To evaluate the optical coherence tomography angiography (OCT angiography) appearance of the superficial and deep capillary plexa in eyes with retinal vein occlusion (RVO) and to compare these findings with those of fluorescein angiography (FA) and spectral-domain optical coherence tomography (SD OCT).
Retrospective observational case series.
Patients presenting with RVO to Creteil University Eye Clinic were retrospectively evaluated. All patients had undergone a comprehensive ophthalmic examination including FA, SD OCT, and OCT angiography.
There were 54 (31 male, 57%) RVO patients with a mean age of 70 years. The perifoveal capillary arcade was visible in 52 of 54 eyes (96%) on OCT angiography and in 45 eyes (83%) on FA; this arcade was disrupted in 48 eyes (92%) and 39 eyes (72%) on OCT angiography and FA, respectively ( P = .002). Perifoveal capillary arcade disruption was correlated with peripheral retinal ischemia ( P = .025). Intraretinal cystoid spaces were observed in 34 eyes (68%) using FA, in 40 eyes (76%) using SD OCT, and in 49 eyes (90%) using OCT angiography ( P = .008 for OCT angiography vs SD OCT and P = .001 for OCT angiography vs FA). Retinal capillary network abnormalities were observed in all patients in both superficial capillary plexus and deep capillary plexus on OCT angiography. Nonperfusion grayish areas were more frequent in the deep capillary plexus (43 eyes, 84%) than in the superficial capillary plexus (30 eyes, 59%, P < .001).
OCT angiography can simultaneously evaluate both macular perfusion and edema. For the first time, an imaging technique enables the evaluation of the deep capillary plexus, which appears to be more severely affected than the superficial capillary plexus in RVO.
Retinal vein occlusion (RVO) is the second most common cause of retinal vascular disease worldwide after diabetic retinopathy, and its prevalence increases with age. Macular edema is the main complication of well-perfused forms of RVO, limiting visual recovery in about half the cases. Fluorescein angiography (FA) is the gold standard in detecting and evaluating retinal perfusion in clinical practice. Dye leakage in the late phases of FA, usually associated with the presence of macular cystoid spaces, is used to identify macular edema.
Recently a noninvasive technique, spectral-domain optical coherence tomography (SD OCT), has provided high-resolution images of both retinal and choroidal structures. SD OCT scans of the retina show bands of different reflectivity that appear to correspond to the histologic layers of the human retina, although strict correlation with histology has not yet been demonstrated.
A more recent development in OCT imaging, called “en face” OCT, combines OCT with transverse confocal scans. En face OCT allows the selection of a frontal OCT image (C-scan) of a single retinal layer. This tomographic image has high pixel-to-pixel correspondence with fundus and angiographic images. Large retinal vessels may be visible on en face OCT because of their reflectivity; however, no information is available about blood flow.
A newly developed amplitude decorrelation algorithm provides information about blood flow by comparing 2 consecutive B-scans. The “AngioVue” optical coherence tomography angiography (RTVue XR; Optovue, Inc, Fremont, California, USA) is the first commercially available OCT device able to provide OCT angiography images. As already reported, the split-spectrum amplitude decorrelation algorithm (SSADA) analyzes OCT scans and enables distinction between static and nonstatic tissue with a scale of flow signals of variable intensity. The SSADA algorithm also improves the signal-to-noise ratio in order to minimize bulk axial motions and artifacts within angiography scans. Thus, by calculating the amplitude of decorrelation signal coming from consecutive B-scans, blood flow can be clearly visualized. It has already been demonstrated that a combination of C-scan OCT angiographies and corresponding conventional B-scans provides clear images of both superficial and deep macular capillary plexa.
Recently, Kuehlewein and associates reported a patient with branch RVO evaluated with FA and swept-source OCT microangiography. Areas of nonperfusion following branch RVO (BRVO) could be precisely delineated at several retinal levels using swept-source OCT microangiography. Moreover, De Carlo and associates described an area of diffuse capillary nonperfusion, continuous with the FAZ and telangiectatic vessels, in a case of central RVO (CRVO).
Jia and associates then reported for the first time the clinical applications of optical coherence tomographic angiography in age-related macular degeneration, while Savastano and associates reported its clinical applications in healthy subjects.
The aim of our study was to evaluate the OCT angiography appearance of the superficial capillary plexus and deep capillary plexus in eyes with RVO, and to compare these findings with those of FA and SD OCT.
The case records of 54 consecutive patients with retinal vein occlusion, who presented at the University Eye Clinic of Creteil between October 1, 2014 and March 31, 2015, were retrospectively evaluated. All patients had undergone a comprehensive ophthalmologic examination, including best-corrected visual acuity (BCVA) using an ETDRS scale, biomicroscopy, FA, and SD OCT, as well as OCT angiography, which involved the central 3 × 3 mm area (10 degrees). The study followed the tenets of the Declaration of Helsinki. This retrospective study was performed in accordance with French legislation and after approval by the University Paris Est ethical committee.
Inclusion and Exclusion Criteria
Patients with a clinical diagnosis of RVO involving the macular area, either naïve or already treated, were retrospectively enrolled in this study. RVO was diagnosed using conventional multimodal imaging (FA and SD OCT). Exclusion criteria were diabetic retinopathy, previous retinal surgery, pathologic myopia, or ocular trauma. Patients with poor-quality images on OCT angiography (signal strength index [SSI] lower than 50) owing to eye movements or media opacities were excluded from this study.
Optical Coherence Tomography Angiography
The OCT angiography used in this study (AngioVue; Optovue Inc) analyzed the OCT images by using SSADA. A 3 × 3 mm area, centered on the fovea, was scanned for all the enrolled patients. The device performs each acquisition at a speed of 100 kHz, 70 000 A-scans per second, using a 840 nm superluminescent diode and with a bandwidth of 45 nm; 320 A-scans made up a B-scan while 320 horizontal and vertical lines separated by 9 μm each were sampled in the scanning area in order to form a 3-dimensional data cube. Volumetric raster scans were obtained from 2 horizontal fast transverse scans and 2 vertical fast transverse scans acquired in 3.4 seconds each. The calculated amplitude decorrelation signal from the consecutive B-scans allowed blood flow, and therefore the capillary network, to be clearly visualized.
Image analysis was performed by automated retinal segmentation derived from the machine software, although with user-dependent adjustments. The retinal vascular network contains multiple plexa. On OCT angiography, however, the intermediate and deep plexa cannot be distinguished and so the simplification into 2 main capillary plexa, as described in 1992 by Snodderly, was required. In normal subjects, the superficial capillary plexus can be seen by using a section starting from the internal limiting membrane (ILM) and selecting sufficient thickness to include the ganglion cell layer, while the 2 components of the deep plexus, which bracket the inner nuclear layer, are included in the section defined by the inner border of the inner nuclear layer (INL) and the middle of the outer plexiform layer. Abnormalities of the superficial capillary plexus and deep capillary plexus surrounding the foveal avascular zone were evaluated using OCT angiography by 2 experienced examiners (F.C., A.M.), masked to each other, in order to evaluate each capillary plexus and interobserver reproducibility.
Several parameters were used to evaluate capillary network abnormalities, at both the superficial and deep capillary level on OCT angiography: disruption of the capillary network (presenting as hypointense, grayish areas, with reduced capillary density); capillary dilation and macular shunting vessels; and well-defined black roundish areas without any signal on OCT angiography and probably corresponding to intraretinal cystoid spaces on en face OCT. In the superficial capillary plexus, since it is only visible at this level, we also evaluated the presence of perifoveal capillary arcade disruption, when extending over 1 quadrant of the entire length.
All these OCT angiography observations were carefully recorded for both superficial capillary plexus and deep capillary plexus and then compared with FA and SD OCT findings. Central macular thickness and retinal cystoid spaces were evaluated on SD OCT images. Disruption of the perifoveal arcade, petaloid macular edema, macular leakage, and macular ischemia (presence of macular capillary exclusion) were evaluated on FA. Peripheral ischemia was defined as a nonperfused retinal area, on FA, of 10 disc diameters or more.
BCVA was converted to the logarithm of the minimal angle of resolution (logMAR) for statistical evaluation. The Fisher exact test was used for qualitative values and Mann-Whitney test for quantitative data analysis SPSS statistics version 19.0 (SPSS Inc, an IBM Company, Chicago, Illinois, USA). The McNemar test was used to compare paired data. Each parameter of the conventional multimodal imaging and of OCT angiography was recorded by 2 independent graders (F.C. and A.M.), masked to each other, at different time points and in different orders. This was done to evaluate the interobserver reproducibility; moreover, the same parameters were analyzed twice by 1 grader (F.C.) to evaluate intraobserver reproducibility. Cohen’s kappa was used to evaluate intra- and interobserver reproducibility. P < .05 was considered to be statistically significant.
Fifty-four patients (54 eyes) were enrolled in this study, 29 of whom had CRVO and the remaining 25 BRVO. Three eyes with RVO were excluded from the study because all scans, owing to motion artifacts, resulted in poor-quality images. Mean age was 69.8 ± 0.41 years and mean BCVA was 0.6 logMAR, with no statistically significant difference between the CRVO and the BRVO groups. Table 1 shows the baseline characteristics of the study patients as well as SD OCT and FA findings. There were statistically significant differences between the CRVO and the BRVO groups in the history of previous macular grid photocoagulation (only present in the BRVO group) and the presence of macular shunting vessels, which were more frequent in the case of BRVO (50% vs 21.4% in CRVO, P = .043).
|Total Group |
N = 54
N = 29
N = 25
|Age (y)||69.8 ± 14.3||67.0 ± 16.3||72.9 ± 11.2||.134|
|Male (%)||31 (57)||19 (65)||12 (48)||.271|
|Naïve patient (%)||27 (52)||14 (25)||13 (52)||1|
|Previous peripheral laser (%)||13 (25)||8 (29)||5 (20)||.528|
|Previous focal or grid laser (%)||4 (7.4)||0 (0)||4 (16)||.04*|
|BCVA (logMAR)||0.60 ± 0.41||0.58 ± 0.42||0.40 ± 0.08||.794|
|Central macular thickness (μm)||422 ± 184||420 ± 205||424 ± 162||.951|
|Macular cysts (%)||40 (75.5)||20 (71.4)||20 (80)||.345|
|Hyperreflective dots (%)||30 (55.6)||13 (44.8)||17 (68)||.075|
|Disruption of the ellipsoid zone (%)||8 (15.4)||3 (10.3)||5 (21.7)||.267|
|Pigment epithelium detachment (%)||19 (35.8)||10 (35.7)||9 (36)||.604|
|Epimacular membrane or vitreomacular traction (%)||6 (11.1)||2 (6.9)||4 (16)||.265|
|Perifoveolar capillary arcade disruption (%)||33 (71.7)||18 (75)||15 (68.2)||.746|
|Macular cysts staining (%)||34 (68)||17 (60.7)||17 (77.3)||.24|
|Late macular leakage (%)||34 (66.7)||19 (67.9)||15 (65.2)||1|
|Shunting vessels (%)||18 (34.6)||6 (21.4)||12 (50)||.043*|
|Cotton-wool spots (%)||8 (15.7)||4 (14.3)||4 (17.4)||1|
|Peripheral ischemia (%)||28 (56)||16 (61.5)||12 (50)||.569|
Table 2 shows the superficial capillary and deep capillary plexus findings on OCT angiography in the CRVO and BRVO groups. Capillary abnormalities (including disruption) and grayish areas (areas with reduced capillary density) were observed in all patients in both superficial and deep capillary plexa. Central cystoid spaces in the superficial capillary plexus were more frequently observed in CRVO than in BRVO (92.9% vs 66.7%, P = .02). Cystoid spaces were more numerous in the deep capillary plexus than in the superficial capillary plexus, with a significant difference between the BRVO and CRVO groups (100% vs 85.7%, P < .001).
|Total Group |
N = 54
N = 29
N = 25
|Superficial capillary plexus|
|Disruption of the perifoveolar capillary arcade (%)||48 (92.3)||26 (92.9)||22 (91.7)||.634|
|Capillary abnormalities||52 (100)||28 (100)||24 (100)|
|Central cysts||42 (80.8)||26 (92.9)||16 (66.7)||.02*|
|Nonperfusion grayish areas||30 (56.6)||17 (58.6)||13 (54.2)||.481|
|Deep capillary plexus|
|Capillary abnormalities||51 (100)||28 (100)||23 (100)|
|Central cysts||41 (78.8)||20 (71.4)||21 (87.5)||.141|
|Nonperfusion grayish areas||43 (84.3)||25 (86.2)||18 (81.8)||.48|
|Cysts equal or more numerous in the deep plexus||46 (92)||24 (85.7)||22 (100)||<.001*|
Table 3 compares the capillary layer abnormalities in the superficial capillary and deep capillary plexus. Grayish areas were more frequent in the deep capillary plexus (84.3%) than in the superficial capillary plexus (58.8%, P < .001). Capillary network dilation was more evident in the deep capillary plexus than in the superficial capillary plexus (78.8% vs 63.5%, P = .039). The disruption in capillary network was also more common in the deep capillary plexus than in the superficial capillary plexus, but the difference was not statistically significant (96% vs 92%, P = .5).
|Superficial Capillary NetworkN = 54||Deep Capillary NetworkN = 54||P Value|
|Disruption in capillary network (%)||48 (92.3)||50 (96.2)||.5|
|Capillary dilation (%)||33 (63.5)||41 (78.8)||.039*|
|Central cysts (%)||42 (80.8)||41 (78.8)||.226|
|Nonperfusion grayish areas (%)||30 (58.8)||43 (84.3)||<.001*|
The comparison between FA, SD OCT, and OCT angiography showed that screening for macular cystoid spaces was easier and more effective using OCT angiography ( Tables 1 and 2 ). Intraretinal cystoid spaces in the superficial capillary plexus and/or deep capillary plexus were observed in 68% of eyes using FA (late phase), in 75.5% of eyes using B-scan SD OCT and in 90.4% of eyes using OCT angiography ( P = .008 for the comparison OCT angiography vs SD OCT and P = .001 for OCT angiography vs FA). When the criteria for the diagnosis of macular edema combined the presence of cystoid spaces and the leakage in the late phase of FA, macular edema was detected in 74% of the study eyes using FA; this is comparable to the detection rate of SD OCT but less than that of OCT angiography ( P = .016) ( Figure 1 ).
The perifoveal capillary arcade was visible in almost all the study eyes on OCT angiography (52 eyes, 96.2% vs 45 eyes, 83.3% with FA) and its disruption was observed more frequently using OCT angiography (48 eyes, 92.3%) rather than FA (39 eyes, 71.7%, P = .025 ( Figure 2 ).