS-OCT). Design Prospective cross-sectional study. Methods Setting: Multicenter. Patient Population: Total of 143 diabetic eyes—27 with no diabetic retinopathy (DR), 47 with nonproliferative DR (NPDR), 51 with NPDR and diabetic macular edema (DME), and 18 with proliferative DR (PDR)—and 64 age-matched nondiabetic control eyes. Observation Procedures: Complete ophthalmologic examination and SS-OCT imaging. En face SS-OCT images of the choroidal vasculature were binarized. Main Outcome Measures: CVD, calculated as the percent area occupied by choroidal vessels in the central macular region (6-mm-diameter circle centered on the fovea), and throughout the posterior pole (12 × 9 mm). The central macular CVV was calculated by multiplying the average CVD by macular area and choroidal thickness (obtained with SS-OCT automated software). Multilevel mixed linear models were performed for analyses. Results Compared to controls (0.31 ± 0.07), central macular CVD was significantly decreased by 9% in eyes with NPDR + DME (0.28 ± 0.06; ß = −0.03, P = .02) and by 15% in PDR (0.26 ± 0.05; ß = −0.04, P = .01). The central macular CVV was significantly decreased by 19% in eyes with PDR (0.020 ± 0.005 mm3, ß = −0.01, P = .01) compared to controls (0.025 ± 0.01 mm3). Conclusions Choroidal vascular density and volume are significantly reduced in more advanced stages of diabetic retinopathy. New imaging modalities should allow further exploration of the contributions of choroidal vessel disease to diabetic eye disease pathogenesis, prognosis, and treatment response. To compare choroidal vascular density (CVD) and volume (CVV) in diabetic eyes and controls, using en face swept-source optical coherence tomography (SS-OCT). Prospective cross-sectional study. Setting: Multicenter. Patient Population: Total of 143 diabetic eyes—27 with no diabetic retinopathy (DR), 47 with nonproliferative DR (NPDR), 51 with NPDR and diabetic macular edema (DME), and 18 with proliferative DR (PDR)—and 64 age-matched nondiabetic control eyes. Observation Procedures: Complete ophthalmologic examination and SS-OCT imaging. En face SS-OCT images of the choroidal vasculature were binarized. Main Outcome Measures: CVD, calculated as the percent area occupied by choroidal vessels in the central macular region (6-mm-diameter circle centered on the fovea), and throughout the posterior pole (12 × 9 mm). The central macular CVV was calculated by multiplying the average CVD by macular area and choroidal thickness (obtained with SS-OCT automated software). Multilevel mixed linear models were performed for analyses. Compared to controls (0.31 ± 0.07), central macular CVD was significantly decreased by 9% in eyes with NPDR + DME (0.28 ± 0.06; ß = −0.03, P = .02) and by 15% in PDR (0.26 ± 0.05; ß = −0.04, P = .01). The central macular CVV was significantly decreased by 19% in eyes with PDR (0.020 ± 0.005 mm3, ß = −0.01, P = .01) compared to controls (0.025 ± 0.01 mm3). Choroidal vascular density and volume are significantly reduced in more advanced stages of diabetic retinopathy. New imaging modalities should allow further exploration of the contributions of choroidal vessel disease to diabetic eye disease pathogenesis, prognosis, and treatment response.