Abstract
Purpose :
Retinal neovascularizations (RNV) in proliferative diabetic retinopathy (PDRP) are treated with panretinal photocoagulation (PRP) to prevent vitreous bleeding, retinal detachment and vision loss. Recent studies have shown non-inferiority of intravitreal Anti-Vascular Endothelial Growth Factor (Anti-VEGF) injections compared to PRP. Current gold standard to diagnose and monitor RNV is indirect ophthalmoscopy and fundus photography. The aim of this study was to qualitatively describe changes of individual RNV under anti-VEGF treatment and PRP using Optical Coherence Tomography Angiography (OCTA), to get further understanding in order to discriminate active from inactive RNV.
Methods :
RNV of 11 eyes of 7 patients with PDRP treated with PRP and/or with intravitreal Anti-VEGF were monitored using Zeiss Plexelite 9000 OCTA device, and qualitatively analyzed using the manufacturers viewing software. Focus was on changes of size, shape, and flow information within the structurally visible RNV using both, rendered en face projections and OCT B-Scans with projected flow information.
Results :
Eyes were followed 1 to 7 months after treatment with Anti-VEGF in 9 eyes, and PRP in 2 eyes. After one month, 8 of 9 eyes showed obvious reduction of RNV flow signal. Main changes occurred within the first 3, but continued over 6 months under treatment. Typically, smaller fan-shaped vessels were regressing quickly, while larger loop vessels remained perfused towards the end of the observation period, but showed reduced tortuosity. Structure of RNV in OCT B-Scan regressed slower, in contrast to the flow signal within.
The two eyes under PRP showed progression of projection into the vitreous while en face size of RNVs remained stable.
Conclusions :
OCTA can show detailed RNV changes under treatment, which are mainly within the first 3 months. This shows the feasibility of this concept for both progressing and regressing RNV, opening the possibility to discriminate active from inactive RNVs using OCTA. Due to the small sample size, no statement can be made about clinically significant differences between treatments and cut-offs for differentiation between active and inactive RNVs.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.