Abstract
Purpose :
Patients exhibiting proliferative diabetic retinopathy (PDR) are at high risk of severe vision loss caused by vitreal hemorrhage or tractive retinal detachment. Early and reliable diagnosis and treatment is crucial for functional preservation. For treatment, anti-VEGF injections might replace pan-retinal photocoagulation (PRP) soon, but carry the need for efficient and non-invasive diagnostic markers at each monthly follow-up to decide re-treatment. Due to its inherent advantages, optical coherence angiography (OCT-A) is expected to replace traditional dye-based angiography in clinical routine in many clinical decisions. The aim of this study was to test OCT-A against FA in terms of visualization of signs of NV.
Methods :
In a cross-sectional exploratory study at the Department of Ophthalmology at the Medical University of Vienna, consecutive patients with PDR in at least one eye underwent routine FA (Heidelberg Engineering HRA) and additional OCT-A (Zeiss Plex Elite 9000) the same day. Eyes with poor image quality were excluded. We analyzed the 55° fovea-centered minute 4 FA images and 12x12mm fovea-centered VRI (vitreoretinal interface) OCT-A slabs for either leakage or pre-retinal decorrelation signal, respectively, compatible with NV. Individual NV at the disc (NVD) and NV elsewhere (NVE) were counted in both modalities and the concordance of presence and number of NV was assessed.
Results :
20 eyes of 11 patients were included. Mean age was 52 ± 10 (range 36 – 67) years. Mean disease duration was 23 ± 12 years. Visual acuity was Sn 0.57 ± 0.41. On FA, 19 eyes had NVEs, and 7 eyes had NVDs, on OCT-A 19 and 6, respectively. In OCT-A and FA, overall NVE/NVD mean count was 3.0 (range 0-9) / 0.3 and 3.15 (range 0-8) / 0.35, respectively. The mean proportion of NVs on OCT-A, compared to FA was 91.46% for NVEs, 85.71% for NVDs (missed 1/7), and 88.41% globally. In no eye, NV count was zero in OCT-A and more in FA.
Conclusions :
OCT-A proved to be an efficient modality to capture NVD/NVE in PDR. VRI slabs exhibited decorrelation signal in 88.41% of all NV visible on a central 55° FA image. This finding suggests to perform a combined approach FA+OCT-A at baseline and to assess OCT-A for continuous follow-up. Nevertheless, more patients will have to be included and modality-specific patterns of NV regression will have to be identified and correlated to treatment necessity.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.