To compare macular vasculature in patients with primary open-angle glaucoma (POAG) and atrophic nonarteritic anterior ischemic optic neuropathy (NAION).
Prospective, cross-sectional study.
Thirty-seven eyes with moderate and advanced POAG, 19 eyes with atrophic NAION, and 40 eyes of normal subjects were imaged using optical coherence tomography angiography (OCT-A). Macular ganglion cell complex (GCC) and peripapillary retinal nerve fiber layer (RNFL) thicknesses were measured in addition to macular superficial and deep vasculature after projection removal using custom software.
Linear models showed that while averaged peripapillary RNFL and macular GCC were not different between NAION and POAG eyes, both were significantly thinner than control eyes. Whole image macular superficial vessel density was significantly lower in NAION and glaucoma eyes ( P = .003 and <.001, respectively) than in normal eyes, with lower vessel density in glaucoma than in NAION eyes ( P = .01). Whole image and parafoveal deep macular vessels in glaucoma eyes (21.0%±8.7%, 24.4%±9.6%) were significantly lower than in control eyes (27.4%±8.6%, 31.9%±10.6%) ( P = .01 and P = .01, respectively). No significant differences in deep vessels were observed between NAION and control eyes. Glaucomatous eyes had lower temporal and inferior parafoveal deep vasculature values than NAION eyes ( P = .007 and .03, respectively).
In NAION and POAG with similar RNFL and macular damage, macular OCT-A shows less involvement of superficial and deep vascular plexus in NAION in contrast to POAG, which might show a primary vascular insult in addition to secondary vascular damage due to ganglion cell damage.
Macular superficial vessel density was significantly lower in NAION and glaucoma eyes than in normal eyes, with lower vessel density in glaucoma than in NAION eyes.
Deep macular vessels in glaucoma eyes were significantly lower than in control eyes.
Glaucoma eyes had lower temporal and inferior parafoveal vasculature values than NAION eyes.
Although primary open-angle glaucoma (POAG) consists of progressive loss of retinal ganglion cells and their axons, nonarteritic anterior ischemic optic neuropathy (NAION) is characterized by acute optic nerve damage. Although NAION is associated with hypoperfusion of the short posterior ciliary arteries and infarction in the retrolaminar region of the optic nerve head, the role of the retinal and macular vascular system in the pathogenesis of POAG and NAION is not well understood.
Optical coherence tomography angiography (OCT-A) can assess the circulation in the retina and optic nerve. Peripapillary and macular vasculature compromise in both NAION and glaucoma eyes has been reported using OCT-A. Several studies have found decreased peripapillary vessel density (VD) at the corresponding location of visual field defects in POAG and NAION. We previously reported that the degrees of peripapillary capillary density in these 2 forms of optic nerve damage were not different. It seems that dropout of peripapillary vessels is not specific to POAG or NAION. Similarly, prior studies have evaluated the role of the macular vasculature in both POAG and NAION. Reduced macular VD was shown in chronic NAION, in contrast with earlier work showing no macular vascular changes. However, in those studies, projection artifacts and duplication of superficial vascular patterns in the deeper retinal layers may have obfuscated the interpretation of data, as the authors had mentioned as limitations. ,
In addition, macular vascular dropout in various types of glaucomatous eyes suggests that the vascular system has an important role in the early diagnosis of glaucoma and in understanding its pathogenesis. Some studies showed that vascular abnormality and macular ganglion cell complex (GCC) thickness change might be interdependent, which raises the question as to whether the vessel dropout is a primary event or is the result of ganglion cell damage. Although several studies noted that total macular GCC thickness does not differ between NAION and moderate and severe POAG, , , the amount of damage to the macular superficial and deep plexus has not been clearly defined, particularly after resolving projection artifacts. These 2 different optic neuropathies with similar ganglion cell loss make a model to test whether microvasculature impairment could vary in different diseases and thus provide further evidence that loss of vascular and neural tissue are independent.
The purpose of the current study was to characterize and compare macular and parafoveal superficial and deep VDs and GCC thickness in NAION and POAG eyes after resolving projection.
Material and Methods
Patients with moderate and severe POAG and post-acute NAION who were examined at New York Eye and Ear Infirmary of Mount Sinai and Farabi Eye Hospital between March 2016 and September 2017 were enrolled in this prospective, cross-sectional, comparative study. The study was approved by the research ethics committee of Farabi Eye Hospital and NYEEI Institutional Review Board. Informed consent was obtained from all patients, and all investigations adhered to the tenets of the Declaration of Helsinki. Participants with age ≥18 years, a spherical refraction within ±5.0 diopters (D), and cylinder correction within ±3.0 D were included. Diagnosis of NAION and POAG was established by 2 authors (M.A.F., R.R.) based on evaluation of the patient’s history, examination, and review of diagnostic testing such as OCT and visual fields, if available. Patients with POAG manifested enlargement of the vertical cup-to-disc ratio, diffuse or focal thinning of the neuroretinal rim, and an open iridocorneal angle on gonioscopy. Presence of a pattern standard deviation outside 95% normal limits, which were confirmed on at least 2 consecutive, reliable tests and glaucoma hemifield test outside normal limits were used for the glaucoma diagnosis. Moderate and severe POAG eyes with 24-2 mean deviation (MD) of less than –6 dB were recruited so that the glaucoma and NAION groups would be similar in terms of severity. ,
Chronic NAION was defined as patients having had a history of sudden visual loss >6 months before enrollment, previous documented optic disc edema, developing a pale optic disc with complete resolution of disc edema at the time of the study, absence of proptosis, and an ophthalmologically normal fellow eye. Patients with other ocular or neurologic disease or evidence of giant cell arteritis with a high erythrocyte sedimentation rate and C-reactive protein, or inflammatory optic neuritis were excluded.
The control group comprised subjects with a best corrected visual acuity of ≥20/30, normal optic disc appearance on fundus examination, normal RNFL thickness by OCT, and IOP <21 mm Hg.
All participants underwent a comprehensive ophthalmic examination, which included assessment of visual acuity, IOP measurement by Goldmann applanation tonometry, slit lamp examination, gonioscopy, fundus examination, VF testing by standard automated perimetry (SAP; Humphrey Field Analyzer; 24-2 Swedish interactive threshold algorithm; Carl Zeiss Meditec, Jena, Germany), and OCT-A imaging.
OCT Angiography and Spectral-Domain OCT
All subjects underwent OCT-A and SD-OCT imaging using the AngioVue imaging system (Optovue, Inc, Fremont, California, USA, RTVue XR version 2018.0.0.18) that has been described previously. Macular 6×6-mm scans centered on the fovea were acquired with the OCT-A AngioVue system. Patients with poor image clarity and scans with signal strength index of <40 were excluded. Retinal layers were automatically segmented to visualize the superficial vascular plexus in a slab from the internal limiting membrane to 9 μm above the junction between the inner plexiform layer and the inner nuclear layer, deep retinal vasculatures from 9 μm above the inner plexiform layer–inner nuclear layer junction to 9 μm below the outer plexiform layer and outer nuclear layer. , Blood flow information as a VD map (%) in superficial slabs was obtained in whole macular image and parafoveal regions. Parafoveal VD was measured in an annulus centered on the fovea with an inner diameter of 1 mm and outer diameter of 3 mm. The parafoveal region was divided into 4 sectors (superior, inferior, temporal, and nasal) in addition to superior and inferior hemispheric areas. We then employed customized MATLAB software (The MathWorks, Inc, Natick, Massachusetts, USA) for calculating deep VD after removing the large vessels of superficial layer that had been projected into the deep layer, which is similar to the method we have described previously for removing large peripapillary vessels. The latest commercial software (PAR algorithm) resolves the flow projection issue in the parafovea, but projection removal is not perfect, and duplication of superficial vessels was still visible in whole macular deep images ( Figure ). To calculate deep VD values, thresholding grayscale deep retinal images of OCT-A were performed to create binary images with threshold values, which were changed manually in each image to remove the large projecting vessels. Then, deep VD was calculated in whole-image, whole parafovea, and each quadrant of the parafovea.
A standard 360°, 3.4-mm-diameter circular scan was used to measure RNFL thickness, and the mean and each sector RNFL values were recorded. The macula cube scanning protocol measured the GCC thickness over a 7 mm diameter and parafoveal GCC was measured over a 3 mm diameter centered on the fovea. Total and superior and inferior hemisphere GCC were recorded for both macular and parafoveal areas.
The distribution of continuous variables was assessed by inspecting histograms and using Shapiro-Wilk W tests of normality. Linear mixed modeling was used for the comparison between groups, after accounting for inter-eye correlation and adjusting for age and multiple comparisons with Bonferroni correction. In this method, the correlation in outcomes between the paired eyes of a participant was accounted for by adding a random effect. In addition, Pearson correlation analysis was used to find associations between thickness of ganglion cell complex, visual field MD, visual acuity, and macular vascular densities. Finally, to evaluate the interobserver reproducibility of our deep vessel measurements, 18 NAION eyes, 12 POAG eyes, and 20 normal controls were randomly selected. Analysis was based on 2 independent series of re-evaluations made by 2 different investigators. The absolute agreement of the measurements conducted by the 2 observers were calculated with the intraclass correlation coefficient (ICC) from a 2-way mixed effect model. All statistical analyses were performed with SPSS software, version 22.0 (IBM Corp, Armonk, New York, USA). P values <.05 were considered significant.
Ninety-six eyes were included after excluding 15 eyes because of segmentation errors: 40 control eyes, 19 chronic NAION, and 37 POAG eyes. One NAION patient and 4 POAG patients had both eyes included in the study. Median time elapsed after visual loss in chronic NAION eyes was 28.6 weeks (range 24-34 weeks), respectively. Demographic information and structural RNFL values are summarized in Table 1 . Mean age in the NAION, POAG, and control eyes was not significantly different. RNFL thicknesses in chronic NAION and POAG were not significantly different from each other (64.5±13.6 vs 59.1±17.4, respectively, P = .46), whereas both were thinner than control eyes. Total macular GCC thickness was lower in chronic NAION eyes (70.01±12.4 μm) and POAG eyes (72.71±13.8 μm) than control eyes, without significant difference between chronic NAION and POAG eyes ( P = .46). Parafoveal GCC thickness was also not different between POAG and NAION eyes, and both were lower than in the control group ( Table 1 ).
|cAION vs Control||Glaucoma vs Control||cAION vs Glaucoma|
|Age, y||57.4 ± 12.2||63.6 ± 10.4||56.7 ± 14.7||>.99||.051||.18|
|Visual acuity, logMAR||0.73 ± 0.74||0.17 ± 0.18||0.03 ± 0.005||<.001||.297||<.001|
|Visual field, MD, dB||−14.0±5.3||−12.7 ± 5.5||−0.65±1.5||<.001||<.001||<.99|
|Average RNFL, μm||64.5 ± 13.6||59.10 ± 17.45||104.18 ± 6.92||<.001||<.001||.46|
|Superior RNFL, μm||77.6 ± 18.9||68.94 ± 18.04||127.42 ± 11.77||<.001||<.001||.18|
|Temporal RNFL, μm||53.1 ± 14.1||54.45 ± 17.53||78.93 ± 8.54||<.001||<.001||>.99|
|Inferior RNFL, μm||79.1 ± 21.8||65.64 ± 21.05||128.15 ± 14.33||<.001||<.001||.04|
|Nasal RNFL, μm||55.3 ± 13.9||52.82 ± 14.63||83.05 ± 12.15||<.001||<.001||<.99|
|Total GCC, μm||70.01 ± 12.4||72.71 ± 13.82||99.80 ± 5.60||<.001||<.001||.463|
|Superior GCC, μm||68.94 ± 13.48||72.91 ± 13.81||99.02 ± 5.83||<.001||<.001||.763|
|Inferior GCC, μm||71.11 ± 12.36||72.69 ± 14.65||100.50 ± 5.80||<.001||<.001||.22|
|Parafovea GCC, μm||95.47 ± 16.60||99.24 ± 12.67||121.40 ± 9.54||<.001||<.001||.777|
|Superior parafoveal GCC, μm||97.58 ± 16.45||96.17 ± 14.21||120.73 ± 10.89||<.001||<.001||.629|
|Inferior parafoveal GCC, μm||93.21 ± 18.28||96.17 ± 14.21||122.00 ± 8.54||<.001||<.001||.884|