Superficial and deep vascular structure of the retina in diabetic macular ischaemia: OCT angiography.

Abstract

Diabetic macular ischaemia (DMI) is a major cause of visual loss in patients with diabetic retinopathy. Although the incidence is not well known, some clinical studies have reported an estimate of approximately 7% of all patients with diabetic retinopathy. The DMI is morphologically appreciable by means of fluorescein angiography (FA) that shows an enlargement of the foveal avascular zone (FAZ) as well as the interruption of capillaries all around the foveal region. However, FA needs dye injection to observe retinal circulation and might have significant adverse events ranging from discomfort and nausea to, in rare cases, anaphylaxis. Optical coherence tomography (OCT) has introduced a new method to visualize the retina by means of cross-sectional images with increased speed and sensitivity over time. This method, known as OCT angiography based on split spectrum amplitude decorrelation angiography (SSADA), is able to detect small changes in retinal vascular microcirculation (Jia et al. 2012). The SSADA has been successfully reported for retinal circulation assessment in healthy eyes (Savastano et al. 2015). The capillary perfusion density values in diabetic eyes were evaluated by Agemy et al. (2015). However, they reported limited information available about OCT angiography changes in DMI. The aim of this study was to evaluate the perifoveal vascular structure in DMI eyes using OCT angiography (OCT-A). Twenty eyes of 10 healthy subjects and 20 eyes of 20 DMI eyes were studied. The mean age of healthy subjects was 50.55 ± 11.07 years. The mean age of DMI patients was 59.5 ± 11.07 years. In DMI patients, the following clinical parameters were considered: diabetes duration (years), HbA1C (%), the presence of peripheral neuropathy involving the inferior and upper limbs, the grade of peripheral retinal ischaemia distributed into four quadrants assessed in FA, the maximum of serum glycaemia ever registered and body mass index (BMI). The OCT-A images and corresponding OCT B-scan were assessed in healthy and DMI eyes using an AngioVue system commercially available XR-Avanti SD OCT device (OptoVue, Freemont, CA, USA). All DMI eyes presenting macular ischaemia by FA were analysed by OCT-A for inclusion in this study, and all OCT-A images had to be of sufficient quality to allow visualization of ‘no flow area’ capillaries for grading of the severity of perifoveal vascular structure in healthy and DMI eyes. The protocol for analysis was SSADA in 3 × 3 mm2 scan volume centred on the foveal region. The automatic detection of ‘no flow area’ in healthy and diabetic eyes of the superficial vascular network (SVN) and deep vascular network (DVN) is shown in Figure 1. The perifoveal vascular structure in healthy eyes was 0.24 mm2 (±0.02) and 0.27 mm2 (±0.02) for the SVN and DVN, respectively, while in DMI eyes, the values were 0.35 mm2 (±0.03) and 0.46 mm2 (±0.02) for the SVN and DVN, respectively (difference by t-tests: p < 0.001), for both the SVN and DVN. Box plots of the changes in the DVN and SVN observed in DMI eyes compared to healthy eyes (T = −7.074; p < 0.001) are shown in Figure 2. In diabetic eyes, the mean FAZ by manual measurement performed by image j software (Version 1.49v, Wayne Rasband; National Institutes of Health, Bethesda, MD, USA; http://imagej.nih.gov/ij) in FA (as described by Conrath et al. 2005) and the extent of perifoveal vascular area by OCT-A did not show any significant differences (mean FAZ at FA: 0.264, SD: 0.054; mean perifoveal vascular area at OCT-A: 0.262, SD: 0.02). No significant correlations were found between perifoveal vascular area measurements and systemic clinical parameters, as well as visual acuity. The present study was a qualitative and quantitative assessment of DMI eyes compared to healthy eyes using AngioVue OCT-A imaging. Our study demonstrated the potential clinical usefulness of the AngioVue system, based on SSADA application, to assess the size and features of DMI in both the SVN and DVN. Optical coherence tomography angiography allows to visualize and assess the appearance of the superficial and deep capillary networks by a quantitative method. This new tool is not applicable to the FA. According to Spaide et al. (2015), FA does not allow separate evaluation of the FAZs of the SVN and DVN because the two networks are overlapped. The potential future application of the methods may be the early identification of perifoveal vascular structure enlargement in eyes with no evidence of DMI with FA and to evaluate this area during the clinical follow-up.