Patterns of Retinal Ganglion Cell Damage in Nonarteritic Anterior Ischemic Optic Neuropathy Assessed by Swept-Source Optical Coherence Tomography

Objective: To evaluate the ability of macular ganglion cell and inner plexiform layer (mGCIPL) and retinal nerve fiber layer (RNFL) thickness measurements by long-wavelength swept-source optical coherence tomography (SS-OCT) to assess retinal ganglion cell (RGC) damage in nonarteritic anterior ischemic optic neuropathy (NAION).

Original Contribution Section Editors: Clare Fraser, MD Susan Mollan, MD Patterns of Retinal Ganglion Cell Damage in Nonarteritic Anterior Ischemic Optic Neuropathy Assessed by Swept-Source Optical Coherence Tomography Jingjing Jiang, MD, Zhijun Wang, MSc, Yi Chen, MD, Aihong Li, MBBS, Chuan Sun, MD, Xinquan Sun, MBBS Downloaded from http://journals.lww.com/jneuro-ophthalmology by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1AWnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8KKGKV0Ymy+78= on 05/04/2022 Objective: To evaluate the ability of macular ganglion cell and inner plexiform layer (mGCIPL) and retinal nerve fiber layer (RNFL) thickness measurements by long-wavelength swept-source optical coherence tomography (SS-OCT) to assess retinal ganglion cell (RGC) damage in nonarteritic anterior ischemic optic neuropathy (NAION). Methods: A retrospective study of 20 patients with unilateral NAION was performed. SS-OCT scanning of the macular and peripapillary areas was performed to measure the total and six-sector thicknesses of macular RNFL (mRNFL) and mGCIPL, as well as peripapillary RNFL (pRNFL) thicknesses in global and 12 clock-hour sectors. Further comparison of these thicknesses between NAION involved eyes and uninvolved counterparts was performed in 12 of the 20 patients at 4 visits. The thickness map and en face images generated by volume data of the posterior pole over a 12 · 9-mm area were used for RNFL analysis. Results: Median time intervals between the visual symptom onset and first thinning occurrences of mGCIPL, mRNFL, and pRNFL were 17 days (95% Confidence Interval [CI] 14– 18 days), 43 days (95% CI 32–48 days), and 70 days (95% CI 62–80 days), respectively. The thickness map indicated a significantly reduced pRNFL in the superior temporal sectors or temporal sectors after 9 weeks, and retinal damage corresponded to the superior hemisphere’s mRNFL and mGCIPL. En face images showed that the RNFL thinning area gradually expanded along the retinal nerve fiber direction and progressed toward the optic nerve head. Conclusions: The patterns of RGC damage in the macular and peripapillary areas of NAION eyes can be revealed by SS-OCT. Objective measurement of SS-OCT is valuable in characterizing NAION. Journal of Neuro-Ophthalmology 2021;41:37–47 doi: 10.1097/WNO.0000000000001025 © 2020 by North American Neuro-Ophthalmology Society Department of Ophthalmology, China-Japan Friendship Hospital, Beijing City, China. The authors report no conflicts of interest. Address correspondence to Xinquan Sun, Department of Ophthalmology, China-Japan Friendship Hospital, No.2 Yinghua East Road, Chaoyang District, Beijing 100029, China; E-mail: sxq2013@hotmail. com Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 N onarteritic anterior ischemic optic neuropathy (NAION) is the most common acute optic neuropathy in middle-aged and older individuals. NAION is characterized by a sudden, painless onset of visual loss with optic disc edema and usually followed by optic disc pallor. An increase in the pallor of the disc is considered a sign of significant axonal retrograde and retinal ganglion cell (RGC) loss caused by NAION (1–3). RGC elements are found in different layers of the retina as follows: ganglion cell axons in the retinal nerve fiber layer (RNFL), ganglion cell bodies in the ganglion cell layer, and ganglion cell dendrites in the inner plexiform layer. Therefore, measuring the RNFL and the combined ganglion cell and inner plexiform layer (GCIPL) could be helpful for evaluating the patterns of RGC damage. Optical coherence tomography (OCT) is widely used in NAION patients to quantify optic disc edema and for monitoring peripapillary RNFL (pRNFL) loss (4–6). As more than 50% RGCs are distributed in the macular region, macular ganglion cell complex (mGCC) assessment has attracted increasing attention in recent years. Macular GCC scans could provide a good alternative to or may complement pRNFL analysis in the detection of optic nerve pathology (7). In a previous study, we assessed the evolution process of macular ganglion cell damage in neuritis patients by evaluating the thickness changes of mGCC and pRNFL (8). It was found that pRNFL and mGCC thinning in NAION measured by OCT correlates well with the visual field in both magnitude and location (1,2), with a high ability to predict visual outcomes (9,10). Meanwhile, mGCC thickness seems even more useful for detecting RGC loss at an early stage compared with pRNFL thickness because former is unaffected by optic disc swelling (11). Measurements of macular RNFL (mRNFL) and macular GCIPL (mGCIPL), which are separated from the mGCC in OCT analysis, could provide even more useful and reliable information for monitoring RGC’s axonal thinning 37 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution and cell body loss, respectively. Several studies have focused on mGCIPL thinning in the acute stage of NAION (9,12– 17), and reports detecting mRNFL thickness are scarce (9,18). Long-wavelength swept-source OCT (SS-OCT), a new generation of high-penetration optical coherence tomography devices, shows improved resolution and is capable of performing a larger range of and more precise RNFL thickness measurements, combining both the pRNFL and mRNFL in thickness maps. SS-OCT incorporates automated retinal segmentation software that allows a highly reliable and more objective evaluation of different layers, including mGCC, mGCIPL, and mRNFL. The high scanning speed also enables sufficient density scans to produce volume maps of the retina for the construction of en face (c-scans) sections (19,20). Previous studies have shown that mGCC or mGCIPL thinning may be a very early sign of RGC damage in NAION eyes (11,12,21,22). It is unclear whether RGC body loss is only a result of RGC axonal loss or occurs before the atrophy of RGC axons. In addition, whether the area of mGCIPL thinning in the macular region corresponds to the RNFL thinning sector of the optic disc is largely unknown. The purpose of this study was to assess the sequence and location of RGC cell body loss and axonal damage through quantitative assessment of mGCIPL and RNFL thicknesses by SS-OCT. This study not only measured the thickness variation of mGCIPL but also evaluated the trajectory of RNFL thinning and the time course of thickness change in both the pRNFL and mRNFL. We sought to determine if the change patterns of mGCIPL and RNFL thicknesses could be a structural biomarker of RGC damage at different stages and could lead to a better understanding of the pathological process of NAION. METHODS Patients Between September 2016 and August 2018, patients with unilateral NAION at the Department of Ophthalmology, China-Japan Friendship Hospital, were enrolled in this observational retrospective study. The study was approved by the Clinical Research Ethics Committee and conducted in accordance with the Declaration of Helsinki. Diagnosis of NAION was based on sudden painless loss of vision, optic disc swelling associated or not with peripapillary hemorrhage, compatible visual field impairment, and normal C-reactive protein levels. The diagnosis was first made by an ophthalmologist based on medical history, best-corrected visual acuity (BCVA) measured by the Snellen chart, intraocular pressure (IOP) measurement with pneumatic tonometry, slit-lamp biomicroscopy, fundus examination, and automated visual field test and confirmed through stable condition at follow-up and exclusion of other causes of optic neuropathy. 38 Twenty consecutive unilateral NAION patients were included according to the following criteria: (1) BCVA, IOP, and SS-OCT measurements at follow-up visits for both involved and uninvolved eyes; (2) presentation within 2 weeks of the symptom onset; and (3) follow-up of more than 4 months. Among these 20 patients, a total of 12 returned for every follow-up appointment within 4 months according to the study time points. These time points, determined by the standard of clinical care (22) and especially the need for close observation in the early stages of NAION, were 0–2 weeks, 3–4 weeks, 5–8 weeks, 9–16 weeks, and .16 weeks. Survival analysis was performed to show the distribution of times when mGCIPL, mRNFL, and pRNFL thinning events were first detected in the 20 patients. One-way repeated-measures analysis of variance (ANOVA) was performed to compare mGCIPL, mRNFL, and pRNFL thicknesses between NAION involved and uninvolved eyes at the initial visit and other 3 follow-up visits in 12 patients. Optical Coherence Tomography Scanning Swept-source (SS)-OCT (DRI Triton; Topcon, Tokyo, Japan) was performed in a dark room by the same operator for B-scans and en face images. Images were obtained by OCT after pupil dilation, and those with poor quality and signal strength index (SSI) ,50 were excluded. The SSOCT incorporates a light source at 1,050 nm and operates at a speed of 100,000 A-scans per second. The macular cube data covering a 6 · 6-mm2 scan area centered on the fovea were used for mGCC analysis. In the SS-OCT report, the mGCC represented as GCL++ was divided into mRNFL and mGCL+ and mGCIPL, in which combined layers of mGCL and mIPL are represented as mGCL+. Meanwhile, mGCL++ and mGCL+ thicknesses were measured in the following 6 sectors around the macula foveal center: nasal superior, superior, temporal superior, temporal inferior, inferior, and nasal inferior regions (Fig. 1). Then, pRNFL thickness was measured by a peripapillary circle scan in a 3.4-mm diameter section around the optic nerve disc. The evaluated pRNFL thickness parameters were total thickness (360° measurement), two-regions thickness (superior and inferior), and thickness at each of the 12 clock-hour sectors as follows: 12 o’clock (superior), 3 o’clock (nasal), 6 o’clock (inferior), and 9 o’clock (temporal) (Fig. 1). The sector thicknesses in the normal range were represented by green backgrounds, and those in abnormal ranges at the 5% and 1% levels by yellow and red backgrounds, respectively. For RNFL analysis, the thickness map and en face images were generated by volume data of the posterior pole over a 12 · 9-mm area consisting of 256 horizontal B-scans. Areas with yellow or red in the deviation map were defined as thinning and uncolored areas indicated normal values. Thicknesses below the normal values at the 5% and 1% levels were represented by yellow and red areas, respectively. The data were subjected to flattening Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 1. Swept-source optical coherence tomography analysis of NAION eyes. A. Sectoral macular GCL++ (combined mRNFL and mGCIPL) thickness map. The macula was divided into 6 sectors with thickness in micrometers in each sector; colors indicate deviation from normal range data. B. Sectoral macular GCL+ (mGCIPL) thickness map. C. Clock-hour pRNFL thickness map and (D) pRNFL thickness profile (T, temporal, 9 o’clock; S, superior, 12 o’clock; N, nasal, 3 o’clock; and I, inferior, 6 o’clock). mRNFL, macular retinal nerve fiber layer; mGCIPL, macular ganglion cell and inner plexiform layer; NAION, nonarteritic anterior ischemic optic neuropathy; pRNFL, peripapillary retinal nerve fiber layer. with the inbuilt software (IMAGEnet 6; Topcon Corporation, Tokyo, Japan) with the inner limiting membrane as a reference plane. En face images were produced at various pixel depths below the reference plane, with a depth scale of 2.6 mm/pixel. Three different depth (10 pixel, 20 pixel, and 50 pixel) and no-head en face images were used for RNFL evaluation. involved eyes at different time intervals were compared with those of the normal uninvolved eyes by one-way repeatedmeasures ANOVA. P , 0.05 was considered statistically significant. Statistical Analysis The 20 NAION patients in this study included 9 men and 11 women, with an average age of 64.6 ± 10.3 years. The involved eyes included 9 right and 11 left eyes, with a mean BCVA of 0.3 ± 0.2 (0.02–0.6) and a mean IOP of 14.83 ± 2.50 mm Hg (10–19 mm Hg) at the initial visit. The mean time from the symptom onset to initial OCT testing was 8.1 ± 3.5 days (3–13 days). The first incidence of thinning changes of the mGCIPL, mRNFL, and pRNFL in NAION eyes over the entire follow-up period was assessed based on the Kaplan–Meier survival analysis (Fig. 2). The median survival times of mGCIPL, mRNFL, and pRNFL thickness thinning were 17 days (95% Confidence Interval [CI], 14–18 days), 43 days (95% CI, 32–48 days), and 70 days (95% CI, 62–80 Statistical analyses were completed with the SAS software, version 9.4 (SAS Institute Inc). Continuous variables were presented as mean ± SD. The time elapsed from the visual symptom onset to sector thickness first falling in the abnormal range in the mGCIPL, mRNFL, or pRNFL was assessed and presented as the median survival time. The Kaplan–Meier survival method and logrank test were used to display and assess differences of cumulative survival rates. The Mann–Whitney U test was performed to detect significant differences between the involved eyes and contralateral normal uninvolved counterparts in NAION patients. The mGCIPL, mRNFL, and pRNFL thicknesses of the RESULTS FIG. 2. Days from the visual symptom onset to thickness decrease in the mGCIPL, mRNFL, and pRNFL in NAION eyes (Kaplan–Meier plot). mRNFL, macular retinal nerve fiber layer; mGCIPL, macular ganglion cell and inner plexiform layer; NAION, nonarteritic anterior ischemic optic neuropathy; pRNFL, peripapillary retinal nerve fiber layer. Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 39 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution days), respectively. The mGCIPL thickness first decreased mainly at 2–3 weeks, followed by an mRNFL thickness decrease mostly at 5–7 weeks, and finally pRNFL thickness decrease mainly at 9–10 weeks. Longitudinal OCT analysis at each visit in NAION and normal uninvolved eyes is shown in Table 1. The total mGCIPL thicknesses in NAION and uninvolved eyes were significantly reduced at 3–4 weeks, 5–8 weeks, and 9–16 weeks (P , 0.01). Thinning of mGCIPL in the nasal superior, superior, and temporal superior sectors was significantly observed at 3–4 weeks (P , 0.01). Thinning of mGCIPL in all 6 sectors was reduced at 5–8 weeks, with no significant change occurring between 5–8 weeks and 9–16 weeks. The total mRNFL thickness of NAION eyes was increased compared with that of the uninvolved eyes at the initial visit, with a significant elevation at 3–4 weeks (P , 0.01); it began to decrease at 5–8 weeks, with a significant reduction at 9–16 weeks (P , 0.01). The thinning of mRNFL in the nasal superior and superior sectors was significantly observed at 5–8 weeks (P , 0.05) and 9–16 weeks (P , 0.01). The total pRNFL thickness of NAION eyes was significantly increased from the initial visit (P , 0.01) to 5–8 weeks later (P , 0.05) and then began to decrease at 9–16 weeks. The superior pRNFL thickness was significantly reduced compared with that of the uninvolved eyes at 9–16 weeks (P , 0.01). Thinning of clockhour pRNFL thickness was significantly observed from 9 to 11 o’clock (3 hours). Figures 3 and 4 show the characteristic patterns of mGCIPL damage and progressive RNFL damage in the macular and peripapillary areas. The SS-OCT color-coded map indicated no evident mGCIPL and RNFL damage at the initial visit (Figs. 3, 4 A-1–D-1). Thinning of the mGCIPL was first observed at 3–4 weeks, and the thinning area tended to be stable after 8 weeks (Figs. 3, 4A). RNFL thinning was first found in the macular area and then in the peripapillary region; the thinning area along the retinal nerve fiber direction gradually expanded over the course of time (Figs. 3, 4B–D). The color fundus and mRNFLpRNFL thickness map showed overt peripapillary nerve fiber thickening in 4 weeks after the disease onset, and the thickness gradually decreased during follow-up. The maps also indicated significantly reduced pRNFL in the superior temporal or temporal sector at 9–16 weeks, and retinal damage corresponded to superior hemisphere’s mRNFL and mGCIPL damage (Figs. 3, 4C–D). The patterns of mRNFL and pRNFL fiber damage in NAION eyes at various pixel depths were detected by en face SS-OCT images (Figs. 5 and 6). RNFL fiber thickening was predominant in the superficial (10-pixel depth), middle (20-pixel depth), and deep (50-pixel depth) layers within 4 weeks of the onset (Figs. 5, 6A-1–D-1, A-2–D-2). En face images of the superficial, middle, and deep layers first showed mild defects of the retinal nerve fiber layer at 5–8 weeks; the superior sector defects became increasingly 40 serious while the damage areas were gradually expanded and progressed toward the optic nerve head during follow-up (Figs. 5, 6A-3–C-3, A-4–C-4, A-5–C-5). Obvious pRNFL bundle defects in the superotemporal and temporal peripapillary sectors were found in no-head en face images after 9 weeks (Figs. 5, 6D-4–D-5). DISCUSSION This study confirmed the value of mGCIPL thickness measurement for detecting RGC damage at the early stage of NAION as previously described (12–14,16). Although most of these studies reported that mGCIPL thinning is first evident at 1–2 months and persisted for 3 months, we have not observed exactly the same. Instead, we found that mGCIPL thickness first decreased at 3–4 weeks in 3 sectors and then in all sectors at 5–8 weeks, whereas thinning was minimally detected after 9 weeks. A reason for these discrepant results might be that we detected sectoral mGCIPL thinning in any of the 6 sectors of the macular area, which is different from observing the average mGCIPL thickness as performed in previous studies (12,14,16). In addition, this study increased the frequency of follow-up at the early stage, refining the first month after NAION onset into 2 weekly follow-up periods. A recent study of GCIPL by spectral-domain OCT (SD-OCT) found that average and minimum GCIPL are abnormally thinned in 43.8% and 75% of NAION eyes at 2 weeks, respectively, with GCIPL’s minimum thickness significantly correlated with functional parameters (12). These conclusions supported our results and suggested that GCIPL averaging or longer than follow-up intervals at the acute stage reduces the sensitivity of neuronal loss detection. Therefore, these results were more reasonable and indicative of early RGC damage in NAION. To the best of our knowledge, only 2 previous reports have investigated the change of mRNFL in NAION patients. One of them found mRNFL thickness is significantly reduced in eyes with NAION for at least 6 months compared with the contralateral uninvolved eyes (9). The other reported a significantly increased mean thickness of mRNFL at the 1-month time point (18). By contrast, this study observed that the total mRNFL thicknesses were increased at 4 weeks, then began to decrease at 5–8 weeks, and significantly decreased after 9 weeks. The total pRNFL thicknesses were increased significantly from the initial visit to 8 weeks and began to decrease at 9–16 weeks. These results suggested that mRNFL thinning occurred earlier than pRNFL thickness reduction. In addition, en face SSOCT images make it possible to separately investigate RNFL values of different depths and sectors within the total measured macular and peripapillary areas. The RNFL values in superficial, middle, and deep layers first showed mild defects in the paramacular area at 5–8 weeks and bundle defects of the peripapillary area after 9 weeks. The results from en face images further confirmed that RNFL damage Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 Thickness Initial Visit (1.1%) (2.3%) (21.2%) (26.6%) (1.5%) (5.5%) (5.7%) 63.2 ± 5.6** 61.8 ± 8.0** 59.6 ± 10.1** 55.0 ± 9.6** 69.3 ± 8.7 65.8 ± 6.9 67.6 ± 9.1 (13.4%) (34.0%) (13.1%) (15.2%) (1.2%) (1.8%) (9.3%) 44.2 ± 7.8** 56.6 ± 14.3* 45.3 ± 8.7* 34.8 ± 6.0** 29.9 ± 4.4* 42.7 ± 7.1 55.8 ± 16.6 (146.1%) 234.8 ± 55.1** (136.9%) 282.9 ± 59.8** (117.5%) 285.1 ± 87.5** (135.5%) 261.4 ± 56.7** (173.1%) 215.7 ± 65.4** (197.5%) 168.3 ± 67.7** (205.0%) 175.0 ± 79.4** (136.9%) 246.3 ± 75.9** (121.3%) 316.8 ± 99.2** (118.3%) 293.6 ± 109.8** (183.0%) 190.9 ± 92.3** (150.7%) 143.2 ± 58.7** (156.2%) 221.0 ± 62.2** (116.8%) 304.6 ± 71.1** (133.7%) 281.8 ± 75.5** 5–8 Weeks 9–16 Weeks Uninvolved Eye P Value (214.4%) 56.9 ± 5.2** (219.5%) 53.7 ± 8.4** (218.1%) 52.8 ± 8.0** (226.4%) 48.9 ± 8.1** (27.4%) 62.6 ± 12.9** (24.8%) 62.0 ± 8.6* (29.0%) 61.3 ± 9.3** (222.9%) 54.1 ± 5.9** (230.1%) 51.3 ± 9.0** (227.5%) 50.4 ± 7.8** (234.5%) 46.7 ± 8.0** (216.3%) 60.0 ± 13.4** (210.3%) 58.5 ± 8.8** (217.5%) 57.8 ± 10.9** (226.7%) (233.2%) (230.8%) (237.5%) (219.8%) (215.3%) (222.2%) 73.8 76.8 72.8 74.7 74.8 69.1 74.3 ± ± ± ± ± ± ± 6.3 7.2 6.6 6.8 6.5 7.0 7.0 0.8119/0.0003/,0.0001/,0.0001 0.6740/,0.0001/,0.0001/,0.0001 0.8103/0.0010/,0.0001/,0.0001 0.2420/,0.0001/,0.0001/,0.0001 0.7586/0.0941/0.0077/0.0024 0.2801/0.2506/0.0374/0.0036 0.3163/0.0543/0.0008/0.0002 (23.5%) (29.2%) (18.6%) (42.6%) (17.7%) (12.4%) (24.0%) 32.4 ± 5.1 37.2 ± 11.7 30.6 ± 5.5** 25.3 ± 4.1 26.7 ± 4.1 33.8 ± 7.3 40.9 ± 11.5 (29.5%) (215.1%) (219.9%) (3.7%) (5.1%) (211.1%) (29.1%) 29.2 ± 4.8** 30.8 ± 7.3** 24.9 ± 3.3** 22.1 ± 3.6 25.4 ± 3.9 32.9 ± 9.6 39.1 ± 12.2 (218.4%) (229.7%) (234.8%) (29.4%) (0.0%) (213.4%) (213.1%) 35.8 43.8 38.2 24.4 25.4 38.0 45.0 ± ± ± ± ± ± ± 5.2 8.3 6.1 3.3 4.4 5.3 9.6 0.0674/0.0053/0.1219/0.0038 0.0162/0.0133/0.1255/0.0005 0.2233/0.0288/0.0040/,0.0001 0.0305/,0.0001/0.5892/0.1105 0.9029/0.0196/0.4756/1.0000 0.7632/0.0815/0.1240/0.1234 0.3286/0.0628/0.3565/0.2013 (107.1%) (100.9%) (91.0%) (105.8%) (130.2%) (138.0%) (129.4%) (99.8%) (98.1%) (98.0%) (130.8%) (80.6%) (104.1%) (97.3%) (110.3%) 137.1 ± 29.0* 158.4 ± 45.8 187.4 ± 69.4 142.3 ± 43.1 117.7 ± 36.0* 94.7 ± 23.9** 96.3 ± 17.9** 151.9 ± 45.0 207.8 ± 82.7 190.2 ± 89.9 115.1 ± 45.3* 80.0 ± 20.5 102.0 ± 24.4 165.3 ± 60.7 161.7 ± 56.7 (20.9%) 107.4 ± 22.1 (12.5%) 112.2 ± 23.7** (25.5%) 150.0 ± 58.9 (12.0%) 117.7 ± 25.8 (25.6%) 100.8 ± 29.1 (33.9%) 81.8 ± 22.0 (26.2%) 82.6 ± 18.2 (23.2%) 130.2 ± 40.1 (30.0%) 161.3 ± 70.8 (28.3%) 160.7 ± 79.7 (39.2%) 98.1 ± 52.9 (0.9%) 60.9 ± 20.6* (25.8%) 69.7 ± 18.8** (7.1%) 98.2 ± 28.7** (20.7%) 114.2 ± 28.1 (25.3%) (220.3%) (0.5%) (27.3%) (7.6%) (15.7%) (8.3%) (5.6%) (0.9%) (8.4%) (18.6%) (223.2%) (235.6%) (236.4%) (214.8%) 113.4 140.8 149.3 127.0 93.7 70.7 76.3 123.3 159.9 148.3 82.7 79.3 108.3 154.4 134.0 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± 10.8 19.3 25.2 22.3 16.4 10.8 15.0 23.5 27.6 26.7 13.0 10.6 14.1 24.9 18.6 ,0.0001/,0.0001/0.0146/0.4069 ,0.0001/,0.0001/0.2336/0.0037 ,0.0001/,0.0001/0.0870/0.9680 ,0.0001/,0.0001/0.2853/0.3535 ,0.0001/,0.0001/0.0474/0.4706 ,0.0001/,0.0001/0.0044/0.1313 ,0.0001/0.0003/0.0074/0.3692 ,0.0001/,0.0001/0.0638/0.6156 ,0.0001/,0.0001/0.0706/0.9491 ,0.0001/0.0002/0.1359/0.6139 ,0.0001/0.0006/0.0262/0.3376 ,0.0001/0.0012/0.9211/0.0115 ,0.0001/,0.0001/0.4451/,0.0001 ,0.0001/,0.0001/0.5731/,0.0001 ,0.0001/,0.0001/0.1227/0.0535 Original Contribution mGCIPL thickness (mm) Total 74.6 ± 10.2 Supernasal 78.6 ± 12.2 Superior 71.9 ± 11.3 Superotemporal 69.8 ± 12.4 Inferotemporal 75.9 ± 10.1 Inferior 72.9 ± 9.7 Inferonasal 78.5 ± 12.2 mRNFL thickness (mm) Total 40.6 ± 6.9 Supernasal 58.7 ± 18.0* Superior 43.2 ± 12.4 Superotemporal 28.1 ± 4.4* Inferotemporal 25.7 ± 5.5 Inferior 38.7 ± 5.4 Inferonasal 49.2 ± 10.8 pRNFL thickness (mm) Total 279.1 ± 70.4** Superior 333.6 ± 79.4** Inferior 324.8 ± 95.8** 1 o’clock 299.1 ± 76.6** 2 o’clock 255.9 ± 63.4** 3 o’clock 210.3 ± 57.0** 4 o’clock 232.7 ± 80.0** 5 o’clock 292.1 ± 78.6** 6 o’clock 353.9 ± 112.3** 7 o’clock 323.8 ± 123.6** 8 o’clock 234.0 ± 94.5** 9 o’clock 198.8 ± 72.1** 10 o’clock 277.5 ± 93.9** 11 o’clock 334.8 ± 125.3** 12 o’clock 313.2 ± 91.9** 3–4 Weeks Values are mean ± SD (mm). *P , 0.05, **P , 0.01; one-way repeated-measures analysis of variance (ANOVA) was used to compare thickness between NAION involved eyes and uninvolved counterparts at different time intervals. mGCIPL, mRNFL, and pRNFL thicknesses (%): percentage increase (positive value) or decrease (negative value) in NAION eyes compared with normal uninvolved eyes. mGCIPL, macular ganglion cell and inner plexiform layer; mRNFL, macular retinal nerve fiber layer; pRNFL, peripapillary retinal nerve fiber layer; SS-OCT, swept-source optical coherence tomography; NAION, nonarteritic anterior ischemic optic neuropathy. 41 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. TABLE 1. Longitudinal comparison of mGCIPL, mRNFL, and pRNFL thicknesses by SS-OCT in patients with NAION Original Contribution FIG. 3. Optical coherence tomographic images showing mGCIPL (A), pRNFL (B), and mRNFL-pRNFL thickness map (C, D) in a patient with nonarteritic anterior ischemic optic neuropathy at 6 follow-up periods. Yellow and red in the deviation map indicate thicknesses below the fifth and first percentiles of normal distribution, respectively. A. Thinning of the mGCIPL was detected at 3–4 weeks, and the thinning area tended to be stable after 8 weeks. B. Thinning of the pRNFL was significantly detected after 8 weeks, in the superior temporal sector. C, D. RNFL thinning was detected in the macular area at 5–8 weeks and gradually expanded to the peripapillary area along the retinal nerve fiber direction. mRNFL, macular retinal nerve fiber layer; mGCIPL, macular ganglion cell and inner plexiform layer; pRNFL, peripapillary retinal nerve fiber layer. 42 Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 4. Optical coherence tomographic images showing mGCIPL (A), pRNFL (B), and mRNFL-pRNFL thickness map (C, D) in a patient with nonarteritic anterior ischemic optic neuropathy at 5 follow-up periods. A. Thinning of the mGCIPL was detected at 3–4 weeks, and the thinning area tended to be stable after 8 weeks. B. Thinning of the pRNFL was detected at 5–8 weeks and in the superior temporal and temporal sectors. C, D. RNFL thinning was detected in the macular area at 3–4 weeks and gradually expanded to the peripapillary area along the retinal nerve fiber direction. mRNFL, macular retinal nerve fiber layer; mGCIPL, macular ganglion cell and inner plexiform layer; pRNFL, peripapillary retinal nerve fiber layer. occurred earlier in the macular area than in the peripapillary area. This study showed that the time and location of thinning were not directly related to the severity of edema. For example, the supernasal mGCIPL thickness (78.6 ± 12.2 mm) was greatest at the initial visit but thinner than average mGCIPL thickness (61.8 ± 8.0 to 63.2 ± 5.6 mm) at the follow-up within 4 weeks. The superior mRNFL thickness was greater than average mRNFL thickness Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 (43.2 ± 12.4 to 40.6 ± 6.9 mm) at the initial visit but smaller than the average thickness (30.6 ± 5.5 to 32.4 ± 5.1 mm) at the second follow-up. Similarly, the 11 o’clock pRNFL thickness was thicker than the average thickness (334.8 ± 125.3 to 279.1 ± 70.4 mm) at the initial visit. However, it was thinner than the average thickness (98.2 ± 28.7 to 107.4 ± 22.1 mm) at the fourth follow-up. In addition, en face images showed that the RNFLs of the superior and inferior half parts were generally symmetrical 43 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 5. En face SS-OCT images of RNFL fibers in the superficial layer (A), middle layer (B), deep layer (C), and radial peripapillary area (D) at 5 follow-up periods in a NAION eye. RNFL thinning at the superficial, middle, and deep layers is shown at 5–8 weeks (A-3–C-3), and the defect area progressed toward the optic nerve head in follow-up visits (A-4–C-4 and A-5–C-5). The border between the damaged and undamaged areas is indicated by white arrows. Significant RNFL bundle defects (dark areas) are shown in nerve head superior temporal area by black arrows (D-4–D-5). RNFL, retinal nerve fiber layer; NAION, nonarteritic anterior ischemic optic neuropathy; SS-OCT, swept-source optical coherence tomography. (Fig. 5A-1–C-1, A-2–C-2), and RNFL swelling at the superior half was even more severe than at the inferior one within 4 weeks (Fig. 6A-1–D-1, A-2–D-2). However, RNFL thinning at the superior part was observed earlier and more obvious than that of the inferior part at followups beyond 4 weeks (Figs. 5, 6A-3–D-3, A-4–D-4, A-5–D5). Thus, the RNFL thickening that was initially serious or near the optic nerve head should not necessarily take longer to resolve. RNFL thinning may be partly explained by the different degrees and the sequence of RGC atrophy in the nerve body or axon. Based on findings regarding the characteristics of RNFL thinning, we speculated that the RGC damage pattern follows a certain time sequence. Peripapillary and paramacular RGC axonal swelling occurred at the early stage of RGC damage, then RGC cell body and inner plexiform layer loss, followed by RGC axonal atrophy in the macular area, and finally RGC axonal atrophy in the peripapillary area. The altitudinal asymmetry of macular thinning has been mentioned in a few previous studies (15,17,22). Our results 44 further show that the incidence rates of mGCIPL thickness reduction are likely to be high in the superonasal, superior, and superotemporal sectors in NAION eyes. We speculate that the RNFL thinning area is more specific in assessing the locational correspondence between mGCIPL damage and RNFL damage compared with the RNFL swelling area. Therefore, by focusing on RNFL thickness reduction, we found an early thinning of mRNFL in superonasal and superior sectors, as well as significant pRNFL thinning in the 9 to 11 o’clock (temporal to superotemporal) region, corroborating en face images demonstrating that overt RNFL bundle defects occur more frequently. These results may be explained in 2 ways. First, ischemic events occur more frequently in the superior vs. inferior region in NAION (23,24). Contreras et al found that the superior quadrant shows a higher percentage of RNFL loss compared with RNFL thickness in the superior, nasal, inferior, and temporal quadrants (25). Altitudinal asymmetry in NAION is associated with distinct upper and lower halves of the arterial circle of Zinn–Haller derived from short posterior Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution FIG. 6. En face SS-OCT images of RNFL fibers in the superficial layer (A), middle layer (B), deep layer (C), and radial peripapillary area (D) at 5 follow-up periods in another NAION eye. RNFL thinning at the superficial, middle, and deep layers is shown at 5–8 weeks (A-3–C-3), and the defect area progressed toward the optic nerve head in follow-up visits (A-4–C-4 and A5–C-5). The borders of defect areas are indicated by white arrows. Wide RNFL bundle defects (dark areas) are indicated by black arrows on no-head en face images (D). NAION, nonarteritic anterior ischemic optic neuropathy; RNFL, retinal nerve fiber layer; SS-OCT, swept-source optical coherence tomography. ciliary arteries (17). Second, the RGC axons are distributed in a specific topographic manner at the optic nerve head. Retinal nerve fibers from the macular area travel horizontally as the papillomacular bundle (PMB), and fibers from the temporal retina arch above and below the macula as superior and inferior arcuate fibers (26). Axons from PMB are arranged temporally in the nerve head. The above results indicated the regional, corresponding relationship of thinning mRNFL sectors and thinning pRNFL clock-hours, which conforms to the characteristic distribution of RGC axons. The patterns of RGC damage in the macular and peripapillary areas vary distinctively in different diseases (27). OCT macular and peripapillary analysis data correlate well with changes in histological measurements of RGC and axonal losses occurring in NAION (3,28). Thus, based on SS-OCT analysis of the time course and trend of thickness changes in mGCIPL and RNFL, we could speculate on the mechanism of formation of the patterns of RGC axon and cell body damage in NAION. Ischemic injury at the optic nerve head level causes early axonal swelling, followed by Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 loss of RGC cell bodies and dendrites, with progressive degeneration of axons from the cell body to the distal part. It is increasingly recognized that injury to RGC axons triggers rapid changes in dendrites (29,30). Dendritic abnormalities occur soon after injury and before overt loss of RGC axons, thus identifying dendrite pathology as an early sign of neurodegeneration (31). In addition, macular vessel density loss precedes detectable structural damage in NAION (32). Early deprivation of blood and nutritional support from cell bodies accelerates RGC cell body and dendrite apoptosis. Meanwhile, it takes time for astrocytes and vascular sheaths surrounding the unmyelinated RGC axons to enter the programmed destruction. de Lima et al indicated that calpain inhibition prevents the degeneration of optic nerve fibers by rescuing axons from the process of axonal degeneration, with no effect on RGC survival after optic nerve damage (33). To the best of our knowledge, this is the first study in which longitudinal and retinal segmented analyses of mGCIPL and RNFL were performed to evaluate RGC damage patterns in NAION eyes. Although this study was 45 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution limited by a small sample size, given the statistically significant findings, the overall results could be generalized. The superior regions of the mGCIPL and RNFL were mainly affected in most NAION cases in this study. Larger studies are required to substantiate the above findings. Another limitation is the retrospective nature of case selection. Although OCT images were not acquired at the same time for all cases, analysis time intervals were set to provide a more representative measurement of the time course of RGC damage. In addition, retinal layer segmentation performed in the current version of SS-OCT has been applied in many studies of macular disease rather than NAION. Therefore, more applications of SS-OCT in NAION are needed to further support our attempt in RGC analysis. In conclusion, this longitudinal study showed the dynamic changes of SS-OCT-derived variables in NAION. The timing and location of mGCIPL and RNFL changes indicate the specific patterns of RGC damage. We speculate that early loss of RGC cell bodies and their dendrites precedes axon atrophy, and ascending degeneration of axons starts at the ganglion cell bodies and progresses toward the distal optic nerve. Standard, objective structural measurement and analysis obtained from SS-OCT scanning contribute to the assessment of NAION disease course and characterization, which could help evaluate prognosis and establish a therapeutic window. Additional histological studies are required to further assess the patterns of RGC damage in NAION. REFERENCES 1. Aggarwal D, Tan O, Huang D, Sadun AA. Patterns of ganglion cell complex and nerve fiber layer loss in nonarteritic ischemic optic neuropathy by fourier-domain optical coherence tomography. Invest Opthalmology Vis Sci. 2012;53:4539. 2. 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. 3. Shariati MA, Park JH, Liao YJ: Optical coherence tomography study of retinal changes in normal aging and after ischemia. Invest Opthalmology Vis Sci. 2015;56:2790. 4. Kernstock C, Beisse F, Wiethoff S, Mast A, Krapp E, Grund R, Dietzsch J, Lagreze W, Fischer D, Schiefer U. Assessment of functional and morphometric endpoints in patients with nonarteritic anterior ischemic optic neuropathy (NAION). Graefes Arch Clin Exp Ophthalmol. 2014;252:515–521. 5. Rebolleda G, Diez-Alvarez L, Casado A, Sanchez-Sanchez C, de Dompablo E, Gonzalez-Lopez JJ, Munoz-Negrete FJ. OCT: new perspectives in neuro-ophthalmology. Saudi J Ophthalmol. 2015;29:9–25. 6. Ackermann P, Brachert M, Albrecht P, Ringelstein M, Finis D, Geerling G, Aktas O, Guthoff R. Alterations of the outer retina in non-arteritic anterior ischaemic optic neuropathy detected using spectral-domain optical coherence tomography. Clin Exp Ophthalmol. 2017;45:496–508. 7. Gonul S, Koktekir BE, Bakbak B, Gedik S. Comparison of the ganglion cell complex and retinal nerve fibre layer measurements using Fourier domain optical coherence tomography to detect ganglion cell loss in non-arteritic anterior ischaemic optic neuropathy. Br J Ophthalmol. 2013;97:1045– 1050. 46 8. Yan YJ, Zhang L, Wang ZJ, Li AH, Sun XQ. Clinical observation on the evolution process of macular ganglion cell complex and peripapillary retinal nerve fiber layer of neuritis patients. Zhonghua Yan Ke Za Zhi. 2018;54:62–68. 9. 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. 10. Sun MH, Liao YJ. Structure-function analysis of nonarteritic anterior ischemic optic neuropathy and age-related differences in outcome. J Neuroophthalmol. 2017;37:258–264. 11. Goto K, Miki A, Araki S, Mizukawa K, Nakagawa M, Takizawa G, Ieki Y, Kiryu J. Time course of macular and peripapillary inner retinal thickness in non-arteritic anterior ischaemic optic neuropathy using spectral-domain optical coherence tomography. Neuroophthalmology. 2016;40:74–85. 12. De Dompablo E, Garcia-Montesinos J, Munoz-Negrete FJ, Rebolleda G. Ganglion cell analysis at acute episode of nonarteritic anterior ischemic optic neuropathy to predict irreversible damage. A prospective study. Graefes Arch Clin Exp Ophthalmol. 2016;254:1793–1800. 13. 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. 14. Akbari M, Abdi P, Fard MA, Afzali M, Ameri A, Yazdani-Abyaneh A, Mohammadi M, Moghimi S. Retinal ganglion cell loss precedes retinal nerve fiber thinning in nonarteritic anterior ischemic optic neuropathy. J Neuroophthalmol. 2016;36:141– 146. 15. Park SW, Heo MY, Ji YS, Heo H. Lack of correspondence between early optic disc oedema and macular ganglion cell loss in non-arteritic anterior ischaemic optic neuropathy. Neuroophthalmology. 2014;38:36–38. 16. 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. 17. Lee YH, Kim KN, Heo DW, Kang TS, Lee SB, Kim CS. Difference in patterns of retinal ganglion cell damage between primary open-angle glaucoma and non-arteritic anterior ischaemic optic neuropathy. PLoS One. 2017;12:e0187093. 18. Keller J, Oakley JD, Russakoff DB, Andorra-Ingles M, Villoslada P, Sanchez-Dalmau BF. Changes in macular layers in the early course of non-arteritic ischaemic optic neuropathy. Graefes Arch Clin Exp Ophthalmol. 2016;254:561–567. 19. Ng DS, Bakthavatsalam M, Lai FH, Cheung CY, Cheung GC, Tang FY, Tsang CW, Lai TY, Wong TY, Brelen ME. Classification of exudative age-related macular degeneration with pachyvessels on en face swept-source optical coherence tomography. Invest Ophthalmol Vis Sci. 2017;58:1054–1062. 20. Hirasawa K, Shoji N. Association between ganglion cell complex and axial length. Jpn J Ophthalmol. 2013;57:429– 434. 21. Lee TH, Heo H, Park SW. Clinical usefulness of spectraldomain optical coherence tomography in glaucoma and NAION. Chonnam Med J. 2016;52:194–200. 22. Erlich-Malona N, Mendoza-Santiesteban CE, Hedges TR III, Patel N, Monaco C, Cole E. Distinguishing ischaemic optic neuropathy from optic neuritis by ganglion cell analysis. Acta Ophthalmol. 2016;94:e721–e726. 23. Gerling J, Meyer JH, Kommerell G. Visual field defects in optic neuritis and anterior ischemic optic neuropathy: distinctive features. Graefes Arch Clin Exp Ophthalmol. 1998;236:188– 192. 24. Saito H, Tomidokoro A, Sugimoto E, Aihara M, Tomita G, Fujie K, Wakakura M, Araie M. Optic disc topography and peripapillary retinal nerve fiber layer thickness in nonarteritic ischemic optic neuropathy and open-angle glaucoma. Ophthalmology. 2006;113:1340–1344. 25. Contreras I, Noval S, Rebolleda G, Munoz-Negrete FJ. Followup of nonarteritic anterior ischemic optic neuropathy with Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited. Original Contribution 26. 27. 28. 29. optical coherence tomography. Ophthalmology. 2007;114:2338–2344. Hogan MJAJ, Weddell JE.: Histology of the Human Eye: An Atlas and Textbook. Philadelphia, PA: Saunders, 1971. Lee EJ, Yang HK, Kim TW, Hwang JM, Kim YH, Kim CY. Comparison of the pattern of retinal ganglion cell damage between patients with compressive and glaucomatous optic neuropathies. Invest Ophthalmol Vis Sci. 2015;56:7012– 7020. Ho JK, Stanford MP, Shariati MA, Dalal R, Liao YJ. Optical coherence tomography study of experimental anterior ischemic optic neuropathy and histologic confirmation. Invest Ophthalmol Vis Sci. 2013;54:5981–5988. Della Santina L, Inman DM, Lupien CB, Horner PJ, Wong RO. Differential progression of structural and functional alterations Jiang et al: J Neuro-Ophthalmol 2021; 41: 37-47 30. 31. 32. 33. in distinct retinal ganglion cell types in a mouse model of glaucoma. J Neurosci. 2013;33:17444–17457. Morgan JE. Retina ganglion cell degeneration in glaucoma: an opportunity missed? A review. Clin Exp Ophthalmol. 2012;40:364–368. Weber AJ, Harman CD. Structure-function relations of parasol cells in the normal and glaucomatous primate retina. Invest Ophthalmol Vis Sci. 2005;46:3197–3207. Fard MA, Ghahvechian H, Sahrayan A, Subramanian PS. Early macular vessel density loss in acute ischemic optic neuropathy compared to papilledema: implications for pathogenesis. Transl Vis Sci Technol. 2018;7:10. de Lima S, Mietto BS, Paula C, Muniz T, Martinez AM, Gardino PF. Rescuing axons from degeneration does not affect retinal ganglion cell death. Braz J Med Biol Res. 2016;49:e5106. 47 Copyright © North American Neuro-Ophthalmology Society. Unauthorized reproduction of this article is prohibited.