Optical Coherence Tomography Angiography Characteristics and Predictors of Visual Outcomes in Patients With Acute and Chronic Nonarteritic Anterior Ischemic Optic Neuropathy

To investigate the correlation between optical coherence tomography angiography (OCTA) characteristics and visual outcomes in patients with acute and chronic nonarteritic anterior ischemic optic neuropathy (NAION

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Optical Coherence Tomography Angiography Characteristics and Predictors of Visual Outcomes in Patients With Acute and Chronic Nonarteritic Anterior Ischemic Optic Neuropathy Yeji Moon, MD, Min Kyung Song, MD, Joong Won Shin, MD, Hyun Taek Lim, MD, PhD Background: To investigate the correlation between optical coherence tomography angiography (OCTA) characteristics and visual outcomes in patients with acute and chronic nonarteritic anterior ischemic optic neuropathy (NAION). Methods: We retrospectively reviewed clinical data and OCTA images of 26 eyes of 26 patients who had been diagnosed with unilateral NAION. OCTA images were acquired from 17 eyes at the acute stage and from 21 eyes at the chronic stage of NAION. We analyzed the peripapillary vessel density (VD) and macular VD in various layers of the retina and choroid for all images. Possible correlations between the OCTA parameters and visual outcomes were investigated. Results: Among the OCTA parameters for the acute stage of NAION, the temporal peripapillary VD was found to be positively correlated with final visual acuity and visual field with statistical significance (P = 0.039 and 0.009, respectively). In the chronic stage of NAION, both peripapillary and superficial macular VDs were positively correlated with visual outcomes. The nasal perifoveal VD in the superficial capillary plexus (SCP) also had a significant correlation with final visual acuity for both acute and chronic stages (the Spearman correlation coefficient = 0.565 and 0.685, respectively). Conclusions: In patients with NAION, significant correlations were found between OCTA parameters and visual outcomes. The temporal peripapillary VD measured during the acute stage was a significant predictor of final visual outcomes. The decreased nasal perifoveal VD in the SCP was strongly associated with poor visual prognosis. Journal of Neuro-Ophthalmology 2021;41:e440–e450 doi: 10.1097/WNO.0000000000001102 © 2020 by North American Neuro-Ophthalmology Society Department of Ophthalmology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea. The authors report no conflicts of interest. Address correspondence to Hyun Taek Lim, MD, PhD, Department of Ophthalmology, Asan Medical Center, University of Ulsan College of Medicine, 388-1 Pungnap-2-dong, Songpa-gu, Seoul, Korea 05505; E-mail: [e-mail] e440 N onarteritic anterior ischemic optic neuropathy (NAION) is the most common type of acute optic neuropathy in patients older than 50 years. It is characterized by acute, painless visual loss, altitudinal visual field defect, and typical ophthalmoscopic findings including optic disc swelling and peripapillary hemorrhage (1). The exact pathogenesis of NAION is unclear, but available evidence indicates that NAION is associated with predisposing factors including constriction of neuronal tissue in small cupless discs, arterial hypertension, arteriosclerosis, and diabetes mellitus (2–5). This optic neuropathy is assumed to be caused by vascular insufficiency of the optic nerve head, which is known to be fed by the posterior ciliary artery (PCA) circulation through the centripetal or recurrent pial branches arising from the peripapillary choroid, the short PCAs, or the circle of Zinn and Haller (6). Optical coherence tomography angiography (OCTA) is a new and noninvasive imaging method that is capable of delineating the microvasculature in 3 dimensions within the retina and around the optic nerve (7). Recently, it has been suggested that peripapillary microvasculature evaluation using OCTA might be useful for characterizing optic neuropathies, especially NAION (8). Furthermore, recent studies have identified correlations between retinal microvasculature and visual function in patients with NAION (9,10). However, most previous studies on OCTA in NAION have used only cross-sectional data acquired at the chronic stage of disease rather than longitudinal data collected at both acute and chronic timepoints. Consequently, these studies could not examine longitudinal changes in the retinal/optic disc microvasculature of NAION over time or evaluate the prognostic value of OCTA associated with final visual outcomes when imaged at the acute stage of NAION. The purpose of this study was to investigate longitudinal changes in peripapillary and macular microvasculature using Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution OCTA in patients with NAION and to determine the relationship between OCTA parameters and final visual outcomes by analyzing the OCTA images obtained at both acute and chronic phases of the disease. This study is a pilot study, which was intended to serve as an introduction to further research work. Despite the small number of participants studied, this research may help not only broaden our horizons on understanding vascular pathogenesis of NAION but also improve our knowledge on the relationship between vascular density and visual function in this optic neuropathy. METHODS Subjects We reviewed medical records of consecutive patients who visited Asan Medical Center (Seoul, Korea) from January 2017 to August 2018 and were diagnosed with unilateral NAION based on standard clinical criteria (11). Patients with evidence of vitreous, retinal, or other optic nerve diseases that could affect central vision or cause visual-field defects, or who presented more than 1 month after visual loss, were excluded. The study was approved by the Institutional Review Board of the Asan Medical Center and abided by the tenets of the Declaration of Helsinki. All subjects underwent a comprehensive ophthalmic examination, including best-corrected visual acuity (BCVA) measurement, slit-lamp biomicroscopy, intraocular pressure measurement, and fundus examination. An automated visual field test was performed (Humphrey Visual Field Analyzer, SITA-Standard 30-2 program; Carl Zeiss Meditec, Dublin, CA). The values of mean deviation (MD) and visual field index (VFI) were used for analysis. OCT and OCTA were performed at the acute and chronic stages of NAION. The acute stage was defined as within 1 month of disease onset and the chronic stage as after more than 3 months from disease onset. The OCT images were obtained using a Cirrus HD OCT system (software version 6.0.2.81; Carl Zeiss Meditec). The thickness of the peripapillary retinal nerve fiber layer (pRNFL) was measured in each of the 4 quadrants of the affected eye around the optic disc (superior [S], inferior [I], nasal [N], and temporal [T]). The thickness of the macular ganglion cell–inner plexiform layer (mGCIPL) was measured in 6 macular sectors (superotemporal [ST], superior [S], superonasal [SN], inferonasal [IN], inferior [I], and inferotemporal [IT]). Optical Coherence Tomography Angiography Image Acquisition and Procession All subjects were examined with the AngioVue OCTA imaging system (software version 2016.2.0.35, Optovue Inc, Fremont, CA). A 6 · 6-mm macular cube scan and a 4.5 · 4.5-mm optic disc cube scan were performed in each Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 patient; the scans were automatically centered at the fovea and the optic disc center by the built-in software. This OCTA system has an 840-nm superluminescent diode with a scanning speed of 70,000 A-scans per second, an axial resolution of 5 mm, and a transverse resolution of 15 mm. A split-spectrum amplitude-decorrelation angiography algorithm was used to identify perfused vessels. Details of this algorithm have been previously described (12). Automatic segmentation was used to analyze the different vascular network layers. Eyes showing a poor image with a signal strength index (SSI) , 45, aberrant segmentation, or movement-related artifacts were excluded. In this study, we used a vessel density (VD) map, in which the VD was calculated as the percentage of vascular areas with blood flow on en face angiograms. Peripapillary parameters were measured as follows: (1) the whole en face image VD (wiVD) was calculated from the 4.5 · 4.5-mm cube scan, and (2) the circumpapillary VD (cpVD) was measured within a 750-mm-wide elliptical annulus extending from the optic disc boundary. These vessel densities were measured in the radial peripapillary capillaries (RPC), defined as the network of capillaries from the inner limiting membrane to the posterior border of the retinal nerve fiber layer. The scanning area was segmented by an annular grid into 8 sectors: nasosuperior (NS), nasoinferior (NI), inferonasal (IN), inferotemporal (IT), temporoinferior (TI), temporosuperior (TS), superotemporal (ST), and superonasal (SN) sectors (Fig. 1A). The average and sectoral cpVDs were automatically calculated. Macular parameters were measured as follows: (1) the wiVD was obtained from a 6 · 6-mm macular scan; (2) the foveal VD was measured within a 1.00-mm circle centered at the fovea; (3) the parafoveal VD was obtained within an annulus centered at the fovea with an inner diameter of 1.00 mm and an outer diameter of 3.00 mm; and finally, (4) the perifoveal VD was obtained within an annulus centered at the fovea with an inner diameter of 3.00 mm and an outer diameter of 6.00 mm. The parafoveal and perifoveal regions were divided into 4 quadrants in each: superior (S), inferior (I), nasal (N), and temporal (T) quadrants (Fig. 1B). All macular parameters were measured in automatically segmented layers of the superficial capillary plexus (SCP), the deep capillary plexus (DCP), and the macular choriocapillaris (mCC). The SCP was defined between 3 mm beneath the internal limiting membrane and 15 mm below the inner border of the IPL, and the DCP was defined as the region between 15 and 70 mm below the inner border of the IPL. The mCC was segmented from 30 to 90 mm below the retinal pigment epithelium (13,14). Statistical Analysis All data with continuous variables are presented as mean ± SD. BCVA was converted to a log of the minimum angle of resolution (logMAR). We used the Kruskal–Wallis test and the Fisher exact test to compare data sets as appropriate. e441 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. The sectorial division analysis of the optical coherence tomography angiography (OCTA) image of the right eye. A. The OCTA en face image of the peripapillary area (4.5 · 4.5 mm) automatically segmented in the radial peripapillary capillary layer. B. The OCTA en face image of the macular area (6 · 6 mm) automatically segmented SCP layer. The same sectorial division analysis was applied to the deep capillary plexus layer and choriocapillaris layer. NS, nasosuperior; NI, nasoinferior; IN, inferonasal; IT, inferotemporal; TI, temporoinferior; TS, temporosuperior; ST, superotemporal; SN, superonasal; S, superior; I, inferior; N, nasal; T, temporal. Correlations were tested using the Spearman correlation coefficient. We considered P values less than 0.05 as statistically significant. All statistical analyses were performed using SPSS version 21.0 (IBM Corp., Armonk, NY). RESULTS A total of 26 eyes from 26 patients with NAION were included in our analysis. Among them, 12 eyes (46.2%) of 12 patients were imaged with OCTA at both the acute and chronic stages of NAION. Additional 5 eyes of 5 patients were scanned only at the acute stage, whereas another 9 eyes of 9 patients were scanned only at the chronic stage. All the OCTA images were included for our analysis. That is to say that we assessed OCTA images from the acute stage in 17 eyes of 17 patients and the chronic stage in 21 eyes of 21 patients with NAION. There were no significant differences in demographic characteristics among the acute- and chronic-stage subjects and total subjects (Table 1). The mean duration of follow-up was 8.5 months (range 6–23). Initial visual function parameters, including BCVA, MD, and VFI, were significantly positively correlated with final visual outcomes. Concerning OCT parameters, there were no significant correlations between the thickness of pRNFL in the acute stage and final visual outcomes. However, mGCIPL thicknesses measured in the acute stage were significantly correlated with final BCVA (BCVA at the final visit) in all macular sectors except for the inferotemporal sector (Table 2). The image quality of OCTA in our study was acceptable. The mean SSIs of the peripapillary OCTA images were 63.9 and 60.1 in the acute and chronic stages, respectively, whereas the mean SSIs of the macular OCTA images were 61.2 and 61.6 in the respective corresponding stages. Tables 3 and 4 show the data of Spearman correlation coefficients between final visual outcomes and the e442 peripapillary/macular VDs measured at the acute and chronic stages of the disease. In the analysis for acute-stage NAION subjects, neither the peripapillary wiVD nor average cpVD showed a significant relationship with final visual outcomes. By contrast, the cpVDs of 3 contiguous temporal sectors (TI, TS, and ST) showed a statistically significant relationship with final VA or visual field outcomes in this pilot study. The correlation coefficients indicated a moderate (0.40–0.59) to strong (0.60–0.79) correlation between temporal cpVDs and visual outcomes (Table 3). With respect to the macular VD parameters, the parafoveal VD of the superior sector was significantly correlated with final visual outcomes in both the SCP and DCP. Besides, perifoveal VDs of superior and nasal sectors were also significantly correlated with final VA in the SCP. All correlation coefficients for macular VDs suggested a moderate correlation between VDs and final visual outcomes. As for the VD of the mCC, none of the microvascular parameters showed a significant correlation with final visual outcomes (Table 3). On the other hand, in the analysis for chronic-stage NAION subjects, there was a strong correlation between the peripapillary wiVD and final visual outcomes. Moreover, most cpVD parameters, including average and temporal contiguous sectoral cpVDs (IT, TI, TS, and ST), showed a significant correlation with final visual outcomes. All correlation coefficients for the cpVDs indicated at least a moderate correlation (Table 4). Concerning the macular VD parameters, most VDs in the SCP, including wiVD, parafoveal, and perifoveal VDs for all sectors, were significantly correlated with final VA or visual field outcomes. In particular, the nasal perifoveal VD was the parameter that showed the highest strength of correlation with final VA (Rho = 20.685) with a statistical significance. Figure 2 highlights the significance of the nasal Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 1. Baseline demographic characteristics of the study subjects Age (mean ± SD)* Sex (male:female)† Hypertension (Y:N)† Diabetes (Y:N)† Hyperlipidemia (Y:N)† Cerebrovascular disease (Y:N)† Heart disease (Y:N)† Malignancy (Y:N)† All Subjects (n = 26) Acute-Stage Subjects (n = 17) Chronic-Stage Subjects (n = 21) P 64.36 ± 8.35 13:13 12:14 10:16 4:22 1:25 4:22 3:23 65.21 ± 8.86 9:8 7:10 6:11 4:13 1:16 3:14 3:14 65.24 ± 8.77 11:10 10:11 8:13 2:19 0:21 2:19 2:19 0.897 1.000 0.950 1.000 0.566 0.729 0.816 0.797 N, no; Y, yes. *The Kruskal–Wallis test. † The Fisher exact test. perifoveal VD as a potential predictor of VA outcome. Three cases of acute NAION presenting with similar degrees of vision loss and diffuse optic disc swelling demonstrated different outcomes of VA in the final follow-up assessment depending on the severity of VD loss in the nasal perifoveal sector during the acute stage. Despite severe initial visual deficits, Case 1 having a well-preserved nasal perifoveal VD at the acute stage exhibited a good visual prognosis, whereas Case 3 having a severe loss of the nasal perifoveal VD at the acute stage displayed a poor visual outcome at 1-year follow-up. None of the VDs measured in the DCP was significantly correlated with final visual outcomes. Similarly, none of the VD parameters measured in the mCC showed a strength of correlation equal to or higher than moderate with final visual outcomes (Table 4). Table 5 illustrates longitudinal changes in the peripapillary VD and macular VD in patients with NAION who underwent serial OCTA from acute to chronic stages. All peripapillary perfusion parameters including wiVD and cpVD showed a statistically significant decrease over the follow-up period from acute to chronic stages of NAION. Macular parameters of the SCP showed a trend of decrease in the VD over the same period, but no statistical significance was found for any retinal sectors. Figure 3 depicts how the peripapillary VD and macular VD had changed over time from 1 week to 1 year after symptom onset as the loss of pRNFL and mGCIPL had gradually developed in 1 TABLE 2. Spearman correlations between clinical data and final visual outcomes (n = 26) BCVA (logMAR) Age (years) Initial BCVA (logMAR) Initial MD (dB) Initial VFI (%) Initial pRNFL thickness (mm) Average Superior Nasal Inferior Temporal Initial mGCIPL thickness (mm) Average Superior Superonasal Inferonasal Inferior Inferotemporal Superotemporal MD (dB) VFI (%) Spearman’s Rho P Spearman’s Rho P Spearman’s Rho P 0.275 0.817 20.514 20.518 0.087 ,0.001 0.004 0.004 0.001 20.399 0.877 0.818 0.498 0.022 ,0.001 ,0.001 20.002 20.439 0.853 0.816 0.496 0.012 ,0.001 ,0.001 20.086 20.231 20.090 0.057 0.173 0.349 0.144 0.342 0.398 0.215 0.147 0.182 20.077 0.125 20.009 0.251 0.202 0.363 0.286 0.484 0.090 0.166 20.093 0.084 20.080 0.341 0.224 0.337 0.352 0.359 20.474 20.518 20.397 20.399 20.537 20.411 20.508 0.037 0.024 0.032 0.030 0.019 0.064 0.027 0.206 0.143 0.261 0.236 0.184 0.063 0.265 0.230 0.305 0.174 0.198 0.256 0.412 0.170 0.263 0.173 0.234 0.246 0.254 0.129 0.302 0.172 0.269 0.200 0.188 0.180 0.324 0.137 Bold values denote statistical significance at the P , 0.05 level. BCVA, best-corrected visual acuity; dB, decibel; logMAR, log of the minimum angle of resolution; MD, mean deviation; mGCIPL, macular ganglion cell–inner plexiform layer; pRNFL, peripapillary retinal nerve fiber layer; VFI, visual field index. Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 e443 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 3. Spearman correlations between peripapillary and macular vessel densities and final visual outcomes in the acute stage of nonarteritic anterior ischemic optic neuropathy (n = 17) BCVA (logMAR) Peripapillary region parameters wiVD cpVD Average Nasosuperior Nasoinferior Inferonasal Inferotemporal Temporoinferior Temporosuperior Superotemporal Superonasal Macular parameters; superficial capillary plexus wiVD Foveal VD Parafoveal VD Temporal Superior Nasal Inferior Perifoveal VD Temporal Superior Nasal Inferior Macular parameters; deep capillary plexus wiVD Foveal VD Parafoveal VD Temporal Superior Nasal Inferior Perifoveal VD Temporal Superior Nasal Inferior Macular parameters; choriocapillaris wiVD Foveal VD Parafoveal VD Temporal Superior Nasal Inferior Perifoveal VD Temporal Superior Nasal Inferior MD (dB) Spearman’s Rho VFI (%) Spearman’s Rho P 20.328 0.116 0.188 0.252 0.052 0.427 20.188 20.427 20.133 20.102 20.114 20.429 20.553 20.074 0.069 0.280 0.109 0.349 0.369 0.356 0.094 0.039 0.409 0.416 0.347 0.273 20.023 0,181 0.082 0.691 0.509 0.497 0.214 0.135 0.223 0.473 0.277 0.394 0.009 0.055 0.050 0.252 0.221 0.231 20.037 0.041 0.129 0.606 0.478 0.273 0.042 0.245 0.260 0.458 0.447 0.337 0.024 0.068 0.195 0.448 20.390 20.238 0.061 0.179 20.211 20.094 0.208 0.359 20.151 20.079 0.281 0.382 20.183 20.560 20.220 0.034 0.241 0.010 0.198 0.449 20.282 20.020 20.208 20.374 0.136 0.470 0.211 0.070 20.231 0.033 20.181 20.303 0.186 0.450 0.243 0.118 20.352 20.552 20.565 20.411 0.083 0.011 0.009 0.051 20.294 20.145 20.139 20.206 0.126 0.290 0.298 0.214 20.241 20.114 20.084 20.127 0.176 0.331 0.375 0.313 20.167 20.239 0.26 0.177 20.187 20.179 0.236 0.246 20.135 20.142 0.302 0.293 20.220 20.583 20.359 20.220 0.199 0.007 0.078 0.199 0.167 0.514 0.130 0.064 0.261 0.017 0.310 0.404 0.185 0.562 0.161 0.124 0.238 0.009 0.269 0.318 20.099 20.273 20.070 20.149 0.352 0.145 0.395 0.284 20.088 20.154 20.338 20.208 0.368 0.277 0.093 0.211 20.038 20.106 20.239 20.160 0.442 0.343 0.178 0.270 0.146 20.185 0.289 0.238 20.154 0.230 0.277 0.187 20.169 0.334 0.258 0.095 0.239 0.122 0.006 0.127 0.178 0.321 0.492 0.313 20.038 0.053 0.119 20.007 0.442 0.420 0.325 0.489 20.077 20.047 0.127 20.065 0.384 0.429 0.314 0.402 0.278 0.173 0.000 0.241 0.140 0.253 0.500 0.176 20.120 0.012 20.127 20.179 0.323 0.481 0.313 0.246 20.140 20.049 20.171 20.194 0.296 0.427 0.256 0.228 P Spearman’s Rho P Bold values denote statistical significance at the P , 0.05 level. BCVA, best-corrected visual acuity; cpVD, circumpapillary VD; dB, decibel; logMAR, log of the minimum angle of resolution; MD, mean deviation; VD, vessel density; VFI, visual field index; wiVD, whole en face image VD. e444 Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 4. Spearman correlations between peripapillary and macular vessel densities and final visual outcomes in the chronic stage of nonarteritic anterior ischemic optic neuropathy (n = 21) BCVA (logMAR) Spearman’s Rho Peripapillary region parameters wiVD cpVD Average Nasosuperior Nasoinferior Inferonasal Inferotemporal Temporoinferior Temporosuperior Superotemporal Superonasal Macular parameters; superficial capillary plexus wiVD Foveal VD Parafoveal VD Temporal Superior Nasal Inferior Perifoveal VD Temporal Superior Nasal Inferior Macular parameters; deep capillary plexus wiVD Foveal VD Parafoveal VD Temporal Superior Nasal Inferior Perifoveal VD Temporal Superior Nasal Inferior Macular parameters; choriocapillaris wiVD Foveal VD Parafoveal VD Temporal Superior Nasal Inferior Perifoveal VD Temporal Superior Nasal Inferior MD (dB) P VFI (%) Spearman’s Rho P Spearman’s Rho P 20.685 ,0.001 0.618 0.002 0.668 0.001 20.485 20.359 20.268 20.391 20.446 20.492 20.339 20.173 20.383 0.018 0.065 0.134 0.049 0.028 0.014 0.072 0.234 0.053 0.643 0.219 0.241 0.314 0.443 0.382 0.582 0.472 0.318 0.001 0.184 0.160 0.095 0.029 0.048 0.004 0.018 0.093 0.680 0.232 0.272 0.390 0.568 0.438 0.630 0.452 0.228 0.001 0.169 0.130 0.050 0.006 0.027 0.001 0.023 0.174 20.528 20.211 0.007 0.179 0.498 20.001 0.011 0.498 0.543 0.053 0.005 0.410 20.358 20.546 20.499 20.478 0.056 0.005 0.011 0.014 0.416 0.614 0.255 0.479 0.030 0.002 0.133 0.014 0.411 0.584 0.246 0.530 0.032 0.003 0.141 0.007 20.464 20.587 20.685 20.492 0.017 0.003 ,0.001 0.012 0.361 0.484 0.555 0.446 0.054 0.013 0.005 0.021 0.386 0.494 0.612 0.545 0.042 0.001 0.002 0.005 20.025 20.276 0.457 0.113 0.113 0.027 0.313 0.453 0.096 0.060 0.340 0.398 20.152 20.149 20.081 0.071 0.255 0.259 0.363 0.380 0.262 0.187 0.192 0.086 0.125 0.208 0.203 0.356 0.226 0.136 0.152 0.015 0,161 0.278 0.255 0.477 0.048 20.150 0.075 0.127 0.419 0.258 0.374 0.291 0.216 0.281 20.036 0.026 0.174 0.109 0.438 0.455 0.164 0.270 20.050 0.000 0.239 0.118 0.414 0.499 20.248 20.379 0.139 0.045 0.266 0.134 0.122 0.281 0.271 0.126 0.117 0.293 20.074 20.352 20.171 20.199 0.375 0.059 0.235 0.194 20.118 0.272 0.386 0.213 0.306 0.116 0.046 0.177 20.089 0.336 0.435 0.263 0.350 0.068 0.028 0.125 20.260 20.186 20.116 20.167 0.127 0.210 0.308 0.234 0.182 0.149 0.223 0.313 0.215 0.259 0.165 0.083 0.177 0.145 0.246 0.288 0.221 0.266 0.142 0.103 Bold values denote statistical significance at the P , 0.05 level. BCVA, best-corrected visual acuity; cpVD, circumpapillary VD; dB, decibel; logMAR, log of the minimum angle of resolution; MD, mean deviation; VD, vessel density; VFI, visual field index; wiVD, whole en face image VD. Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 e445 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 2. Comparison of the peripapillary and macular vessel densities between the acute and chronic stages of NAION. Three index cases of NAION presenting with similar degrees of vision loss and optic disc swelling portrayed different visual outcomes at their final follow-up visits. Case 1, having a well-preserved nasal perifoveal VD during the acute stage, exhibited a good visual prognosis, whereas Case 3, having a prominent loss of the nasal perifoveal VD, displayed a poor visual prognosis. Case 2, having a mild loss of the nasal perifoveal VD, lay in between. e446 Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution TABLE 5. Longitudinal changes in the peripapillary and macular vessel densities in patients undergoing serial OCTA from acute to chronic stages of NAION (n = 12) Peripapillary parameters wiVD* cpVD† Macular parameters: superficial capillary plexus wiVD Foveal VD Parafoveal VD† Perifoveal VD† Macular parameters: deep capillary plexus wiVD Foveal VD Parafoveal VD† Perifoveal VD† Macular parameters: choriocapillaris wiVD Foveal VD Parafoveal VD† Perifoveal VD† Acute Stage Chronic Stage P 44.56 ± 3.89 46.00 ± 5.48 39.49 ± 4.17 38.41 ± 4.14 0.014 0.008 42.79 15.49 44.96 43.46 41.57 12.10 43.98 41.72 4.82 7.19 6.56 3.78 0.239 0.084 0.097 0.065 44.32 ± 7.79 27.77 ± 11.71 48.82 ± 12.80 45.07 ± 8.47 0.216 0.313 0.278 0.138 ± ± ± ± 5.54 9.24 4.17 5.25 41.22 ± 5.36 30.33 ± 12.34 47.70 ± 3.54 40.94 ± 5.97 69.09 67.62 66.24 69.68 ± ± ± ± 3.70 6.92 3.51 3.71 71.54 70.46 69.81 72.16 ± ± ± ± ± ± ± ± 3.07 3.59 4.21 3.94 0.065 0.187 0.053 0.128 Bold values denote statistical significance at the P , 0.05 level. *The unit of the VD is % area. † The VD denotes the average VD for the whole sectors. cpVD, circumpapillary VD; NAION, nonarteritic anterior ischemic optic neuropathy; OCTA, optical coherence tomography angiography; VD, vessel density; wiVD, whole en face image VD. exemplary case. A reduction of the peripapillary VD in OCTA became apparent after 2–4 weeks after the symptom onset and then worsened gradually over another 4–6 months; after that, it tended to remain stable in most cases of our study. Compared with this, a reduction of the macular SCP-VD became apparent even at Week 1 and then was gradually aggravated until 1-year follow-up. In particular, the loss of the macular VD of the SCP was present, although mild, before the loss of mGCIPL was evident. Macular parameters of the DCP showed no significant difference in the VD between the acute and chronic stages of the disease. Choriocapillaris perfusion parameters revealed a slight increase in the macular VD but did not reach statistical significance. DISCUSSION In this study, the relationships between retinal microvascular perfusion measured by OCTA and visual outcomes of NAION have been assessed, with the specific aim to determine whether there are certain OCTA parameters predictive of final visual outcomes in patients with NAION. This study led to 4 main findings: (1) the peripapillary microvascular perfusion (cpVD) in the temporal sector, measured during the acute stage of NAION, showed a significant positive correlation with final visual outcomes; (2) the nasal perifoveal VD measured in the acute-stage SCP also had a significant positive correlation with final visual outcomes; (3) the peripapillary wiVD as well as the cpVD Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 was shown to strongly correlate with final VA and visual field in chronic-stage NAION; (4) there was a significant correlation between VDs in the chronic-stage macular SCP and final visual outcomes, among which the nasal perifoveal VD had the strongest correlation with final VA. Previous studies investigating peripapillary OCTA parameters in NAION have reported a significantly reduced VD in the RPC of chronic-stage (old atrophic) NAION patients compared with healthy individuals (8,15–18). This rarefaction of the RPC found in their patients with chronic NAION was associated with pRNFL and mGCIPL thinning in OCT and poor visual function (15,17). Similarly, we observed that the reduced cpVD in the RPC had a significant correlation with poor visual function in patients with chronic NAION. In our study, however, the cpVD of the temporal sector measured in the acute stage also showed a significant positive correlation with the final visual outcomes. Recently, Sharma et al (8) reported that the temporal sectors were more severely affected than other peripapillary sectors in the acute stage of NAION within 1 week after disease onset, and their 5 of the 6 eyes involving temporal sectors showed poor visual function (logMAR VA $1.0). Based on the findings of the previous studies and our current study, it could be speculated that OCTA may detect the peripapillary microvascular changes even in the acute stage of NAION having a marked optic disc swelling and the rarefaction of temporal peripapillary capillaries may be associated with a poor visual outcome. e447 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 3. Longitudinal changes of the peripapillary (top row) and macular (upper middle row) microvasculature over time in relation to the optic disc (lower middle row) and OCT features (bottom row). These serial data from a 71-year-old woman depict the pattern of retinal microvascular changes along the time course of changes in the pRNFL and mGCIPL. A reduction of the peripapillary VD in OCTA became apparent after 2–4 weeks after symptom onset and then worsened gradually over another 4 months. A reduction of the macular VD in the SCP became apparent even at Week 1 and then was gradually aggravated. The loss of the macular VD can be detected, even before the loss of mGCIPL was evident. It is of note that these peripapillary changes in microvascular perfusion are not limited to NAION. Other optic neuropathies such as glaucoma have shown the decreased VD in the RPC compared with healthy controls, which has been reported as a characteristic feature associated with disease severity (19–22). This association indicates that the rarefaction of the RPC (reduced cpVD) demonstrated in various optic neuropathies including NAION may be a result of an accompanying retinal ganglion cell loss and decreased metabolic demand, which was regarded as a common final pathway of virtually all optic neuropathies. The thinning of mGCIPL secondary to peripapillary axonal degeneration in NAION has already been found to occur within 1 month after disease onset (23–25). The thickness of the papillomacular bundle within the mGCIPL has also been found to be correlated with VA (26). In our study, the thinning of mGCIPL in the acute stage, especially in the nasal sectors of mGCIPL, which include the papillomacular bundle, was strongly correlated with final VA (Table 2). These results were highly consistent with the data from our OCTA analysis. The macular VD in the SCP had a significant correlation with final VA, especially for the nasal perifoveal sectors, when measured during not only the chronic but also the acute stage of NAION. As illustrated in Figure 2, the nasal perifoveal VD might help predict final VA as a potential predictor of VA outcome. e448 The degree of the sectoral perifoveal VD loss in the acute stage was well correlated with final VA. This finding seems plausible, considering that the nasal perifoveal sector includes the papillomacular bundle. As mentioned above, the loss of the peripapillary or macular VD in NAION could be considered, at least partly, the result of axonal degeneration. However, it is unclear at what stage of disease the loss of the VD secondary to axonal degeneration occurs in the retina or can be detected in OCTA. It may seem to take a longer time to occur than the axonal degeneration itself. On this ground, it would be worth mentioning that the loss of the macular VD of the SCP may be manifest in the hyperacute phase of the disease, even before the loss of mGCIPL was evident in the OCT deviation map, as shown in the case of Figure 3. It might be explained partly by the difference in imaging software between OCTA and OCT, where the macular VD map of OCTA directly represents the actual degree of the VD in color code; by contrast, the macular deviation map of OCT shows color value as the probability of abnormality (e.g., the red value indicates occurring in less than 1% of the normal population). It is also possible for the microvascular loss in the SCP located in the nerve fiber layer and the ganglion cell layer to be detected earlier than the loss of mGCIPL becomes obvious. This finding has important implications for clinicians, in that it can provide a potential Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution quantitative parameter predictive of final visual outcome of NAION even in the acute stage of the disease. It has been proposed that the topographic pattern of VD loss was well correlated with visual outcomes, just likely presented in cases of pRNFL or GCL loss in OCT images (9). Our findings further support the idea of the topographic accordance between the pattern of the macular VD and visual field defect. The loss of the superior perifoveal VD in the SCP uniformly has corresponded to the inferior altitudinal visual field defect in our patients, as seen in Figures 2 and 3. In a previous study on atrophic NAION, the macular VD was significantly lower in the SCP and DCP, and the VD was positively correlated with VA. However, in our study, we were not able to demonstrate any significant correlations between the macular VD in the DCP and final visual outcomes. This discrepancy may be due to the small number of samples, technical problems related to automatic segmentation, and a wide spectrum of pathophysiologic mechanisms involved in the development of optic disc swelling. Nevertheless, it might be reasonable to assume that the SCP-VD is more relevant to the pathogenesis of NAION than the DCP-VD because the former is anatomically associated with the retinal nerve layer and the ganglion cell layer. Regarding the analysis of choroidal microvascular in patients with NAION, our study found no significant alteration of the mCC-VD and no significant correlation between the mCC-VD and visual outcomes. There have been only a few studies attempting to analyze the macular choroidal vasculature of NAION. One study has reported that there were no significant differences in the VD of the mCC between patients with NAION and normal controls (9). On the contrary, in another case–control study, the authors have reported a decrease of the macular choroidal VD in patients with NAION compared with normal controls (27). These contradictory results may be attributed to the limited quality of the choroidal OCTA scan and the technical difference of the choroidal segmentation methods. Further studies with more focus on choroidal vasculature in NAION are therefore suggested. This study has several limitations. First, this study involved a relatively small number of subjects as a pilot study, which may limit the statistical strength of the analysis. Second, to investigate the relationships between retinal VDs measured during the acute and chronic stages and visual outcomes, we used the cross-sectional data from the acute and chronic stages independently. However, to assess the serial changes in the retinal VDs in patients with NAION, longitudinal data collected over time from the acute to the chronic stages of the disease in the same subjects would be highly desirable. For this purpose, we collected longitudinal data on clinical information and OCTA images from 12 patients who had been assessed from the onset of NAION to the chronic atrophic stage of Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 the disease. To the best of our knowledge, this is the first study, although preliminary, to investigate the longitudinal changes in the OCTA microvascular parameters from acute to chronic stages in association with final visual outcomes. Finally, the current limits of OCTA imaging technology should be taken into account. Segmentation imperfections and flow projection artifacts could induce measurement errors, especially in deeper layers, such as the DCP and CC. To overcome this limitation, we additionally excluded OCTA images of seemingly aberrant process of creation and the images of low SSI. In conclusion, we found that peripapillary and macular microvascular changes were significantly correlated with final visual outcomes in patients with NAION. In the acute stage, peripapillary perfusion in the temporal sector, measured as the cpVD in the RPC, and macular perfusion in the nasal perifoveal sector of the SCP were identified to be potential parameters correlated with final visual outcomes. In the chronic stage, whole peripapillary and superficial macular perfusion, measured as the wiVD beyond the boundary of the geographic sectors, were significant factors associated with the final VA and visual field. Further studies with larger longitudinal sample size are needed to validate these findings. STATEMENT OF AUTHORSHIP Category 1: a. Conception and design: Y. Moon and H. T. Lim; b. Acquisition of data: Y. Moon, M. K. Song, J. W. Shin, and H. T. Lim; c. Analysis and interpretation of data: Y. Moon and H. T. Lim. Category 2: a. Drafting the manuscript: Y. Moon, M. K. Song, J. W. Shin, and H. T. Lim; b. Revising it for intellectual content: Y. Moon, M. K. Song, J. W. Shin, and H. T. Lim. Category 3: a. Final approval of the completed manuscript: Y. Moon, M. K. Song, J. W. Shin, and H. T. Lim. REFERENCES 1. Arnold AC. Pathogenesis of nonarteritic anterior ischemic optic neuropathy. J Neuroophthalmol. 2003;23:157–163. 2. Cestari DM, Gaier ED, Bouzika P, Blachley TS, De Lott LB, Rizzo JF, Wiggs JL, Kang JH, Pasquale LR, Stein JD. Demographic, systemic, and ocular factors associated with nonarteritic anterior ischemic optic neuropathy. Ophthalmology. 2016;123:2446–2455. 3. Hayreh SS, Joos KM, Podhajsky PA, Long CR. Systemic diseases associated with nonarteritic anterior ischemic optic neuropathy. Am J Ophthalmol. 1994;118:766–780. 4. Lee MS, Grossman D, Arnold AC, Sloan FA. Incidence of nonarteritic anterior ischemic optic neuropathy: increased risk among diabetic patients. Ophthalmology. 2011;118:959–963. 5. Burde RM. Optic disk risk factors for nonarteritic anterior ischemic optic neuropathy. Am J Ophthalmol. 1993;116:759– 764. 6. Hayreh SS. Ischemic optic neuropathy. Prog Retin Eye Res. 2009;28:34–62. 7. Akil H, Falavarjani KG, Sadda SR, Sadun AA. Optical coherence tomography angiography of the optic disc: an overview. J Ophthalmic Vis Res. 2017;12:98–105. 8. Sharma S, Ang M, Najjar RP, Sng C, Cheung CY, Rukmini AV, Schmetterer L, Milea D. Optical coherence tomography e449 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. angiography in acute non-arteritic anterior ischaemic optic neuropathy. Br J Ophthalmol. 2017;101:1045–1051. Augstburger E, Zéboulon P, Keilani C, Baudouin C, Labbé A. Retinal and choroidal microvasculature in nonarteritic anterior ischemic optic neuropathy: an optical coherence tomography angiography study. Invest Ophthalmol Vis Sci. 2018;59:870– 877. Gaier ED, Wang M, Gilbert AL, Rizzo JF III, Cestari DM, Miller JB. Quantitative analysis of optical coherence tomographic angiography (OCT-A) in patients with non-arteritic anterior ischemic optic neuropathy (NAION) corresponds to visual function. PLoS One. 2018;13:e0199793. Characteristics of patients with nonarteritic anterior ischemic optic neuropathy eligible for the Ischemic Optic Neuropathy Decompression Trial. Arch Ophthalmol. 1996;114:1366– 1374. Jia Y, Tan O, Tokayer J, Potsaid B, Wang Y, Liu JJ, Kraus MF, Subhash H, Fujimoto JG, Hornegger J, Huang D. Split-spectrum amplitude-decorrelation angiography with optical coherence tomography. Opt Express. 2012;20:4710–4725. Chansangpetch S, Lin SC. Optical coherence tomography angiography in glaucoma care. Curr Eye Res. 2018;43:1067– 1082. Lauermann JL, Eter N, Alten F. Optical coherence tomography angiography offers new insights into choriocapillaris perfusion. Ophthalmologica. 2018;239:74–84. Liu CH, Kao LY, Sun MH, Wu WC, Chen HS. Retinal vessel density in optical coherence tomography angiography in optic atrophy after nonarteritic anterior ischemic optic neuropathy. J Ophthalmol. 2017;2017:e9632647. Wright Mayes E, Cole ED, Dang S, Novais EA, Vuong L, Mendoza-Santiesteban C, Duker JS, Hedges TR III. Optical coherence tomography angiography in nonarteritic anterior ischemic optic neuropathy. J Neuroophthalmol. 2017;37:358– 364. Hata M, Oishi A, Muraoka Y, Miyamoto K, Kawai K, Yokota S, Fujimoto M, Miyata M, Yoshimura N. Structural and functional analyses in nonarteritic anterior ischemic optic neuropathy: optical coherence tomography angiography study. J Neuroophthalmol. 2017;37:140–148. Higashiyama T, Ichiyama Y, Muraki S, Nishida Y, Ohji M. Optical coherence tomography angiography in a patient with e450 19. 20. 21. 22. 23. 24. 25. 26. 27. optic atrophy after non-arteritic anterior ischaemic optic neuropathy. Neuroophthalmology. 2016;40:146–149. Ghasemi Falavarjani K, Tian JJ, Akil H, Garcia GA, Sadda SR, Sadun AA. Swept-source optical coherence tomography angiography of the optic disk in optic neuropathy. Retina. 2016;36(suppl 1):S168–S177. Chen JJ, AbouChehade JE, Iezzi R Jr, Leavitt JA, Kardon RH. Optical coherence Angiography demonstration of retinal changes from chronic optic neuropathies. Neuroophthalmology. 2017;41:76–83. Wang X, Jia Y, Spain R, Potsaid B, Liu JJ, Baumann B, Hornegger J, Fujimoto JG, Wu Q, Huang D. Optical coherence tomography angiography of optic nerve head and parafovea in multiple sclerosis. Br J Ophthalmol. 2014;98:1368–1373. Mammo Z, Heisler M, Balaratnasingam C, Lee S, Yu DY, Mackenzie P, Schendel S, Merkur A, Kirker A, Albiani D, Navajas E, Beg MF, Morgan W, Sarunic MV. Quantitative optical coherence tomography angiography of radial peripapillary capillaries in glaucoma, glaucoma suspect, and normal eyes. Am J Ophthalmol. 2016;170:41–49. Kupersmith MJ, Garvin MK, Wang JK, Durbin M, Kardon R. Retinal ganglion cell layer thinning within one month of presentation for non-arteritic anterior ischemic optic neuropathy. Invest Ophthalmol Vis Sci. 2016;57:3588–3593. Park SW, Ji YS, Heo H. Early macular ganglion cell-inner plexiform layer analysis in non-arteritic anterior ischemic optic neuropathy. Graefes Arch Clin Exp Ophthalmol. 2016;254:983–989. Larrea BA, Iztueta MG, Indart LM, Alday NM. Early axonal damage detection by ganglion cell complex analysis with optical coherence tomography in nonarteritic anterior ischaemic optic neuropathy. Graefes Arch Clin Exp Ophthalmol. 2014;252:1839–1846. Rebolleda G, Sanchez-Sanchez C, Gonzalez-Lopez JJ, Contreras I, Munoz-Negrete FJ. Papillomacular bundle and inner retinal thicknesses correlate with visual acuity in nonarteritic anterior ischemic optic neuropathy. Invest Ophthalmol Vis Sci. 2015;56:682–692. Wang H, Meng ZY, Li SG, Wang JJ, Sun J, Li HY. Macular evaluation of the retinal and choroidal vasculature changes in anterior ischemic optic neuropathy-a case control study. BMC Ophthalmol. 2018;18:341. Moon et al: J Neuro-Ophthalmol 2021; 41: e440-e450 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.