Prevalence and Distribution of Venous Loops and Association with Retinal Ischemia in Diabetic Retinopathy Using Widefield Swept-Source Optical Coherence Tomography Angiography (WF SS-OCTA)

Abstract

Purpose : Retinal venous loops (RVLs) are rare findings in diabetic retinopathy (DR). Bek reported 7.7% prevalence of RVLs in patients with proliferative DR (PDR) using color fundus photography (CFP). However, previously reported prevalence of RVLs may be underestimated due to limited detection of vascular abnormalities on CFP. This study aimed to determine RVL prevalence, distribution and associated microvascular changes using WF SS-OCTA.

Methods : Retrospective, observational study including diabetic patients with proliferative DR (PDR), non-proliferative DR (NPDR) and no DR. All patients were imaged with WF SS-OCTA using Angio 6×6mm and Montage 15×15mm scan protocols centered on the fovea. Images were independently evaluated by two graders for the presence of RVLs and other DR lesions including nonperfusion areas (NPAs) and neovascularization (NV). RVLs were classified as type I or type II according to the branching level of the feeder vessel. Type I RVLs emerge from larger veins encompassing the first and second branching level from the central retinal vein; type II RVLs emerge from the third or higher branching level. A binary logistic regression model (outcome: presence of RVLs) was used for statistical analyses.

Results : One hundred and ninety-five eyes of 132 diabetic patients (n = 31 without DR; 57 NPDR and 107 PDR) were included. Among them, 22 eyes (11%) had RVLs on WF SS-OCTA. The prevalence of RVLs was higher in PDR eyes compared to NPDR eyes (18% vs. 5%, p<0.05, Chi-square test). Type II RVLs were more prevalent than type I RVLs (p<0.001, Chi-square test). In terms of distribution, RVLs were more likely to originate from superior (vs. inferior) and temporal (vs. nasal) veins (p<0.05, Chi-square test). More advanced severity of DR (OR = 4.495, 95% CI: 1.177-17.161, p = 0.028) and younger age (OR = 0.961, 95% CI: 0.926-0.998, p = 0.037) were associated with the presence of RVLs. NPAs and NV area/number were not significantly correlated with the presence of RVLs.

Conclusions : WF SS-OCTA is useful for the identification of RVLs in patients with DR. Severity of DR, but not ischemia-related imaging biomarkers, was associated with the presence of RVLs in our cohort. Further longitudinal studies may be needed to identify the role of RVLs in DR progression and retinal ischemia.

This is a 2021 ARVO Annual Meeting abstract.