Widefield Swept-Source OCT Angiography Metrics Associated with the Development of Diabetic Vitreous Hemorrhage <i>A Prospective Study</i>
Ying Cui,Ying Zhu,Edward S. Lu,Rongrong Le,Ines Lains,Raviv Katz,Jay C. Wang,Itika Garg,Yifan Lu,Rebecca Zeng,Dean Eliott,Demetrios G. Vavvas,Deeba Husain,Joan W. Miller,Leo A. Kim,David M. Wu,John B. Miller
DOI: https://doi.org/10.1016/j.ophtha.2021.02.020
IF: 14.277
2021-01-01
Ophthalmology
Abstract:Purpose: To investigate the association among widefield swept-source (SS) OCT angiography (OCTA) metrics and systemic parameters and vitreous hemorrhage (VH) occurrence in eyes with proliferative diabetic retinopathy (PDR). Design: Prospective, observational study. Participants: Fifty-five eyes from 45 adults with PDR, with no history of VH, followed up for at least 3 months. Methods: All patients underwent widefield SS OCTA (Montage 15 x 15 mm and high-definition (HD)-51 line scan) imaging. Images were evaluated independently by 2 graders for quantitative and qualitative widefield SS OCTA metrics defined a priori. Systemic and ocular parameters and widefield SS OCTA metrics were screened using least absolute shrinkage and selection operator and logistic or Cox regression for variable selection. Firth's bias-reduced logistic regression models (outcome, occurrence of VH) and Cox regression models (outcome, time to occurrence of VH) were used to identify parameters associated with VH occurrence. Main Outcome Measures: Occurrence of VH. Results: Over a median follow-up of 363 days (range, 28-710 days), 13 of 55 PDR eyes (24%) demonstrated VH during the follow-up period. Presence of extensive neovascularizations (odds ratio, 8.05; 95% confidence interval [CI], 1.43-58.56; P = 0.02), defined as neovascularizations with total area of more than 4 disc diameters, and forward neovascularizations (odds ratio, 5.42; 95% CI, 1.26-35.16; P = 0.02) that traversed the posterior hyaloid face into the vitreous were associated with the occurrence of VH. The presence of flat neovascularizations (odds ratio, 0.25; 95% CI, 0.04-1.01; P = 0.05) confined to the posterior hyaloid face was associated with a lower risk of VH with borderline significance. Similarly, presence of extensive neovascularizations (hazard ratio, 18.24; 95% CI, 3.51-119.47; P < 0.001) and forward neovascularizations (hazard ratio, 9.60; 95% CI, 2.07-68.08; P = 0.002) was associated significantly with time to development of VH. Conclusions: Widefield SS OCTA is useful for evaluating neovascularizations and their relationship with the vitreous. The presence of forward and extensive neovascularizations was associated with the occurrence of VH in patients with PDR. Larger samples and longer follow-up are needed to verify the risk factors and imaging biomarkers for diabetic VH. (C) 2021 by the American Academy of Ophthalmology