Macula structural and vascular differences in glaucoma eyes with and without high axial myopia

48 Aims 49 To assess the thickness of various retinal layers, and the superficial vessel density (sVD) in 50 the macula of glaucomatous eyes and their associations with axial length (AL) and visual 51 field mean deviation (VFMD) to identify parameters useful for glaucoma management in 52 myopic eyes. 53

54 248 glaucoma patients (401 eyes) participating in the Diagnostic Innovations in Glaucoma 55 Study observational cohort representing 3 axial myopia groups (non-myopia: n=146 eyes; 56 mild myopia: n=208 eyes; high myopia (AL>26 mm): n=47 eyes) who completed macular 57 OCT and OCT-Angiography imaging were included. The cross-sectional associations of AL 58 and VFMD with the thickness of the ganglion cell inner plexiform layer (GCIPL), macular 59 retinal nerve fiber layer (mRNFL), ganglion cell complex (GCC), sVD and macular choroidal 60 thickness (mCT) were evaluated. 61

62 Thinner Global GCIPL and GCC were significantly associated with worse VFMD (R2=35.1%; 63 and R2=33.4%; respectively p0.350). Thicker mRNFL showed 64 a weak association with increasing AL (R2=3.4%; p=0.001) and a positive association with 65 VFMD (global R2=20.5%; p<0.001). Lower sVD was weakly associated with increasing AL 66 (R2=2.3%; p=0.016) and more strongly associated with more severe glaucoma VFMD 67 (R2=31.8%; p<0.001). Thinner mCT was associated with increasing AL (R2=17.3% p<0.001) 68 and not associated with VFMD (P=0.262). mRNFL was thickest while mCT was thinnest in all 69 sectors of high myopic eyes. 70

71 GCIPL and GCC thinned with increasing severity of glaucoma but were not significantly 72 associated with axial length. GCIPL and GCC thickness may be useful clinical parameters to 73 identify glaucoma in myopic eyes. 74 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 4 R ez a po ur J

75 76 With its potential vision threatening risk and with its prevalence increasing globally, myopia, 77 especially high myopia, has become a major concern around the world.1 78 Although optical coherence tomography (OCT) based measurements of peripapillary retinal 79 nerve fiber layer (pRNFL) thickness can accurately discriminate between healthy and 80 glaucomatous eyes,2 there is concern that in myopic eyes (and especially in high myopic 81 eyes) the diagnostic accuracy of OCT measures is decreased. Optic disc changes in myopic 82 eyes such as morphologic changes in the parapapillary region and optic disc enlargement 83 pose significant challenges to the use of optical imaging and clinical optic disc evaluation to 84 detect and monitor glaucoma (Figure 1).1,3 This is due in part to difference in the regional 85 arrangement of the peripapillary retinal nerve fibers between myopic eyes and emmetropic 86 eyes that may result in sectoral values incorrectly classified as outside normal limits by 87 instrument-specific software analysis in healthy myopic eyes.4 5 88 89 Approximately 50% of the retinal ganglion cells are concentrated within 10 degree of the 90 fovea6 making the macula an useful region for diagnosing optic neuropathies including 91 glaucoma, especially in myopic eyes because myopic axial elongation primarily affects the 92 optic nerve head region. Previous studies have reported that early glaucomatous damage 93 can be detected in the macula region7 and that measurements of the ganglion cell inner 94 plexiform layer (GCIPL) can be used for detecting glaucoma in highly myopic eyes.8-12 95 However, little information is available about differences in the topographic distribution of the 96 thickness of the various macular retinal layers and the retinal vessel density in glaucomatous 97 eyes with and without myopia. Sectoral measurements of the underlying macular vasculature 98 may offer additional insight into differences in glaucomatous eyes with and without myopia. 99 100 Several studies using OCT-Angiography (OCTA) have demonstrated a strong relationship 101 between macular capillary density and the severity of glaucoma.13 14 Furthermore recent 102 studies have reported the peripapillary choroid to be thinner in highly myopic eyes compared 103 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 5 R ez a po ur J to non-myopic eyes.15 16 However, few studies have assessed macular choroidal thickness in 104 highly myopic eyes17 18 and to date, to the best of our knowledge, no study has documented 105 the local distribution of macular choroidal thickness in glaucomatous eyes with and without 106 high myopia. 107 108 The purpose of this study was to characterize the local distribution of GCIPL, GCC, macular 109 retinal nerve fiber layer (mRNFL), choroidal thickness and vessel density in glaucoma eyes 110 with and without axial myopia. By better understanding how the topographic distribution of 111 these parameters varies with axial length and severity of disease, macula parameters that 112 may be useful for detecting and monitoring glaucoma in myopic eyes can be identified. 113 114

115 Study Sample 116 This cross-sectional study included all glaucoma patients enrolled in the University of 117 California, San Diego Diagnostic Innovations in Glaucoma Study (DIGS; clinicaltrials.gov 118 identifier NCT00221897) with available axial length measurements and good quality macula 119 OCT scans acquired between 2015 and 2020. The study was approved by the institutional 120 review board of the University of California San Diego and according to the tenets of the 121 Declaration of Helsinki written informed consent was obtained from all patients. As described 122 previously;19 participants underwent a complete ophthalmologic examination including 123 assessment of refractive error, axial length measurement (IOLMaster, Carl Zeiss Meditec, 124 Dublin, CA), visual field testing, simultaneous stereophotography of the optic disc and 125 macula, and macular OCT and OCTA imaging. Study participants were ≥ 18 years with best-126 corrected visual acuity ≥ 20/40 and open anterior chamber angles at baseline. 127 128 Visual field (VF) testing was performed using the standard Humphrey Field Analyzer 24-2 129 Swedish interactive thresholding algorithm. Repeatable glaucomatous VF damage was 130 defined as the presence of glaucomatous optic nerve head (ONH) damage based on masked 131 assessment by two trained observers and glaucomatous VF damage.19 ONH 132 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 6 R ez a po ur J stereophotographs of highly myopic eyes were graded for glaucoma by two experts (CB and 133 JR) after training with a senior consultant (JBJ). Diagnosis was defined by consensus 134 between the two graders and adjudication by the senior consultant in case of disagreement. 135 136 Myopia definition 137 Because a change in refractive error can occur after refractive or cataract surgery, myopia 138 was classified by axial length into the following 3 groups. 139 -No myopia: axial length ≤ 24.0 mm 140 -Mild myopia: 24.0mm 26.0 mm 142 143 Optical coherence tomography and optical coherence tomography angiography 144 imaging 145 OCT imaging of the macula was performed with the Spectralis OCT (version 6.10; 146 Heidelberg Engineering Inc, Heidelberg, Germany). Details of this instrument have been 147 previously described.14 Macula horizontal posterior pole (p-Pole) scans covering an area of 148 30° x 25° (6 x 6 mm) were obtained. GCIPL, mRNFL and GCC (GCIPL + mRNFL) thickness 149 measurements were generated from each retinal layer from the central 1-, 3-, and 6-mm 150 circles as inner rings (1- and 3-mm circle) and outer rings (3- and 6-mm circle) according to 151 the Early Treatment Diabetic Retinopathy Study defined sectors (temporal, superior, nasal, 152 and inferior). 153 154 OCTA imaging of the macula was performed with the Avanti AngioVue OCT system (version 155 2017.1.0.151; Optovue, Inc., Fremont CA, USA).20 Macular whole image vessel density was 156 calculated on a 3 x 3 mm² field macula scan (304 B-scans x 304 A-scans per B-scan) 157 centered on the fovea. Whole image vessel density of the temporal, superior, nasal, and 158 inferior sectors were reported. Macular parafoveal superficial VD (sVD) was calculated within 159 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 7 R ez a po ur J an annulus centered on the fovea, with an inner diameter of 1 mm and an outer diameter of 160 2.5 mm. 161 162 All images were reviewed by the Imaging Data Evaluation and Analysis (IDEA) Reading 163 Center for image quality, and accurate segmentation of the mRNFL, ganglion cell layer 164 (GCL), and inner plexiform layer (IPL). The automated Spectralis software segmentation was 165 manually corrected if needed, according to the standard IDEA Reading Center protocols.19 166 167 Choroidal thickness measurement using deep learning 168 As choroidal thickness is not available from standard software, custom deep learning-based 169 software was developed to automatically measure mCT.21 A trained grader (JR) manually 170 segmented the Bruch’s Membrane (BM) / RPE complex and the posterior boundary of the 171 choroid in 120 p-Pole scans in the SPX software (version 1.9.204.0; Heidelberg Engineering 172 Inc, Heidelberg, Germany) in a subset of 20 eyes, which was used as ground truth to train a 173 deep convolutional neural network model (BCDU-Net).22 Two thousand two hundred fifty 174 seven scans (753 eyes) with automated choroid segmentation were reviewed for accuracy 175 (JR). The overall performance of the deep learning algorithm for segmenting the choroid was 176 very good with 400/401 (99.8%) eyes included (no myopia: 146/146 (100%), mild myopia: 177 208/208 (100%) and high myopia: 46/47 (97.9%)). 178 179 Macular choroidal thickness (mCT) was obtained for the inner and outer rings of the macular 180 p-Pole scans defined above. Each ring was subdivided into temporal, superior, nasal, and 181 inferior sectors and global and sectoral choroidal thickness was calculated. The MCT was 182 defined as the perpendicular distance between the posterior border of BM / retinal pigment 183 epithelium (RPE) complex and the posterior boundary of the choroid. 184 185 Statistical Analyses 186 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 8 R ez a po ur J Data is presented as mean (95% confidence interval (CI)) and count (percentage) for 187 continuous and categorical variables, respectively. Patient and eye characteristics were 188 compared across myopia groups using analysis of variance (ANOVA) and chi-squared tests 189 for continuous and categorical patient-level variables (respectively) and linear mixed effects 190 models for continuous eye-level variables, with a random intercept to account for within-191 patient correlation. Univariable and age and VFMD adjusted multivariable models were 192 applied to evaluate the association between axial length and ocular parameters. P-values 193 less than 0.05 were considered statistically significant. All statistical analyses were 194 performed using R (version 3.6.3). 195 196

197 Four-hundred-one glaucoma eyes of 248 patients were included with 146 eyes (87 patients) 198 in the non-myopic group, 208 eyes (125 patients) in the mild myopic group, and 47 highly 199 myopic eyes (36 patients) (Table 1). All p-values are reported as age-adjusted. 200 No axial myopia (n=87; 146 eyes) Mild axial myopia (n=125; 208 eyes) High axial myopia (n=36; 47 eyes) Overall (n=248, 401 eyes) P-value Age- adjusted p-value Age 76.5 (74.1, 78.8 73.0 (71.1, 75.0) 67.5 (63.7, 71.4) 73.4 (72.0, 74.8) <0.001 1,2,3 Gender Female 55 (63.2%) 55 (44.0%) 13 (36.1%) 123 (49.6%) 0.005 1,2 Male 32 (36.8%) 70 (56.0%) 23 (63.9%) 125 (50.4%) Race African Descent 21 (24.1%) 22 (17.6%) 3 (8.3%) 46 (18.5%) 0.057 2,3 Asian Descent 7 (8.0%) 15 (12.0%) 10 (27.8%) 32 (12.9%) European Descent 57 (65.5%) 84 (67.2%) 21 (58.3%) 162 (65.3%) Axial length (mm) 23.4 (23.3, 23.6) 24.9 (24.8, 25.0) 26.5 (26.3, 26.7) 24.5 (24.4, 24.7) <0.001 1,2,3 <0.0011,2,3 SE (dpt) −0.00 (−0.41, 0.41) −1.50 (−1.85, −1.16) −2.84 (−3.42, −2.26) −1.12 (−1.44, −0.79) <0.0011,2,3 <0.0011,2,3 CCT 537.9 (529.5, 546.3) 534.5 (527.4, 541.7) 535.5 (523.0, 548.1) 535.9 (530.2, 541.6) 0.803 0.637 VFMD (db) -5.75 (-7.05, -4.45) -7.10 (-8.19, -6.00) -7.18 (-9.32, -5.04) -6.61 (-7.43, - 5.80) 0.251 IOP (mmHg) 14.6 (13.7, 15.5) 13.8 (13.1, 14.6) 14.1 (12.6, 15.5) 14.1 (13.6, 14.7) 0.423 0.0931 Cataract Surgery Yes 63 (43.2%) 78 (37.5%) 14 (29.8%) 155 (38.7%) 0.236 No 83 (56.8%) 130 (62.5%) 33 (70.2%) 246 (61.3%) Refractive Surgery Yes 0 (0.0%) 12 (5.8%) 2 (4.3%) 14 (3.5%) 0.003 1 No 146 (100.0%) 196 (94.2%) 45 (95.7%) 387 (96.5%) Table 1: Glaucoma patient and eye characteristics by myopia group. 201

are presented as mean (95% confidence interval) or percentage. Race was compared using 202 a chi-squared test. Continuous variables were compared using ANOVA (for age) or linear mixed models 203 (for eye level data). 204 No myopia: AL ≤ 24.0mm; Mild myopia: AL: >24mm and ≤ 26.0mm; High myopia: AL >26.0mm 205 Missing 13 a , 2b , and 8c values. 206 1 No vs. Mild Myopia p < 0.05; 2 No vs. High Myopia p < 0.05; 3 Mild vs. High Myopia p < 0.05 207 Abbreviations: BMO; Bruch’s membrane opening, IOP; intraocular pressure, MD; mean deviation 208 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 9 R ez a po ur J The participants in the high myopia group were significantly younger (mean [95% CI]) 67.5 209 [63.7, 71.4] years) than the members of the mild (73.0 [71.1, 75.0] years) myopic group and 210 the non-myopic individuals (76.5 [74.1, 78.8] years) groups (p<0.001). There was a trend of a 211 higher proportion of individuals of Asian descent in the high myopia group compared to the 212 no-myopic and the mild myopic group (p=0.06). 213 214 There was no significant difference in intraocular pressure (p=0.09), central corneal thickness 215 (p=0.64) and BMO area (p=0.51) among the three groups (Table 1). Non-myopes tended 216 (p=0.131) to have less severe glaucoma than the mild myopic group and the high myopic 217 group (mean visual field mean deviation (MD) -5.75 dB, -7.10 dB and -7.18 dB, respectively). 218 219 Macular Thickness Measurements 220 A total of 43 eyes were excluded from the analysis for not meeting image quality criteria 221 (Spectralis quality score >15 dB or segmentation failure) with 14/140 (10.1%), 12/220 (5.5%) 222 and 17/64 (26.6%) eyes excluded from the no-, mild- and high-axial myopia groups, 223 respectively. Macular thickness measures are presented in Figure 2 and Supplemental Table 224 1. 225 226 Associations with Axial Length 227 There were no statistically significant associations between global and sectoral GCC or 228 GCPIL thickness measurements and axial length except for a weak association of the GCIPL 229 outer nasal sector (R²=1.9%, p=0.016). All mRNFL measurements, except for the inner 230 temporal and outer inferior sector were significantly (all p<0.024) but relatively weakly (all R² 231 < 5%) associated with axial length. We found weak associations between global and sectoral 232 vessel density measurements and axial length (all R²<3.2%, all p<0.05) and stronger 233 associations between choroidal thickness measures and axial length (R² range: 9.6%-19.3%, 234 all p<0.001) (Table 2). 235 236 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 10 R ez a po ur J Patients (eyes) Univariable Regression Multivariable Regression Estimate R 2 (p-value) Estimate R2 (%) (p-value) Age N=248 (401) -0.26 (-0.59, 0.07) 0.1 (0.121) -0.25 (-0.58, 0.07) 0.1 (0.130) VFMD N=248 (401) -0.35 (-0.97, 0.27) 0.4 (0.272) -0.53 (-1.17, 0.12) 0.8 (0.110) Spectralis GCIPL (µm) Global N=248 (401) -0.29 (-1.21, 0.63) 0.1 (0.531) -0.41 (-1.19, 0.36) 0.4 (0.299) Inner ring N=248 (401) -0.32 (-1.58, 0.94) 0.1 (0.618) -0.36 (-1.42, 0.69) 0.2 (0.498) Outer ring N=248 (401) -0.26 (-0.92, 0.40) 0.2 (0.443) -0.45 (-1.05, 0.14) 0.7 (0.136) Inner temporal N=248 (401) 0.14 (-1.26, 1.54) 0.0 (0.843) 0.29 (-0.86, 1.43) 0.1 (0.625) Inner superior N=248 (401) -0.60 (-1.90, 0.71) 0.3 (0.370) -0.82 (-2.01, 0.37) 0.6 (0.180) Inner nasal N=248 (401) -0.61 (-1.96, 0.74) 0.3 (0.380) -0.76 (-1.99, 0.47) 0.5 (0.226) Inner inferior N=248 (401) -0.21 (-1.69, 1.26) 0.0 (0.780) -0.11 (-1.37, 1.15) 0.0 (0.859) Outer temporal N=248 (401) -0.20 (-1.01, 0.62) 0.1 (0.637) -0.20 (-0.91, 0.51) 0.1 (0.577) Outer superior N=248 (401) -0.21 (-0.92, 0.50) 0.1 (0.567) -0.47 (-1.14, 0.20) 0.6 (0.170) Outer nasal N=248 (401) -0.49 (-1.28, 0.30) 0.5 (0.225) -0.92 (-1.66, -0.18) 1.9 (0.016) Outer inferior N=248 (401) -0.18 (-0.84, 0.48) 0.1 (0.595) -0.27 (-0.90, 0.37) 0.2 (0.411) Spectralis RNFL (µm) Global N=248 (401) 0.33 (-0.05, 0.71) 0.9 (0.092) 0.59 (0.24, 0.95) 3.4 (0.001) Inner ring N=248 (401) 0.32 (0.06, 0.59) 1.8 (0.018) 0.49 (0.22, 0.76) 3.9 (<0.001) Outer ring N=248 (401) 0.33 (-0.21, 0.87) 0.5 (0.233) 0.70 (0.21, 1.19) 2.6 (0.005) Inner temporal N=248 (401) 0.04 (-0.16, 0.24) 0.1 (0.671) 0.19 (-0.01, 0.39) 1.0 (0.066) Inner superior N=248 (401) 0.55 (0.18, 0.92) 2.7 (0.004) 0.74 (0.35, 1.12) 4.4 (<0.001) Inner nasal N=248 (401) 0.44 (0.09, 0.79) 1.9 (0.014) 0.62 (0.25, 0.98) 3.4 (<0.001) Inner inferior N=248 (401) 0.27 (-0.09, 0.63) 0.7 (0.141) 0.44 (0.09, 0.80) 1.8 (0.015) Outer temporal N=248 (401) 0.07 (-0.13, 0.28) 0.2 (0.483) 0.23 (0.03, 0.44) 1.6 (0.024) Outer superior N=248 (401) 0.44 (-0.28, 1.16) 0.5 (0.231) 0.83 (0.15, 1.51) 1.8 (0.017) Outer nasal N=248 (401) 0.60 (-0.23, 1.43) 0.6 (0.160) 1.14 (0.36, 1.93) 2.7 (0.004) Outer inferior N=248 (401) 0.17 (-0.57, 0.91) 0.1 (0.654) 0.59 (-0.09, 1.26) 0.9 (0.088) Spectralis GCC (µm) Global N=248 (401) 0.03 (-1.19, 1.25) 0.0 (0.964) 0.18 (-0.88, 1.25) 0.0 (0.737) Inner ring N=248 (401) 0.00 (-1.41, 1.41) 0.0 (0.998) 0.12 (-1.10, 1.35) 0.0 (0.844) Outer ring N=248 (401) 0.07 (-1.04, 1.17) 0.0 (0.909) 0.25 (-0.75, 1.24) 0.1 (0.625) Inner temporal N=248 (401) 0.18 (-1.24, 1.61) 0.0 (0.800) 0.47 (-0.72, 1.66) 0.2 (0.438) Inner superior N=248 (401) -0.06 (-1.59, 1.48) 0.0 (0.943) -0.08 (-1.53, 1.36) 0.0 (0.910) Inner nasal N=248 (401) -0.16 (-1.68, 1.37) 0.0 (0.840) -0.14 (-1.58, 1.31) 0.0 (0.852) Inner inferior N=248 (401) 0.06 (-1.68, 1.80) 0.0 (0.947) 0.33 (-1.20, 1.85) 0.1 (0.676) Outer temporal N=248 (401) -0.12 (-1.05, 0.82) 0.0 (0.805) 0.03 (-0.80, 0.87) 0.0 (0.940) Outer superior N=248 (401) 0.23 (-1.07, 1.53) 0.0 (0.725) 0.36 (-0.87, 1.58) 0.1 (0.569) Outer nasal N=248 (401) 0.12 (-1.31, 1.55) 0.0 (0.873) 0.22 (-1.13, 1.56) 0.0 (0.750) Outer inferior N=248 (401) -0.02 (-1.30, 1.27) 0.0 (0.978) 0.31 (-0.88, 1.51) 0.1 (0.608) Avanti GCC (µm) Whole image N=204 (317) -0.04 (-1.44, 1.36) 0.0 (0.954) 0.02 (-1.23, 1.26) 0.0 (0.979) Parafovea N=204 (317) -0.15 (-1.64, 1.33) 0.0 (0.839) -0.13 (-1.45, 1.19) 0.0 (0.850) Temporal N=204 (317) -0.11 (-1.60, 1.38) 0.0 (0.888) 0.07 (-1.20, 1.33) 0.0 (0.920) Superior N=203 (316) -0.15 (-1.81, 1.50) 0.0 (0.857) -0.16 (-1.76, 1.43) 0.0 (0.839) Nasal N=204 (317) -0.11 (-1.71, 1.48) 0.0 (0.891) -0.25 (-1.75, 1.26) 0.0 (0.750) Inferior N=204 (316) -0.31 (-2.14, 1.52) 0.0 (0.739) -0.28 (-1.89, 1.34) 0.0 (0.739) Avanti Vessel Density (%) Whole image N=204 (317) -0.35 (-0.87, 0.17) 0.7 (0.191) -0.54 (-0.97, -0.10) 2.3 (0.016) Parafovea N=204 (317) -0.42 (-0.97, 0.13) 0.9 (0.138) -0.63 (-1.09, -0.18) 2.8 (0.007) Temporal N=204 (317) -0.38 (-0.95, 0.18) 0.7 (0.183) -0.49 (-0.96, -0.01) 1.5 (0.046) Superior N=203 (316) -0.31 (-0.88, 0.26) 0.4 (0.293) -0.59 (-1.09, -0.09) 2.0 (0.022) Nasal N=204 (317) -0.48 (-1.05, 0.08) 1.1 (0.096) -0.75 (-1.24, -0.25) 3.2 (0.004) Inferior N=204 (316) -0.57 (-1.26, 0.11) 1.0 (0.102) -0.78 (-1.36, -0.21) 2.7 (0.008) Spectralis Choroid (µm) Global N=247 (400) -10.95 (-14.96, -6.94) 8.9 (<0.001) -15.17 (-18.96, -11.38) 17.3 (<0.001) Inner Ring N=247 (400) -11.51 (-16.18, -6.84) 7.5 (<0.001) -16.04 (-20.49, -11.58) 14.6 (<0.001) Outer Ring N=247 (400) -10.81 (-14.70, -6.92) 9.2 (<0.001) -14.97 (-18.64, -11.30) 17.9 (<0.001) Inner Temporal N=247 (400) -8.05 (-12.53, -3.57) 4.1 (<0.001) -12.33 (-16.64, -8.02) 9.6 (<0.001) Inner Superior N=247 (400) -12.56 (-17.72, -7.40) 7.3 (<0.001) -17.30 (-22.26, -12.34) 13.8 (<0.001) Inner Nasal N=247 (400) -16.87 (-22.38, -11.36) 11.1 (<0.001) -21.73 (-27.06, -16.41) 17.9 (<0.001) Inner Inferior N=247 (400) -9.28 (-14.19, -4.37) 4.5 (<0.001) -14.19 (-18.87, -9.51) 10.8 (<0.001) Outer Temporal N=247 (400) -6.56 (-10.30, -2.83) 4.0 (<0.001) -11.03 (-14.47, -7.60) 11.8 (<0.001) Outer Superior N=247 (400) -13.07 (-17.89, -8.24) 8.9 (<0.001) -17.45 (-22.12, -12.78) 15.5 (<0.001) Outer Nasal N=247 (400) -14.37 (-18.70, -10.04) 12.8 (<0.001) -18.10 (-22.34, -13.86) 19.3 (<0.001) Outer Inferior N=247 (400) -10.15 (-14.80, -5.50) 6.0 (<0.001) -14.92 (-19.32, -10.51) 13.1 (<0.001) Table 2. Ocular associations with axial length 237 *Linear mixed models slope estimates (with 95% confidence intervals) from univariable and multivariable models 238 adjusted for age and VFMD. ^R 2 reported as a percentage 239 Abbreviations: GCC; Ganglion cell complex, GCIPL; Ganglion cell inner plexiform layer, RNFL; Retinal nerve fiber layer, VFMD; 240 Visual field mean deviation 241 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 11 R ez a po ur J Associations with Severity of Glaucoma (Visual Field MD) 242 In multivariable models adjusted for age and axial length we found relatively strong 243 associations between thinner global and sectoral GCIPL measures and worse VFMD (R² 244 range: 14.0%-38.1%, all p<0.001). Thinner mRNFL was also significantly associated with 245 worse VFMD in all sectors (R² range: 3.1%-23.8%, all p0.285). Thinner global and sectoral Spectralis and Avanti 247 GCC measures were significantly associated with worse VFMD (R² range: from 20.2% to 248 37.0% and 18.6% to 35.4%, all p<0.001, respectively). In addition, we found a relatively 249 strong association between lower macular vessel density and worse VFMD (R² ranged from 250 20.3% to 33.2%, all p<0.001). Macular choroidal thickness was not associated with VFMD 251 (Table 3). 252 253 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 12 R ez a po ur J Patients (eyes) Univariable Regression Multivariable Regression Estimate R 2 (p-value) Estimate R2 (%) (p-value) Age N=248 (401) -0.006 (-0.025, 0.014) 0.0 (0.558) -0.005 (-0.024, 0.015) 0.0 (0.636) Axial length N=248 (401) 0.002 (-0.007, 0.010) 0.0 (0.679) 0.001 (-0.008, 0.009) 0.0 (0.860) Spectralis GCIPL (µm) Global N=248 (401) 0.88 (0.77, 0.99) 35.7 (<0.001) 0.86 (0.75, 0.97) 35.1 (<0.001) Inner ring N=248 (401) 1.25 (1.10, 1.40) 38.1 (<0.001) 1.23 (1.08, 1.38) 37.5 (<0.001) Outer ring N=248 (401) 0.51 (0.42, 0.59) 23.7 (<0.001) 0.49 (0.41, 0.57) 23.1 (<0.001) Inner temporal N=248 (401) 1.40 (1.23, 1.58) 38.6 (<0.001) 1.39 (1.21, 1.57) 38.1 (<0.001) Inner superior N=248 (401) 1.02 (0.85, 1.19) 24.0 (<0.001) 0.99 (0.82, 1.16) 23.1 (<0.001) Inner nasal N=248 (401) 1.07 (0.90, 1.24) 25.1 (<0.001) 1.05 (0.88, 1.21) 24.4 (<0.001) Inner inferior N=248 (401) 1.45 (1.25, 1.64) 35.6 (<0.001) 1.43 (1.23, 1.62) 34.8 (<0.001) Outer temporal N=248 (401) 0.73 (0.63, 0.84) 31.5 (<0.001) 0.72 (0.61, 0.82) 30.7 (<0.001) Outer superior N=248 (401) 0.43 (0.33, 0.53) 14.7 (<0.001) 0.41 (0.31, 0.51) 13.8 (<0.001) Outer nasal N=248 (401) 0.46 (0.35, 0.56) 13.6 (<0.001) 0.43 (0.33, 0.53) 13.0 (<0.001) Outer inferior N=248 (401) 0.40 (0.31, 0.49) 14.7 (<0.001) 0.39 (0.30, 0.48) 14.0 (<0.001) Spectralis RNFL (µm) Global N=248 (401) 0.26 (0.21, 0.31) 18.8 (<0.001) 0.27 (0.22, 0.32) 20.5 (<0.001) Inner ring N=248 (401) 0.08 (0.04, 0.12) 3.5 (<0.001) 0.09 (0.05, 0.13) 4.4 (<0.001) Outer ring N=248 (401) 0.44 (0.37, 0.51) 26.2 (<0.001) 0.45 (0.38, 0.52) 27.5 (<0.001) Inner temporal N=248 (401) 0.01 (-0.02, 0.04) 0.1 (0.599) 0.02 (-0.01, 0.05) 0.3 (0.285) Inner superior N=248 (401) 0.09 (0.04, 0.15) 2.5 (0.001) 0.10 (0.05, 0.16) 3.1 (<0.001) Inner nasal N=248 (401) 0.02 (-0.04, 0.07) 0.1 (0.543) 0.03 (-0.03, 0.08) 0.3 (0.315) Inner inferior N=248 (401) 0.19 (0.13, 0.25) 10.4 (<0.001) 0.20 (0.14, 0.25) 11.3 (<0.001) Outer temporal N=248 (401) 0.05 (0.02, 0.09) 2.8 (<0.001) 0.06 (0.03, 0.10) 4.0 (<0.001) Outer superior N=248 (401) 0.52 (0.42, 0.62) 20.1 (<0.001) 0.53 (0.43, 0.63) 21.0 (<0.001) Outer nasal N=248 (401) 0.63 (0.53, 0.74) 22.3 (<0.001) 0.65 (0.54, 0.75) 23.4 (<0.001) Outer inferior N=248 (401) 0.56 (0.46, 0.67) 23.0 (<0.001) 0.58 (0.48, 0.68) 23.8 (<0.001) Spectralis GCC (µm) Global N=248 (401) 1.14 (0.99, 1.29) 34.0 (<0.001) 1.13 (0.98, 1.28) 33.4 (<0.001) Inner ring N=248 (401) 1.33 (1.16, 1.51) 34.5 (<0.001) 1.32 (1.15, 1.49) 33.9 (<0.001) Outer ring N=248 (401) 0.95 (0.81, 1.08) 29.1 (<0.001) 0.94 (0.81, 1.08) 28.6 (<0.001) Inner temporal N=248 (401) 1.41 (1.23, 1.59) 37.5 (<0.001) 1.41 (1.23, 1.59) 37.0 (<0.001) Inner superior N=248 (401) 1.12 (0.91, 1.32) 20.9 (<0.001) 1.10 (0.89, 1.31) 20.2 (<0.001) Inner nasal N=248 (401) 1.10 (0.90, 1.29) 20.8 (<0.001) 1.09 (0.89, 1.28) 20.2 (<0.001) Inner inferior N=248 (401) 1.64 (1.41, 1.87) 32.8 (<0.001) 1.63 (1.39, 1.86) 32.2 (<0.001) Outer temporal N=248 (401) 0.78 (0.66, 0.91) 28.0 (<0.001) 0.78 (0.66, 0.90) 27.3 (<0.001) Outer superior N=248 (401) 0.95 (0.77, 1.13) 20.8 (<0.001) 0.94 (0.76, 1.12) 20.4 (<0.001) Outer nasal N=248 (401) 1.09 (0.91, 1.27) 22.9 (<0.001) 1.08 (0.90, 1.26) 22.4 (<0.001) Outer inferior N=248 (401) 0.96 (0.79, 1.13) 22.4 (<0.001) 0.96 (0.79, 1.14) 22.1 (<0.001) Avanti GCC (µm) Whole image N=204 (317) 1.23 (1.05, 1.42) 31.2 (<0.001) 1.23 (1.04, 1.42) 30.9 (<0.001) Parafovea N=204 (317) 1.33 (1.13, 1.53) 31.9 (<0.001) 1.32 (1.12, 1.52) 31.6 (<0.001) Temporal N=204 (317) 1.42 (1.21, 1.63) 35.7 (<0.001) 1.41 (1.20, 1.62) 35.4 (<0.001) Superior N=203 (316) 1.13 (0.88, 1.37) 18.9 (<0.001) 1.12 (0.87, 1.37) 18.6 (<0.001) Nasal N=204 (317) 1.11 (0.90, 1.32) 20.2 (<0.001) 1.10 (0.89, 1.32) 20.0 (<0.001) Inferior N=204 (316) 1.67 (1.40, 1.93) 32.3 (<0.001) 1.65 (1.39, 1.92) 31.9 (<0.001) Avanti Vessel Density (%) Whole image N=204 (317) 0.46 (0.38, 0.53) 31.1 (<0.001) 0.44 (0.37, 0.52) 31.8 (<0.001) Parafovea N=204 (317) 0.48 (0.40, 0.56) 31.0 (<0.001) 0.47 (0.39, 0.54) 31.9 (<0.001) Temporal N=204 (317) 0.50 (0.42, 0.59) 31.7 (<0.001) 0.49 (0.41, 0.58) 31.6 (<0.001) Superior N=203 (316) 0.42 (0.33, 0.50) 21.6 (<0.001) 0.40 (0.32, 0.49) 22.1 (<0.001) Nasal N=204 (317) 0.39 (0.31, 0.48) 19.9 (<0.001) 0.38 (0.30, 0.47) 20.3 (<0.001) Inferior N=204 (316) 0.62 (0.53, 0.72) 32.8 (<0.001) 0.61 (0.51, 0.70) 33.2 (<0.001) Spectralis Choroid (µm) Global N=247 (400) 0.35 (-0.11, 0.80) 0.3 (0.137) 0.25 (-0.19, 0.69) 0.2 (0.262) Inner Ring N=247 (400) 0.51 (-0.04, 1.06) 0.5 (0.068) 0.39 (-0.13, 0.92) 0.4 (0.144) Outer Ring N=247 (400) 0.31 (-0.14, 0.75) 0.2 (0.177) 0.21 (-0.22, 0.63) 0.2 (0.335) Inner Temporal N=247 (400) 0.43 (-0.17, 1.04) 0.4 (0.159) 0.27 (-0.30, 0.85) 0.2 (0.355) Inner Superior N=247 (400) 0.67 (0.01, 1.33) 0.7 (0.046) 0.49 (-0.14, 1.12) 0.5 (0.127) Inner Nasal N=247 (400) 0.77 (0.10, 1.44) 0.8 (0.025) 0.59 (-0.05, 1.22) 0.6 (0.071) Inner Inferior N=247 (400) 0.48 (-0.14, 1.10) 0.4 (0.130) 0.33 (-0.26, 0.93) 0.2 (0.274) Outer Temporal N=247 (400) 0.40 (-0.09, 0.88) 0.5 (0.108) 0.25 (-0.20, 0.71) 0.3 (0.281) Outer Superior N=247 (400) 0.27 (-0.33, 0.88) 0.1 (0.376) 0.12 (-0.46, 0.69) 0.0 (0.687) Outer Nasal N=247 (400) 0.34 (-0.20, 0.87) 0.2 (0.217) 0.18 (-0.32, 0.69) 0.1 (0.477) Outer Inferior N=247 (400) 0.46 (-0.11, 1.02) 0.4 (0.118) 0.32 (-0.23, 0.86) 0.2 (0.253) Table 3. Ocular characteristics associations with Visual Field mean deviation 254 *Linear mixed models slope estimates (with 95% confidence intervals) from univariable and multivariable models 255 adjusted for age and axial length. 256 ^R2 reported as a percentage 257 Abbreviations: GCC; Ganglion cell complex, GCIPL; Ganglion cell inner plexiform layer, RNFL; Retinal nerve fiber layer 258 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 13 R ez a po ur J Secondary Analysis of Differences by Axial Myopia Group 259 As a secondary analysis, we compared macular thickness- and vascular measurement 260 differences across the three axial myopia groups and adjusted for age and VFMD. In 261 general, mRNFL was thickest in high myopic eyes in all sectors, while mCT was significantly 262 thinner in all sectors in high myopic eyes. The pattern of other macular thickness and vessel 263 density measurements were less consistent across the three axial myopia groups 264 (Supplemental Table 1 and Figures 2A-2F.) 265 266 Specifically, global and sectoral GCC (both Spectralis and Avanti) and GCIPL thickness 267 values were similar across myopic groups, except for the inner and outer nasal rings, and 268 inferior outer ring GCIPL (all p≤ 0.033) (see Figures 2A, 2C and 2D). Compared to no and 269 mild myopia groups, thicker mRNFL was generally found in high myopes globally and in 270 specific sectors (age and VF adjusted MD global: p=0.031, global inner: p=0.051, global 271 outer p=0.042, inner superior ring p=0.001 and outer superior ring (p=0.017) (Figure 2B). 272 Parafoveal vessel density tended to be slightly higher in non-myopes compared to mild and 273 high myopes, but only reached statistical significance in the nasal sector (mean [95% CI]); 274 nasal vessel density in high myopes (42.7% [40.7%, 44.6%]) and non-myopes (44.5% 275 [43.3%, 45.7%]) (p=0.011) (Figure 2E). Global and sectoral mean MCT was significantly 276 thinner in high myopes compared to mild and non-myopes (all p<0.001, See Figure 2F). 277 278

279 The results of this work have implications for diagnosing glaucoma in the challenging 280 patients with high myopia. Specifically, our results suggest that macula measurements can 281 be useful measurements to diagnose and monitor glaucoma in myopic eyes as the GCIPL 282 and GCC thinned with increasing severity of glaucoma but are not associated with axial 283 length. Except for choroidal thickness, all other macula thickness measures obtained in this 284 study were associated with the severity of glaucoma. Because ganglion cell-related macular 285 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 14 R ez a po ur J thickness measurements are strongly associated with VFMD but do not vary with axial 286 length, GCIPL and GCC show promise for detecting glaucoma in myopic eyes. 287 288 Because the macula is devoid of morphometric variations such as tilt and peripapillary 289 atrophy, it’s diagnostic role in detecting glaucoma in highly myopic eyes is gaining more 290 attention recently. Specifically, there is evidence that myopia can lead to a high rate of false-291 positives in the measurement of the peripapillary RNFL (pRNFL).23 Several studies have 292 examined the diagnostic ability of GCIPL, pRNFL and GCC and reported that GCIPL 293 thickness has been reported as superior 8 9 24-27 or comparable to pRNFL thickness28 for 294 diagnosing glaucoma in myopic eyes. Shoji et al. found that GCC parameters had high 295 diagnostic accuracy to detect glaucoma in highly myopic eyes and that the diagnostic ability 296 was higher than that of the pRNFL.8 In another study, these authors reported that GCC 297 parameters were not significantly related to refractive errors and had good accuracy to detect 298 glaucoma in non-myopes and in high myopes.9 Similarly, Kim et al. determined that in highly 299 myopic eyes, the accuracy of glaucoma detection based on the macular GCC thickness was 300 comparable to that based on the pRNFL thickness.29 These findings and those of other 301 studies8 9 27 are consistent with our results that GCIPL and GCC thickness measured using 302 both Spectralis and Avanti showed no association between the GCIPL thickness and axial 303 length suggesting GCIPL thickness is less sensitive to changes due to axial elongation. 304 Moreover, in our study GCIPL and GCC measurements showed the strongest association 305 with VFMD, suggesting that both are useful for measuring ganglion cell loss associated with 306 glaucoma in both non-highly myopic and highly myopic eyes. These results are generalizable 307 across instruments as both Spectralis GCC and Avanti GCC showed similar results. 308 309 High myopia is characterized by marked structural changes in the retina and choroid and the 310 corresponding vasculature.30 31 With the introduction of the non-invasive technique, OCTA 311 images can provide a microvascular map from different retinal layers. To the best of our 312 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 15 R ez a po ur J knowledge this is the first study comparing both macular tissue thickness and vascular 313 measurements in axial non-myopic, mild and high myopic glaucomatous eyes. 314 In the current study, the macular vessel density showed a weak association with axial length 315 and a moderate association with VFMD. Previous studies reported conflicting results. This 316 inconsistency can be explained in part by differences in study populations and image 317 acquisition and analysis protocols. For instance, we employed a 3 x 3 mm imaging area 318 whereas Yang et al. employed a larger 6 x 6 mm area.31 A large scan size can be more 319 sensitive to image artefacts but also may identify microvasculature dropout in outer regions.31 320 We found only a weak association between superficial macula vessel density and axial 321 length but a moderate association between vessel density and VFMD. Our results suggest 322 that although myopic changes might affect vessel density in the macula, effects due to 323 glaucoma are much stronger as indicated by the stronger association to the VFMD and 324 therefore may also be useful for monitoring glaucoma in myopic eyes.32 325 326 In terms of choroidal thickness, as axial length increased, the choroid thinned in all sectors. 327 Previously reported results have demonstrated choroidal thinning in highly myopic eyes.33-35 328 Ho et al.33 reported that subfoveal choroidal thickness decreased by 6.20 µm for each diopter 329 of myopia and was thinnest in the nasal sectors in all groups, which is similar to the 330 distributions of non-axial myopes, mild axial myopes and high axial myopes in our study. The 331 choroid is a highly vascular layer, supplied by the posterior ciliary arteries and provides the 332 retinal photoreceptors and the retinal pigment epithelium with oxygen and nourishment.36 Our 333

did not show an association between choroidal thickness and VFMD which suggests 334 that choroidal thickness likely is not a useful metric for differentiating glaucomatous from 335 healthy eyes or for monitoring glaucomatous progression. 336 337 The current study has several limitations. First, individuals with high myopia were younger. 338 As retinal tissue is known to thin in older eyes37 we adjusted for age and VFMD in all 339 analyses. In addition, we compared the 3 axial myopic groups after age-matching and found 340 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 16 R ez a po ur J similar results (data not shown) with respect to the pattern of the retinal and vascular 341 measurements in the three groups. Second, it has been suggested that axial length might 342 affect retinal vessel density measurements and lead to incorrect scaling in OCTA imaging, 343 which should be taken into account when interpreting our results.38 Moreover, as vessel 344 density measurements vary across instruments,39 these sVD results are not necessarily 345 generalizable to macula vessel density measurements from other OCTA instruments or to 346 macula deep layer vessel density measurements. In addition, axial elongation often leads to 347 retinal layer segmentation errors and measurement failures. However, we reviewed the OCT 348 images meticulously for segmentation errors and excluded data with uncorrectable 349 segmentation failures. Finally, the sample size of the high myopic group was relatively small 350 compared to the other two groups and the mean axial length was only 26.5 mm. We can 351 therefore not generalize our results to eyes with longer axial length. 352 353 In conclusion, GCIPL and GCC thickness can be useful measurements to diagnose and 354 monitor glaucoma in myopic eyes as they thinned with increasing severity of glaucoma but 355 did not vary with axial length. Macular sVD may also be useful for detecting glaucoma in 356 myopic eyes, however we found a weak association between vessel density and axial length 357 which needs to be explored further. 358 359 Funding 360 Grant support: 361 JR: German Research Foundation research fellowship grant recipient (RE 4155/1-1) and 362 German Ophthalmological Society Grant 363 MC: K99EY030942 364 SM: Tobacco-Related Disease Research Program T31IP1511 365 RNW: National Eye Institute R01EY029058, an Unrestricted grant from Research to Prevent 366 Blindness (New York, NY) 367 LMZ: National Eye Institute R01EY011008, R01EY019869, R01EY027510, P30EY022589 368 369 Competing Interests: 370 None: JR, CB, JD, AB, JAP, MC, LH, JBJ, RCP, SM, HH, MAF 371 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 17 R ez a po ur J RNW: Consulting: Bausch & Lomb, Eyenovia, Aerie Pharmaceuticals, Allergan; Research 372 Funding or Equipment: Bausch & Lomb, Heidelberg Engineering, Carl Zeiss Meditec, Konan 373 Medical, Genentech, Optos, Optovue, Centervue; Patent: Toromedes, Carl Zeiss Meditec-374 Zeiss 375 LMZ: Research Funding and Equipment: Heidelberg Engineering; Research Equipment: 376 Optovue Inc, Carl Zeiss Meditec Inc, Topcon Medical Systems Inc; Patent: Carl Zeiss 377 Meditec. 378 379 380 381 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 18 R ez a po ur J

382 1 . Marcu s MW , de V rie s MM, Junoy Mon toli o FG , et a l . Myopia a s a ri sk fac to r fo r ope n -angle 383 g lauc oma: a s y s temat ic r evie w and meta -analy s i s. Op ht hal mo lo gy 2 011;118(10 ):1 989-94 e 2. 384 do i: 10.10 16/ j . ophth a.2011 .03 .012 [pub l ishe d Online F i r s t: 201 1 / 06 / 21 ] 385 2 . Leu ng CK, C h eung CY , We inreb R N, e t al . Re tinal ne rve fib e r l ayer ima ging with spec tral -doma i n 386 op t i cal cohe renc e tomography : a v ariabil i t y and d iagno stic pe r fo rman ce stu dy. 387 Op ht halm olo gy 2009;116 ( 7 ):1257 -63, 63 e1-2. doi : 10.1 016/j. ophth a.200 9.04 .013 [publi shed 388 Online Fir s t: 2009/05 /26 ] 389 3 . Jona s J B, Gu sek GC , Nauman n GO . Op tic disk mo r pho me tr y in hig h myopi a. Gra efe's archi ve f or 390 c linic al a nd ex perime nt al op ht halm olo gy = A l br e cht v on Gr a e fe s Ar c hiv fu r k lini sch e und 391 e xperime n telle O p ht hal m ologi e 1988;2 2 6(6):587 -90. doi : 10.1 007/ BF02169 209 [ publi s h ed 392 Online Fir s t: 1988/01 /01 ] 393 4 . Leu ng CK, Moha m ed S , Leun g KS, et al . Re t i nal n erve fib er lay e r mea sur ement s i n myo pia: An 394 op ti cal cohe renc e t omography study . I n v e s t ig a t iv e opht hal m o lo g y & v is u al s c i en ce 395 2 006;47(12) :5 171-6. d oi : 10.116 7 / iov s.0 6-0545 [publ i s h e d O nl in e F irst: 2006/11 / 24] 396 5 . Leu ng CK, Y u M, Wein reb R N, et a l. Re tinal ne rve fi be r laye r i maging with spec tral -d omain op t ic al 397 c oherenc e tomogra phy : inte rpr eting th e RNFL maps in h e althy m yopic eye s. Inv est ig at i v e 398 o ph tha lmol og y & vi sua l sci ence 2012;5 3( 11):7194-20 0 . doi: 1 0.1167 /iov s.12 -97 26 [publi shed 399 Online Fir s t: 2012/09 /22 ] 400 6 . Curcio C A , All en KA. Topog raphy of ga n glion c ells in human r etin a. The Jour nal of co mp ara t i ve 401 n eurol ogy 1990;30 0(1) :5-25 . doi : 10.100 2 / c ne.9030 00103 [pub li s h ed O n line F i r s t: 402 1 990/10/ 0 1 ] 403 7 . Hood D C, Raz a AS, d e Mora e s C G, e t al . G la ucoma tou s da mage o f th e mac ula. P rogre ss i n retin al 404 a nd ey e res earch 2 013;32:1 -21 . doi : 10 .1 016/j.pre teye re s.2012 .08.003 [p ubli shed Online 405 F irst: 201 2/ 0 9 / 22 ] 406 8 . Shoji T, Sa to H, I s h id a M, e t a l. A ss es sme nt of glau com atou s ch ang es in subjec ts w ith high myopia 407 using s pe ctral doma in op t ica l c ohe renc e t om ography. Inve s tiga tive oph th almo l og y & vi s u al 408 sci ence 2011;5 2(2) :1098-102 . doi: 10 .116 7/iovs .10-59 22 [publi sh ed Onlin e Fir st: 2 010/11/06 ] 409 9 . Shoji T, Naga oka Y , Sat o H, e t al . Impa c t of hig h myopi a on th e per f o rma nce o f S D-OCT p a r a me ter s 410 to d ete ct gla ucoma . Gr a e fe' s archiv e for cl inica l and expe r im e nt al op ht hal m ology = Albrec ht 411 v on Gr ae fe s Archiv f ur klini s c he und exp e r i ment elle O p h t h almol ogi e 2012 ;250 (12) :184 3-9. 412 do i: 10.10 07/s0041 7 -012-1994 -8 [publi s hed Onlin e F ir s t: 20 12 / 0 5/ 05 ] 413 1 0. Akas hi A, Kanamo ri A, N ak amura M, e t al. The abili ty o f mac ular par ame t e rs a nd circum papilla r y 414 retin a l nerve f i ber l aye r by thr ee S D- OCT inst r u men t s to di agno s e hig hly my opic gl au coma. 415 In v es t ig at i v e op ht ha l m o l o g y & v i s ua l s ci en ce 201 3;54(9):6025 -3 2. doi : 1 0.1167/i o vs.13 -416 1 2630 [publ i s he d O nl ine F irs t: 2013/08 /0 3] 417 1 1. Nakan i s hi H , Akagi T, Hanga i M , et a l . Effec t of A x ial Leng t h on Macu la r Gangli on Cell Co mplex 418 Th i c kn e s s a nd o n E ar l y Gl a u co m a D ia gn os is b y S p e ctr a l- D o m ai n O p t ic a l C o he r e n ce 419 T omography . Jo urna l of gl auco m a 2016; 2 5(5): e481-90 . doi: 10. 1097 /I JG.00 0000000 0000330 420 [pub li s h e d Onlin e F irst: 2015 / 11/10 ] 421 1 2. Hung KC, W u P C, P oon YC, et al . Mac ul ar Diagno s tic Abi li ty in O CT f o r A s s e ssi ng Glau coma in High 422 My opia. Op tome try a nd vi si on s c ienc e : offic ial p u blic a t i on of t he A me r i can Ac ad e my of 423 Op to metry 2 016;93(2) :12 6-35 . doi: 10.10 97/O PX.0000 00000000 0776 [pub li s he d O n line F irs t: 424 2 015/12/ 2 6 ] 425 1 3. Wu J, Se ba s tia n RT, Chu CJ, et al . Red uce d Macul ar Ve ssel Den si ty a nd Capi ll ar y Pe r fu si on in 426 Gla ucoma De tec te d Us i n g OCT Angio gra phy. Cu r rent e ye re searc h 2019;4 4(5 ):53 3-40. doi : 427 1 0.1080/0271 3683.201 8.15631 95 [publi she d Onlin e Fi r st: 2018 /12/24 ] 428 1 4. Moghimi S, Z angw ill L M, Pent ead o R C, e t al . Macul ar and O ptic Nerve H e ad V es se l De ns i t y and 429 P r og re ssive Re ti n al Ne rve F iber L ayer L o ss in Gla ucoma. O pht h al mo l og y 2018;12 5(11):1720 -430 2 8. doi: 1 0.1016 /j.ophtha. 2018 .05. 006 [p ubli s h ed O nli n e Fi rst: 2018/06/17 ] 431 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 19 R ez a po ur J 1 5. Gupt a P, Che ung CY, Saw SM, e t a l. P eripa pill ary choroida l thick ne ss in young A si an s w ith high 432 my opia. Inv es tig ativ e op hth alm o logy & v is u a l s c ienc e 20 15;56(3 ):147 5 -81. doi : 433 1 0.1167/iov s.14 -15 742 [publi s h ed O nli n e Fi rst: 2015 /02/07 ] 434 1 6. Yang H, Luo H , G a r di n er SK, e t a l. Fa c tor s I n f l uenci ng O ptical Cohe rence Tomo grap hy Perip a pillary 435 Ch or oi dal Th ickn e s s : A Mul tic ent er S tudy. I nve s tig ative oph th almo l ogy & vi s u al s c i enc e 436 2 019;60(2) :79 5-806. d oi : 10.116 7 / iov s.1 8-25407 [pu bli s h ed Onlin e F irst: 2019 / 02 /28] 437 1 7. Flo r e s -Mo r e no I , L ugo F , Duker JS, e t al. Th e rel ation shi p be t w e en a xial length a nd ch or oi dal 438 thic kne s s in ey e s with hig h myopi a. Amer ic an journal o f o ph thal m olo gy 2013;155 (2 ):314-19 439 e 1. doi: 10.1016 /j.aj o .2012.0 7.015 [publ i s he d O nl ine F i rst: 2012 /10/06 ] 440 1 8. Ikun o Y, Ta no Y. Re ti nal a nd cho roida l biome try in high ly my opic e yes wit h spe c t r a l -domain 441 op ti cal cohe renc e t omography . I n v e s t i gat i ve o pht h alm o l og y & v i su a l s c i en c e 442 2 009;50(8) :38 76-80. d oi : 10.116 7 / iov s.0 8-3325 [publ i s h e d O nl in e F irst: 2009/03 / 13] 443 1 9. Sa mple P A, Gi rkin CA, Zan gwill LM, e t a l. The Af r ic an D e sce nt and Gla uc oma Ev alua tio n St udy 444 ( A DA G E S) : d es i g n a nd b as e l ine d a ta. A rc h O ph thal mol 2009;127 (9 ) : 1136 -45. d oi : 445 1 0.1001/arch o phth a lmol.20 09.187 446 2 0. Ghaha ri E , Bowd C, Zang will LM, et a l . Mac ular Ve s s el Den s i ty in Glauc oma tou s Eye s With F oca l 447 L amina Cribr osa D e fe cts . J G la ucoma 201 8;27 ( 4 ):3 42-49 . doi : 448 1 0.1097/I JG .0000000 0000009 22 [publi sh ed O nli ne Fi r st: 2018 /02/21 ] 449 2 1. Rezapo ur J, B ow d C, Dohlema n J, e t a l. The i nflu ence o f ax ial my opia on optic di s c cha r a c teri stic s 450 of glauc oma e ye s . Sc i Rep 2 021;11(1):8 8 54. doi: 10.1038 / s41598 -02 1 -88406-1 [p ub lis h ed 451 Online Fir s t: 2021/04 /25 ] 452 2 2. Hongmei Song, W e ngu an Wan g, Sa ny uan Zhao, J i an bing Shen, Kin -Man L am; P yram id Dilated 453 Dee p er ConvL STM for Vide o Sali en t O bje ct De tecti onP roc eedi ng s o f t h e Europe an 454 Co n fere nce o n Computer Vi si on (EC C V), 20 18, pp. 715-731. 455 2 3. Biswa s S, Lin C , Leu ng CK. E valua ti on of a My opic N o rmativ e Dat aba se for A na ly s is of R etina l 456 N e r v e F i be r La y er Th i c kn es s . JAMA Op hth alm ol 2 016;134(9 ):1 032-9 . doi : 457 1 0.1001/jama oph thalmol .2 016.234 3 [pu blish ed Onlin e Fir st : 2016 /07/22 ] 458 2 4. Seo l BR, Jeoung JW , Park KH. Gl auc o ma Det ec tion Ability o f Macul ar G a nglio n Cell -Inn er 459 P lexifo rm L aye r T hick nes s in Myopic P re perime tric Gla ucom a. I nve sti ga tiv e o pht hal m ology & 460 v isu al scie nce 2015;56(13 ):8306 -13. d oi: 10.11 67/ i ov s.15 -18141 [publi sh ed Onlin e Fir s t: 461 2 016/01/ 0 1 ] 462 2 5. Kim Y K, Yoo BW, J e oung J W, et al . Gl auco ma- D i agno s tic Abil ity o f Gangl ion Ce ll -Inn er Pl ex ifo r m 463 L aye r Thick nes s Di f fer enc e Acr o ss T emp oral Ra ph e in Hig hly Myo pic E yes. In ves t ig at ive 464 o ph tha lmol og y & vi sua l sci ence 2016;5 7( 14):5856-63 . doi: 10 .1167 /iov s.16 -201 16 [publi shed 465 Online Fir s t: 2016/11 /02 ] 466 2 6. Baek SU, Kim KE, Kim YK, et a l. Devel opment o f Topo graphic Sco r ing Sy s t em f or Ide n t i f y ing 467 Gla ucoma in Myopi c Ey e s : A Spec tral -Do ma in OCT St udy . O pht ha lm o l og y 2018; 125(11):171 0 -468 1 9. doi: 1 0.1016 /j.ophtha. 2018 .05. 002 [p ubli s h ed O nli n e Fi rst: 2018/06/12 ] 469 2 7. Yang Z, T atham AJ, W einr eb R N , e t al . Diag no stic abili t y of mac ul ar ga nglion c ell inner pl e xiform 470 l ayer me a s ur eme nts in glauc oma u sing swe pt sourc e an d sp ec tra l domain op tical cohe r en c e 471 tomog r a phy. PL oS One 2015;10 (5 ):e 012595 7. doi: 1 0.1371/jou rnal .p one .0125 957 472 2 8. Choi Y J, Jeou ng JW, Pa r k KH, et al. G l a ucoma det ectio n a bili t y of gan glion c ell- in ner plex if or m 473 l ayer thic kne s s by sp ectr al-d oma in opt ic al c oherenc e tomogr aphy in hig h myopia . 474 In v es t ig at i v e op ht ha l m o l o g y & v i s ua l s ci en ce 201 3;54(3):2296 -3 04. doi : 10.1167/i ov s .12 -475 1 0530 [publ i s he d O nl ine F irs t: 2013/03 /0 7] 476 2 9. Kim NR, L ee E S, Se ong G J, e t al . Com paring th e gang lion c el l c omplex and r e ti na l nerve fi bre lay e r 477 me asu r e men ts by Fourie r dom ain O CT to det ec t glau coma in high myopia . The B r itis h jo ur n a l 478 o f o pht hal mol o gy 201 1;95(8):111 5 -21. d oi: 10.1136/bjo .2 010.1824 93 [pu bli shed Onlin e Fir s t: 479 2 010/09/ 0 2 ] 480 3 0. Ye J , W a ng M, S h en M, et a l. De ep Re tina l C api llary P l exus Decr ea sing Co rr e l at ed With the Ou te r 481 Re t i nal L aye r Alt era tion and Vi sual Ac ui ty Impa irment in Pa t ho logic al Myop ia . I nves t igat iv e 482 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 20 R ez a po ur J o ph tha lmol og y & vi sua l sci ence 2020;6 1( 4):45. doi : 10.1167 / i ov s.6 1.4.45 [pub li s h e d Online 483 F irst: 202 0/ 0 4 / 29 ] 484 3 1. Yang D, Ca o D, Zhang L, e t a l . Mac ula r and perip apill ary ve ssel de n s i t y in myopi c ey es o f young 485 Ch ine s e adults . Clin ical & e xperi me n t a l o pto metry 2 020;103(6) :8 30-37 . doi : 486 1 0.1111/cx o.1304 7 [pub lis hed Onli ne Fir s t: 2 020/02 / 1 4] 487 3 2. Le e K, Mae ng KJ, Ki m JY, e t al . Diagno stic abili ty of ve s s e l d en sity mea s ur ed by spec tral -doma i n 488 op t i cal cohe renc e tomography ang iogr ap hy for g lauc oma in pati ent s with hi gh myop ia. 489 S cie ntific rep or t s 2020; 10 (1):3027 . doi: 1 0.10 38/s415 98-020 -6005 1-0 [publi sh ed Online Fir st: 490 2 020/02/ 2 3 ] 491 3 3. Ho M, Li u DT, Ch an V C, e t al. Choroi d al thic kne s s mea sur ement in myopic eye s by enhan ce d 492 de pth op tica l cohe r e nce tomog r a phy. O p hth almo l ogy 2 013;120(9 ):1 909-14 . do i: 493 1 0.1016/j.oph tha . 2013.02. 005 [publi s he d O nli ne Fi rs t: 2013/05 /21] 494 3 4. Harb E, Hyma n L, G w iazda J , et a l . Ch oroida l Thic kne s s Pro f i l e s in Myopi c Ey es of Young Adult s i n 495 the C or r ec tion o f Myop ia Ev alua tion Trial Coho rt. Am J O ph t ha l mol 2015 ;160(1 ):6 2-71 e2. 496 do i: 10.10 16/ j . ajo.201 5.04 .018 497 3 5. Wei WB , Xu L, Jona s JB , e t a l . Sub fo veal choroid al thick ne ss: the B e ijing E ye S t u dy. 498 Op ht halm olo gy 2013;120 ( 1 ):175 -80. do i : 10.101 6/ j. oph tha.2012 . 07.048 [pu bli she d Online 499 F irst: 201 2/ 0 9 / 27 ] 500 3 6. Li ns en m eier R A, P ad nick - S ilve r L . Me tabol ic de p endenc e of ph oto rece p tors o n the choroi d in the 501 no r ma l and de tache d r e ti na. I n ves t i gat ive o pht h al m o lo gy & v is u al s c ie n c e 2000;4 1 (10):3117 -502 2 3. [publi sh ed Onlin e Fir st: 2000/09/01 ] 503 3 7. Hammel N , B e lghith A, B o wd C, e t al. Rate an d Pa tt ern o f Rim Ar e a Los s in He a lthy and 504 P r og re ssing Glau coma Eye s. Op ht halm ol og y 201 6;123(4):760 -7 0. d oi: 505 1 0.1016/j.oph tha . 2015.11. 018 [publi s he d O nli ne Fi rs t: 2016/01 /10] 506 3 8. Sa mpson D M , Gong P , An D , et a l . Axia l Le ngth Va riati on Impact s o n S uper ficia l Retinal Ve ssel 507 Den sity a nd F ov eal Av a scular Z one Are a Mea s ur ement s U s ing O p t ic al C oher ence 508 T omography Angiograp hy. In ves t igat i v e op ht h al m o lo gy & v is u al s c ie n c e 2017;5 8( 7):3065-72 . 509 do i: 10.11 67/ i ovs .17-21551 [publi shed O nline Fir s t: 20 17/ 0 6/ 18 ] 510 3 9. Li XX, Wu W, Zho u H, et al. A qu anti ta tive compa r i son o f fi ve op tic al c ohe renc e tomography 511 a ngiog r a phy sy ste ms i n cli nica l p er for m a nce. I nt e r na t io n al jo u r n al o f opht h a l m o lo g y 512 2 018;11(11) :1 784-95. d oi : 10.18 240 / i jo.2 018. 11.09 [pub l ishe d O n li ne F irs t: 2018 /11/20] 513 514 515 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint 21 R ez a po ur J Figure Legends 516 517 Figure 1: Optic disc photograph (left), optical coherence tomography optic nerve head en 518 face image (middle) and optical coherence tomography macula posterior pole image (right) 519 of an eye with (A) no axial myopia (axial length = 23.8 mm), (B) mild axial myopia (axial 520 length = 24.7 mm and (C) high axial myopia (axial length = 29 mm). 521 522 Figure 2: Sectoral and global thickness distribution of the Spectralis GCIPL thickness (1A), 523 Spectralis macular RNFL thickness (1B), Spectralis GCC thickness (1C), Avanti GCC 524 thickness (1D), Avanti macular vessel density (1E) and Spectralis macular choroidal 525 thickness (1F) in non-myopic, mild myopic and highly myopic glaucoma eyes. 526 Abbreviations: GCC; ganglion cell complex, GCIPL; ganglion cell inner plexiform layer, 527 RNFL; retinal nerve fiber layer 528 . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint . CC-BY-NC 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.(which was not certified by peer review)preprint The copyright holder for thisthis version posted September 10, 2021. ; https://doi.org/10.1101/2021.09.02.21263045doi: medRxiv preprint