The optic nerve head ONH) is an important anatomic

Transcription

1 Glaucoma Deep Optic Nerve Head Morphology Is Associated With Pattern of Glaucomatous Visual Field Defect in Open-Angle Glaucoma Jong Chul Han, Jae Hwan Choi, Do Young Park, Eun Jung Lee, and Changwon Kee Department of Ophthalmology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, Korea Correspondence: Changwon Kee, Department of Ophthalmology, Samsung Medical Center, Sungkyunkwan University School of Medicine, 81 Irwon-ro, Gangnam-gu, Seoul 06351, Korea; Submitted: April 16, 2018 Accepted: June 29, 2018 Citation: Han JC, Choi JH, Park DY, Lee EJ, Kee C. Deep optic nerve head morphology is associated with pattern of glaucomatous visual field defect in open-angle glaucoma. Invest Ophthalmol Vis Sci. 2018;59: PURPOSE. We investigated the relationship between visual field (VF) defect pattern and deep optic nerve head (ONH) morphology in open angle glaucoma (OAG) eyes using spectraldomain optical coherence tomography (SD-OCT). METHODS. Of 278 OAG eyes, 169 with and 55 without externally oblique border tissue (EOBT) were included in the final analysis. Enhance depth imaging SD-OCT was used to measure the deep ONH parameters, such as EOBT length, ONH tilt angle, and optic canal (OC) obliqueness. The extents and locations of the maximum deep ONH parameters were explored according to VF pattern in OAG eyes. RESULTS. OAG eyes with EOBT showed longer axial length (AL; 25.6 vs mm; P < 0.001) and greater superior VF defects (67.5% vs. 49.1%; P ¼ 0.02) compared to OAG eyes without EOBT. Multivariate logistic regression analysis revealed that relatively inferior location of maximum deep ONH parameters was associated with superior hemifield defects dominant in OAG (P < in all parameters). In addition, the locations of maximum deep ONH parameters were consistent with dominant VF defect locations in OAG eyes with AL < 24.5 mm. The presence of paracentral scotoma in OAG was associated with worse mean deviation (MD) and relatively inferior location of deep ONH parameters. CONCLUSIONS. The locations of maximum deep ONH parameters were associated with the location of dominant VF defects and presence of paracentral scotoma in OAG eyes. Keywords: optic nerve head, optic nerve head morphology, border tissue, open angle glaucoma, visual field The optic nerve head ONH) is an important anatomic structure in glaucoma pathogenesis. 1,2 It contains retinal ganglion cell (RGC) axons and lamina cribrosa (LC) composed of LC beams, which are covered by astrocytes and contain capillaries. 3 In addition, peripheral structures of the ONH, such as Bruch s membrane openings (BMO) and Elschnig s border tissues, were reported as the important ONH biomarkers for understanding glaucoma pathogenesis. 4 6 Previously, ONH surface morphology, such as parapapillary atrophy (PPA) or optic disc tilt, was shown to be associated with glaucomatous retinal nerve fiber layer (RNFL) defects and associated visual field (VF) defects especially in myopic open-angle glaucoma (OAG) eyes Recently, using optical coherence tomography (OCT), it became possible to search the deep structure of the ONH, such as the LC or BMO. 4,11,12 Reis et al. 4 observed deep ONH structures, such as BMO or the border tissue of Elschnig, using OCT and classified the border tissues into externally and internally oblique types according to their morphology. In their previous study, externally oblique type border tissue was found more frequently especially at the inferotemporal ONH area in glaucoma compared to normal control eyes. In our recent study, we defined the deep ONH-associated parameters, such as externally oblique border tissue (EOBT) length, ONH tilt angle and optic canal (OC) obliqueness using BMO and border tissue, and reported that the maximum values of the deep ONH parameters were associated with axial length (AL). 13 In myopic normal tension glaucoma (NTG) eyes, the locations of the parameters showed topographic correspondence with RNFL defect. 13 Furthermore, the deep ONH parameters could demonstrate deep ONH vertical asymmetry even in ONH without definite PPA or disc tilt. Thus, we speculated that deep ONH vertical asymmetry may exist in most eyes with various ALs, and affect the pattern of glaucomatous VF defect in OAG. To evaluate our hypothesis, we investigated whether there is correspondence between the locations of dominant VF defects and maximum deep ONH parameters in OAG eyes. To understand whether the locational correspondence exists irrespective of AL, the relationship was evaluated in OAG eyes with AL < 24.5 mm (relatively short AL). In addition, we divided the included eyes into OAG with and without paracentral scotoma and explored the characteristics of maximum deep ONH parameters according to the presence of paracentral scotoma. MATERIALS AND METHODS This cross-sectional observational study included OAG patients who visited Samsung Medical Center (Seoul, South Korea) for their first ophthalmic examination between January 2015 and March This study followed all guidelines for experimental Copyright 2018 The Authors iovs.arvojournals.org j ISSN: This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

2 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3843 investigation in human subjects, was approved by the Samsung Medical Center Institutional Board, and adhered to the tenets of the Declaration of Helsinki. Each patient underwent a comprehensive ophthalmic examination, including slit-lamp biomicroscopy, Goldmann applanation tonometry, manifest refraction, gonioscopic examination, color fundus photography (TRC-50DX; Topcon Medical System, Inc., Oakland, NJ, USA), automated perimetry using a central 30-2 Humphrey Field Analyzer (HFA model 640; Humphrey Instruments, Inc., San Leandro, CA, USA) with the Swedish interactive threshold algorithm standard, AL measurement (IOL Master; Carl Zeiss Meditec, Jena, Germany), and ultrasound pachymetry (Tomey SP-3000; Tomey Ltd., Nagoya, Japan). The extent of VF defect was measured using mean deviation (MD), pattern standard deviation (PSD), and visual field index (VFI). Reliability VF analysis was defined as a falsenegative rate < 15%, false-positive rate < 15%, and fixation loss < 20%. IOPs were measured at the first and second visits without IOP-lowering medications. Average IOP values were used in the analysis. The following criteria were required for diagnosis of OAG: (1) glaucomatous optic disc changes, such as increased cupping (vertical cup-to-disc ratio > 0.7), diffuse or focal neural rim thinning, disc hemorrhage, or RNFL defects at diagnosis; (2) glaucomatous VF defects positive by more than one reliable test for at least two of the following three criteria: (a) a cluster of three points with a probability <5% on the pattern deviation map in at least one hemifield, including at least one point with a probability less than 1% or a cluster of two points with a probability less than 1%; (b) a glaucoma hemifield test result outside the normal limits; or (v) a PSD of 95% outside the normal limits; and (3) an open angle on gonioscopic examination confirmed to have no identified causes of secondary glaucoma present. The exclusion criteria were eyes with media opacities, such as a corneal or vitreous opacity, moderate to severe cataract; systemic disease or ocular diseases that could affect VF test results; high myopia with AL > 28 mm that can have myopic degeneration or retinal schisis around the ONH; and baseline MD of 12 db or less, for which glaucomatous VF pattern would be difficult to be determined on a VF test. If the patient has OAG in both eyes, then only the eye with less severe MD was included in the analysis. Enhanced Depth Imaging OCT Enhanced depth imaging spectral-domain OCT (EDI SD-OCT; Heidelberg Engineering, Heidelberg, Germany) was used to measure deep ONH parameters. The methods have been described previously. 13 Briefly, EDI scans were obtained using 48 radial-line B-scans (each at an angle of 3.758) centered on the optic disc. Each EDI scan included an average of 20 OCT frames; every other section (24 radial EDI scans in total) was selected. OCT machine scaling was adjusted into 1:1 lm before measurement. Images were taken on a day of VF evaluation. Magnification error was corrected using a formula provided by the manufacturer based on results of autorefraction keratometry and focus setting during image acquisition. If the scan section contained a poor-quality OCT image that did not offer interpretable information with respect to BMO or border tissue due to prelaminar tissue or overlying vessels, the next image was used. If >3 of 24 radial scans were unrecognizable, the eye was excluded. Measurement of the Extent and Angular Location of Parameters Reflecting Deep ONH Morphology In the previous study, we defined the parameters reflecting deep ONH structures, such as EOBT length, ONH tilt angle and OC obliqueness and demonstrated that the parameters were correlated with AL. 13 Briefly, EOBT length was defined as the length between the end points of the EOBT tissue and ONH tilt angle was defined as the angle between the BMO and optic canal planes. The BMO plane was defined as the line connecting both BMOs. The optic canal plane was defined as the line connecting the nasal BMO and innermost margin of the EOBT. OC obliqueness was defined as the angle formed by a vertical line and the EOBT. The maximal value of each parameter was defined as the maximum deep ONH parameter. The angular location of the deep ONH parameters was measured with help of the infrared (IR) scanning laser ophthalmoscopy (SLO) image included in the SD-OCT. In case the fovea was not included in the IR photo, we aligned the IR and red-free photos using Photoshop CS5 (Adobe System, San Jose, CA, USA). The line connecting the center of the BMO and fovea was defined as the foveal-bmo (FoBMO) axis. The angular location of each parameter was defined as the angle between the location of each parameter and the FoBMO axis. If the angular location of each parameter was below the FoBMO axis, the location was assigned a positive value. Otherwise, the location of each parameter was assigned a negative value (Fig. 1). The presence, extent, and angular location of each parameter were assessed by two investigators (JCH and JWC) in a masked fashion using built-in SD-OCT software. Each investigator determined the presence of the parameters. Disagreement between the investigators was resolved by a third adjudicator (CK). In measurement of the extents and angular locations of the parameters, the average values of the two investigators (JCH and JWC) were used in the final analysis. Classification of the VF Defect Pattern in OAG We divided VF defect pattern based on the following criteria: (1) dominant VF defect location (superior vs. inferior dominant) and (2) presence of paracentral scotoma. To determine the dominant involvement location, we calculated superior and inferior average hemifield value (26 points in each hemifield) in pattern deviation plots. When one hemifield had a greater absolute value, we regarded it as the dominant VF defect location. The existence of paracentral scotoma was defined as the glaucomatous VF defect in one hemifield within 108 of fixation, with at least 1 point at P < 1% lying at the two innermost paracentral points irrespective of VF abnormality outside the central ,15 When the glaucomatous defect involved both hemifields, only the hemifield with the dominant VF defect was used to determine the existence of paracentral scotoma to minimize false-positive cases. Statistical Analysis Intrabserver (two consecutive measurements by JCH) and interobserver (measured by JCH and JWC) reproducibility were assessed by calculation of intraclass correlation coefficients (ICCs). A total of 20 randomly selected fundus red-free photos and EDI OCT photos were used to analyze intraobserver ICCs. Independent t-test or Mann-Whitney U test was used to compare variables between the OAG eyes with and without EOBT presence and between the OAG eyes with and without paracentral scotoma. The v 2 test or Fisher s exact test was used to compare categorical data. Continuous variables were expressed as mean 6 SD (range). To calculate odds ratio (OR) of potential risk factors for superior dominant VF defect and presence of paracentral scotoma in OAG eyes, univariate and multiple logistic regression analyses were performed. Parameters with P < 0.2 on univariate analysis were included in the multivariate logistic regression analysis and ORs with 95%

3 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3844 FIGURE 1. Measurements of the extents and the locations of deep ONH parameters, such as EOBT length, ONH tilt angle, and OC obliqueness. (A) BMOs (yellow dots) were marked at 24 sections. The FoBMO axis (yellow line) was defined as the line connecting the BMO center and fovea and used as the reference line. The locations of maximum deep ONH parameters were measured as angular location from the FoBMO axis. When the directions were located above the FoBMO axis, they were assigned negative values (green arrow). (B) Three deep ONH structures were measured: EOBT length, defined as the length between the end points of EOBT tissue (red line); ONH tilt angle, defined as the angle between the BMO (white line) andoc (white dotted line) planes; BMO plane was the line connecting both BMOs and the OC plane was defined as the line connecting the nasal BMO and innermost margin of the EOBT; and OC obliqueness, defined as the angle made by a vertical line (red dotted line) and EOBT (red line). (C) The angular location of each maximum deep ONH parameter was located below the FoBMO axis, and the location was assigned a positive value. (D) DeepONH parameters located at the inferotemporal area exhibited greater extents compared to the parameters located at the superotemporal area. confidence intervals (CIs) were calculated. As we described in the previous study, there were correlations among maximum values of EOBT length, ONH tilt angle, and OC obliqueness. We used three multivariate analysis models to divide these parameters in different models. In addition, logistic analysis also was performed in OAG with nonmyopic eyes (AL < 24.5 mm). In the preliminary analysis of included OAG with phakic eyes, there were no emmetropic eyes with AL > 24.5 mm; thus, AL 24.5 mm was regarded as the cutoff value to define myopia. All statistical analyses were performed with SPSS software version 18.0 (SPSS, Inc., Chicago, IL, USA). A probability value of P < 0.05 was considered statistically significant. RESULTS Our study included 278 OAG eyes from 278 patients. Among these OAG eyes, 54 (19.4 %) were excluded because they had poor EDI OCT images (24), unreliable VF tests (10), advanced glaucoma with MD < 12 db (17), or myopia with AL > 28 mm (3). Only 224 eyes were included in the analysis. Intraand interobserver ICCs showed good agreements for assessments of the extent of each parameter (Table 1). When we divided the included OAG eyes into two groups based on the presence of EOBT, 169 OAG eyes demonstrated EOBT structure in at least at one section, while 55 did not show EOBT structure at any section. Parameters, such as age, sex, IOP, and central corneal thickness (CCT) did not show significant differences between the two groups, but the OAG eyes with EOBT had longer AL compared to those without EOBT (P < 0.001). Among the VF global indexes, such as MD, PSD, and VFI, there were no differences between the two groups. However, superior dominant VF defects were found more frequently in the OAG with than without EOBT (P ¼ 0.014; Table 2). TABLE 1. Intraobserver and Interobserver Reproducibility in Measurement of Parameters Intraobserver ICC (95% CI) Interobserver ICC (95% CI) Extent Maximum EOBT length, lm ( ) ( ) Maximum ONH tilt angle, deg ( ) ( ) Maximum OC obliqueness, deg ( ) ( ) Angular location Maximum EOBT location, deg ( ) ( ) Maximum ONH tilt location, deg ( ) ( ) Maximum OC obliqueness location, deg ( ) ( )

4 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3845 TABLE 2. Baseline Characteristics of the Included Open Angle Glaucoma Eyes EOBT (þ), N ¼ 169 EOBT ( ), N ¼ 55 P Value Age, y (21 78) (22 74) 0.247* Male sex, n (%) 115 (68.0) 30 (54.5) IOP, mm Hg (9 28) (11 27) 0.796* AL, mm ( ) ( ) <0.001* CCT, lm ( ) ( ) 0.079* VF characteristics MD, db ( 11.9 to 0.7) ( ) 0.336* PSD, db ( ) ( ) 0.730* VFI, % (56 99) (69 99) 0.475* Superior hemifield defect dominant, n (%) 114 (67.5) 27 (49.1) Paracentral involvement, n (%) 59 (34.9) 13 (23.6) Values are mean 6 SD (range). P < 0.05 are shown in bold. * Independent t-test or Mann-Whitney U test. v2 test. Locational Correspondence Between Deep ONH Structures and VF Defect in OAG We selected 169 OAG eyes with EOBT. When we divided the OAG eyes into two groups according to the dominant VF defect location, 114 (67.5%) were included in the superior and 55 (32.5%) in the inferior defect dominant groups. The parameters, such as age, sex, IOP, AL, CCT, and MD, did not show significant differences between two groups. The maximum deep ONH parameters, such as EOBT length, ONH tilt angle, and OC obliqueness, did not demonstrate significant differences in extent. However, in the superior dominant VF defect group, the maximum deep ONH parameters were significantly located inferiorly compared to the inferior VF defect dominant group (P < 0.001, respectively). When the angular locations were divided into binary values, such as positive and negative directions based on FoBMO axis, the relationship between the locations of maximum deep ONH parameters and dominant VF defect was statistically significant (P < 0.001, respectively; Table 3). In the univariate and multivariate logistic regression analysis for factors associated with the superior hemifield defect dominant group, the maximum values of EOBT length, ONH tilt angle, and OC obliqueness were located relatively inferiorly with statistical significance (P < in all; Table 4). When we selected only OAG eyes with AL < 24.5 mm, the superior VF defect dominant group also represented significantly inferiorly located maximum deep ONH parameters compared to the inferior VF defect dominant group (Table 5). Factors Associated With Paracentral Scotoma When we divided the OAG eyes into two groups according to the presence of paracentral scotoma, 57 eyes (33.7%) had paracentral scotoma. Factors, such as age, sex, IOP, AL, and CCT were not different between the two groups. MD values were significantly worse in OAG eyes with than without paracentral scotoma (P ¼ 0.003). Deep ONH deformation, such as maximum EOBT length, ONH tilt angle, and OC obliqueness, showed relatively inferiorly location in eyes with than TABLE 3. Comparisons of Parameters According to Dominant VF Involvement in OAG With EOBT Superior Dominant VF Defect, n ¼ 114 Inferior Dominant VF Defect, n ¼ 55 P Value Age, y (21 to 74) (21 to 78) 0.615* Male sex, n (%) 75 (65.8) 40 (72.7) IOP, mm Hg (9 to 24) (11 to 28) 0.980* AL, mm (22.3 to 27.9) (22.9 to 27.7) 0.424* CCT, lm (457 to 642) (400 to 609) 0.205* MD, db ( 11.9 to 0.1) ( 11.6 to 0.7) 0.097* Deep ONH parameters Extent Maximum EOBT length, lm (139 to 1269) (154 to 1242) 0.894* Maximum ONH tilt angle, deg (4.0 to 29.0) (4.6 to 43.3) 0.346* Maximum OC obliqueness, deg (12.6 to 78.8) (9.7 to 81.2) 0.093* Angular location Maximum EOBT location, deg ( 56.3 to 86.3) ( 63.8 to 82.1) <0.001* Maximum ONH tilt angle location, deg ( 48.2 to 85.9) ( 78.5 to 82.7) <0.001* Maximum OC obliqueness location, deg ( 45.1 to 86.2) ( 78.8 to 86.3) <0.001* Inferior dominant location Maximum EOBT location, n (%) 90 (78.9) 29 (51.0) <0.001 Maximum ONH tilt angle location, n (%) 88 (77.2) 27 (49.1) <0.001 Maximum OC obliqueness location, n (%) 94 (82.4) 31 (56.4) <0.001 Values are mean 6 SD (range). P < 0.05 are shown in bold. * Independent t-test or Mann-Whitney U test. v2 test.

5 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3846 TABLE 4. Logistic Regression Analysis: Factors Associated With Superior Hemifield Dominant Defect Univariate Analysis Multivariate Analysis 1 Multivariate Analysis 2 Multivariate Analysis 3 OR 95% CI P OR 95% CI P OR 95% CI P OR 95% CI P Age, y Male sex AL, mm IOP, mm Hg CCT, lm MD, db Maximal EOBT length, lm Maximal EOBT < <0.001 length location, deg Maximal ONH tilt angle, deg Maximal ONH tilt < <0.001 location, deg Maximal OC obliqueness, deg Maximal OC obliqueness location, deg < <0.001 P < 0.05 are shown in bold. without paracentral scotoma (P ¼ 0.019; P ¼ 0.016; P ¼ 0.150; Table 6). In the univariate logistic regression for factors associated with paracentral scotoma, worse MD was significantly associated with paracentral scotoma (P ¼ 0.002). Among the maximum deep ONH parameters, relatively inferior locations of maximum EOBT and ONH tilt angle were significantly associated with paracentral scotoma in univariate and multivariate analysis (P ¼ 0.021, P ¼ in maximum EOBT length; P ¼ 0.018, P ¼ in maximum ONH tilt angle). Maximum OC obliqueness demonstrated relatively inferior location in paracentral scotoma, but it did not reach statistical significance in univariate and multivariate analysis (P ¼ 0.150, P ¼ 0.186; Table 7). DISCUSSION Our study showed the relationship between VF defect pattern and deep ONH morphology, such as EOBT length, ONH tilt angle, and OC obliqueness in OAG eyes. First, OAG eyes with EOBT had longer AL and greater proportion of superior dominant VF defects than those without EOBT. In OAG eyes without EOBT, no significant difference was found between TABLE 5. Comparisons of Parameters According to Dominant VF Involvement in OAG With Axial Length < 24.5 mm Superior Dominant VF Defect, N ¼ 31 Inferior Dominant VF Defect, N ¼ 10 P Value Age, y * Male sex, n (%) 19 (61.3) 8 (80.0) IOP, mm Hg * AL, mm * CCT, lm * MD, db * Deep ONH deformation Extent Maximum EOBT length, lm * Maximum ONH tilt angle, deg * Maximum OC obliqueness, deg * Angular location Maximum EOBT location, deg * Maximum ONH tilt angle location, deg * Maximum OC obliqueness location, deg * Inferior dominant location Maximum EOBT location, n (%) 26 (83.9) 5 (50.0) Maximum ONH tilt angle location, n (%) 28 (90.3) 5 (50.0) Maximum OC obliqueness location, n (%) 28 (90.3) 6 (60.0) Values are mean 6 SD. P < 0.05 are shown in bold. * Independent t-test or Mann-Whitney U test. v2 test.

6 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3847 TABLE 6. Comparisons of Parameters According to Presence of Paracentral Scotoma in OAG With EOBT Paracentral Scotoma ( ), N ¼ 112 Paracentral Scotoma (þ), N ¼ 57 P Value Age, y (21 to 73) (22 to 78) 0.423* Male sex, n (%) 77 (68.8) 38 (66.7) IOP (mm Hg) (9 to 28) (10 to 28) 0.263* AL, mm (22.7 to 27.9) (22.3 to 27.8) 0.876* CCT, lm (400 to 642) (487 to 605) 0.725* MD, db ( 11.5 to 0.7) ( 11.9 to 0.1) 0.002* Deep ONH deformation Extent Maximum EOBT length, lm (139 to 1217) (179 to 1269) 0.501* Maximum ONH tilt angle, deg (4.3 to 29.0) (4.0 to 43.3) 0.862* Maximum OC obliqueness, deg (9.7 to 78.8) (18.6 to 81.2) 0.958* Angular location Maximum EOBT length location, deg ( 63.8 to 86.3) ( 48.8 to 86.3) 0.019* Maximum ONH tilt angle location, deg ( 78.5 to 85.9) ( 45.0 to 83.2) 0.016* Maximum OC obliqueness location, deg ( 78.8 to 86.3) ( 45.0 to 86.3) 0.150* Values are mean 6 SD. P < 0.05 are shown in bold. * Independent t-test or Mann-Whitney U test. v2 test. the proportions of superior and inferior dominant VF defects. Second, there was correspondence between the locations of deep ONH parameters and dominant VF defects, and the relationship also was significant when only OAG eyes with AL < 24.5 mm were included (Fig. 2). Third, the location of deep ONH morphology was associated with the presence of paracentral scotoma in OAG eyes. In OAG eyes with paracentral scotoma, the maximum deep ONH parameters were distributed relatively inferiorly from the FoBMO axis, whereas, in OAG without paracentral scotoma, the maximum values were on average close to the foveal direction (Fig. 3). In myopic OAG eyes, correspondence between the locations of RNFL defects and ONH morphology, such as disc tilt and PPA, has been well known. 7 9,16 Tensile stress around ONH was suggested as the possible main mechanism to induce glaucomatous defect in myopic eyes and, thus, it was suggested that myopic OAG should be differentiated from nonmyopic OAG in that they may have different mechanisms to induce glaucomatous defect and also demonstrate different clinical features, such as progression rate and pattern However, in ONH without definite PPA or disc tilt, few studies reported the locational correspondence between ONH morphology and VF defect because it is difficult to evaluate regional differences based on the angular locations of ONH by observing ONH surface. OCT enabled the deep ONH structures to be observed. As a result, small morphologic differences in border tissue and BMO at each angular location could be measured. 19 The previous study reported that the externally oblique type border tissue was more common especially at inferotemporal areas in TABLE 7. Logistic Regression Analysis: Parameters Associated With the Paracentral Scotoma in OAG With EOBT Univariate Analysis Multivariate Analysis 1 Multivariate Analysis 2 Multivariate Analysis 3 OR 95% CI P OR 95% CI P OR 95% CI P OR 95% CI P Age, y Male sex AL, mm IOP, mm Hg CCT, lm MD, db Maximal EOBT length, lm Maximal EOBT length location, deg Maximal ONH tilt angle, deg Maximal ONH tilt location, deg Maximal OC obliqueness, deg Maximal OC obliqueness location, deg P < 0.05 are shown in bold.

7 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3848 FIGURE 2. Images of representative cases of locational correspondence between deep ONH parameters and VF defect in OAG. (A) The right eye of a 65-year-old female with OAG had an AL of mm. BMOs (white arrow head) and inner end point of border tissues (white arrow) located differently according to the angular locations. The B-scans at the inferotemporal area (þ458; green arrow) showed greater EOBT length compared to the opposite direction ( 458; green dot arrow) and the FoBMO axis direction (yellow line). The maximum deep ONH parameters located around the inferotemporal ONH area. ONH also showed notching and RNFL defect at the inferotemporal area and a corresponding VF defect was found in the superior hemifield. (B) The left OAG eye of a 60-year-old female with an AL of mm. In the B-scan images, maximum deep ONH parameters were greater at the superotemporal location (blue dotted arrow) than the FoBMO axis (yellow line) and inferotemporal location (blue arrow). Fundus photo showed a round type ONH with rim notching at the superotemporal area and a corresponding VF defect was found in the inferior hemifield. glaucoma eyes compared to normal controls. 4 In our previous study, we demonstrated that the maximum EOBT length was associated with AL. In addition, the dominant locations of EOBT were the inferotemporal area of ONH and were consistent with RNFL defect location in myopic NTG. 11 In our study, we demonstrated correspondence between the locations of dominant VF defects and maximum deep ONH parameters in OAG eyes, and the locational correspondence also was shown when only OAG eyes with AL 24.5 mm were included. The locational correspondence between glaucomatous VF defects and ONH morphology seems not just the specific feature of glaucomatous defect in myopic eyes, but it is

8 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3849 FIGURE 3. Images of representative cases of OAG with and without paracentral scotoma according to deep ONH parameter. (A) The right OAG eye of a 50-year-old male with an AL of mm. Fundus photo demonstrates a round ONH with superior notching and thinning and inferior VF defect without paracentral scotoma. BMOs (white arrow head) and inner end point of border tissues located differently according to the different angular locations. Deep ONH parameters, such as EOBT length, ONH tilt angle, and optic canal obliqueness had greatest values at FoBMO axis direction (yellow line) compared to the superotemporal (green dotted arrow) and inferotemporal (green arrow) directions. (B) The left OAG eye of a 61-yearold female with an AL of mm. Fundus photo exhibits a round ONH with rim notching at the inferotemporal area with corresponding VF defect with paracentral scotoma. The border tissue morphology was externally oblique at the inferotemporal direction (blue arrow), but internally oblique at the superotemporal direction (blue dotted arrow) and FoBMO axis direction (yellow line). The maximum values of deep ONH parameters existed near the inferotemporal direction. likely to be the features in the OAG eyes with most ranges of AL. IOP, the most important factor in glaucoma, generates load on ONH tissues at all levels of IOP that is presumed to damage retinal ganglion cell axons. 1,2 When ONH is exposed to IOP that is difficult to endure, the ONH structural changes, such as deeper prelaminar and LC depth or larger cupping, can occur. 20,21 However, the ONH changes, such as large cupping or IOP itself, hardly explain why glaucomatous thinning or notching usually appears at inferotemporal ONH areas at first in most types of glaucoma. We hypothesized that IOP-related stress around the ONH may be different according to ONH geometric characteristics, and it can offer different vulnerability around the ONH. All eyes, even in short AL eyes, would have experienced axial elongation during the development period. Deep ONH structural deformation, such as elongated

9 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j 3850 EOBT length, enlarged ONH tilt angle, and OC obliqueness, might be formed differently according to the location around the ONH with axial growth. 22 Different structural deformation might offer different stress and strain on collagen tissues of the LC and parapapillary sclera under given IOP, and could trigger remodeling of the extracellular matrix Our hypothesis seems to be supported by our results. When the eyes were divided into two groups based on the presence of EOBT, the OAG eyes without EOBT showed no significant difference in the proportion between superior and inferior VF defect dominant cases (51.7% vs. 48.3%). Considering OAG eyes with EOBTs were associated with superior dominant VF defects, it is likely that, if the eyes show no EOBT, then the superior dominant VF defect pattern may be weak. In our study, paracentral scotoma in OAG eyes was associated with worse MD and relatively inferior locations of maximum deep ONH parameters. Considering that the chance to involve the central VF increase with worse MD, it is natural to suggest worse MD as the risk factor of paracentral scotoma in OAG eyes. The OAG eyes with paracentral scotoma also were associated with relatively inferior location of deep ONH deformation compared to OAG eyes without paracentral scotoma (þ22.48 vs. þ10.88). We speculated that the results may be associated with vulnerable macula regions that exists at the inferior macula zone. Hood et al. 28 showed that there is a vulnerable macula area at which glaucomatous defects can occur in early stage glaucoma. This region may be vulnerable because the region could be exposed to maximum stress related to maximum deep ONH deformation, and simultaneously have relatively sparse RGC density compared to macula center. Previously, it was known that approximately 50% of RGC reside within 168 of the fovea. 29 Thus, in the macula region, VF defect may not be detected until approximately 35% of RGCs are damaged. 30 Interestingly, our study exhibited that the average angular locations of maximum deep ONH parameters were just outside the area with redundant RGC. Maximum deep ONH deformation on the relatively vulnerable region might induce paracentral scotoma in OAG eyes. Previously, several reports indicated that myopic eyes had relations with paracentral scotoma in OAG eyes. 15,31,32 Kimura et al. 32 suggested that OAG with high myopia can induce parapapillary bundle defect. Sawada et al. 15 also showed that disc tilt angle was associated with the presence of paracentral scotoma in myopic OAG and LC defects at the temporal location of ONH were suggested as the reason for the relationship between disc tilt angle and paracentral scotoma. The results of the previous studies were different from those of our study in that our study showed that the locations of the maximum deep ONH parameters were more important for paracentral scotoma than the extents of the parameters. It is difficult to compare our study results with previous studies because our study also included nonmyopic OAGs in the analysis. However, when the paracentral scotoma group was divided into two groups using K-clustering analysis, our study also demonstrated one group with a long AL and greater maximum parameters, though the number of cases was small (data not shown). We postulated that OAG with high myopia may induce paracentral scotoma even with deep ONH deformation located at the direction close to the fovea. If the great value of myopic ONH deformation was concentrated on the macula, it could induce macula RGC damage near to the macula and parapapillary bundle area at a fast rate. Our study has several limitations. First, relatively small OAG cases were included in the analysis. Therefore, in the analysis for factors associated with paracentral scotoma, maximum OC obliqueness location did not reach statistical significance. However, considering that the deep ONH parameters were correlated with each other, 13 the clinical implications of the parameters appear to be similar though the statistical significance was not shown. Second, in some cases, measurement of the deep ONH parameters could be difficult because EOBTs had a curved shape or small size. To minimize measurement errors, we regarded the EOBT length as the shortest distance between both end points of EOBT tissue and checked ICC values. Third, the definitions of dominant VF defect location and paracentral scotoma were determined arbitrarily. To minimize selection bias, the definition of dominant VF defect location was simply determined by the difference between superior and inferior mean deviation values. To reduce false-positive rates, paracentral scotoma was confirmed only when it existed in the dominant VF defect location. The arbitrary definitions of dominant VF defect location and paracentral scotoma could have affected the results. Fourth, other risk factors, such as low blood pressure and ocular perfusion pressure, were not considered in our study. In the previous study, systemic factors, such as systemic hypotension and migraine, suggested to be potential risk factor of paracentral scotoma in glaucoma. 14 Because glaucoma is the multifactorial disease, other factors might have influenced the results. In summary, our study showed that the locations of maximum deep ONH parameters, such as EOBT length, ONH tilt angle, and OC obliqueness, seem to be associated with the characteristics of VF defect in OAG. Our study results may help us to understand the pathogenesis of glaucoma, especially the reason why different regional vulnerability exists around ONH in glaucoma. Acknowledgments Supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2017R1D1A1B ). Disclosure: J.C. Han, None; J.H. Choi, None; D.Y. Park, None; E.J. Lee, None; C. Kee, None References 1. Burgoyne CF, Downs JC, Bellezza AJ, Suh JK, Hart RT. The optic nerve head as a biomechanical structure: a new paradigm for understanding the role of IOP-related stress and strain in the pathophysiology of glaucomatous optic nerve head damage. Prog Retin Eye Res. 2005;24: Bellezza AJ, Hart RT, Burgoyne CF. The optic nerve head as a biomechanical structure: initial finite element modeling. Invest Ophthalmol Vis Sci. 2000;41: Burgoyne CF. A biomechanical paradigm for axonal insult within the optic nerve head in aging and glaucoma. Exp Eye Res. 2011;93: Reis AS, Sharpe GP, Yang H, Nicolela MT, Burgoyne CF, Chauhan BC. Optic disc margin anatomy in patients with glaucoma and normal controls with spectral domain optical coherence tomography. Ophthalmology. 2012;119: Jonas JB, Xu L. Histological changes of high axial myopia. Eye (Lond). 2014;28: Kim YW, Lee EJ, Kim TW, Kim M, Kim H. Microstructure of beta-zone parapapillary atrophy and rate of retinal nerve fiber layer thinning in primary open-angle glaucoma. Ophthalmology. 2014;121: Park HY, Lee K, Park CK. Optic disc torsion direction predicts the location of glaucomatous damage in normal-tension glaucoma patients with myopia. Ophthalmology. 2012;119:

10 Deep ONH Morphology and VF Defect in OAG IOVS j August 2018 j Vol. 59 j No. 10 j Cho BJ, Park KH. Topographic correlation between beta-zone parapapillary atrophy and retinal nerve fiber layer defect. Ophthalmology. 2013;120: Teng CC, De Moraes CG, Prata TS, Tello C, Ritch R, Liebmann JM. Beta-Zone parapapillary atrophy and the velocity of glaucoma progression. Ophthalmology. 2010;117: Choi JA, Park HY, Shin HY, Park CK. Optic disc tilt direction determines the location of initial glaucomatous damage. Invest Ophthalmol Vis Sci. 2014;55: Reis AS, O Leary N, Yang H, et al. Influence of clinically invisible, but optical coherence tomography detected, optic disc margin anatomy on neuroretinal rim evaluation. Invest Ophthalmol Vis Sci. 2012;53: Chauhan BC, O Leary N, Almobarak FA, et al. Enhanced detection of open-angle glaucoma with an anatomically accurate optical coherence tomography-derived neuroretinal rim parameter. Ophthalmology. 2013;120: Han JC, Lee EJ, Kim SB, Kee C. The characteristics of deep optic nerve head morphology in myopic normal tension glaucoma. Invest Ophthalmol Vis Sci. 2017;58: Park SC, De Moraes CG, Teng CC, Tello C, Liebmann JM, Ritch R. Initial parafoveal versus peripheral scotomas in glaucoma: risk factors and visual field characteristics. Ophthalmology. 2011;118: Sawada Y, Araie M, Ishikawa M, Yoshitomi T. Multiple temporal lamina cribrosa defects in myopic eyes with glaucoma and their association with visual field defects. Ophthalmology. 2017;124: Sung MS, Kang YS, Heo H, Park SW. Optic disc rotation as a clue for predicting visual field progression in myopic normaltension glaucoma. Ophthalmology. 2016;123: Han JC, Lee EJ, Kim SH, Kee C. Visual field progression pattern associated with optic disc tilt morphology in myopic open-angle glaucoma. Am J Ophthalmol. 2016;169: Doshi A, Kreidl KO, Lombardi L, Sakamoto DK, Singh K. Nonprogressive glaucomatous cupping and visual field abnormalities in young Chinese males. Ophthalmology. 2007;114: Strouthidis NG, Grimm J, Williams GA, Cull GA, Wilson DJ, Burgoyne CF. A comparison of optic nerve head morphology viewed by spectral domain optical coherence tomography and by serial histology. Invest Ophthalmol Vis Sci. 2010;51: Yang H, Downs JC, Girkin C, et al. 3-D histomorphometry of the normal and early glaucomatous monkey optic nerve head: lamina cribrosa and peripapillary scleral position and thickness. Invest Ophthalmol Vis Sci. 2007;48: Kim YW, Kim DW, Jeoung JW, Kim DM, Park KH. Peripheral lamina cribrosa depth in primary open-angle glaucoma: a swept-source optical coherence tomography study of lamina cribrosa. Eye (Lond). 2015;29: Kim TW, Kim M, Weinreb RN, Woo SJ, Park KH, Hwang JM. Optic disc change with incipient myopia of childhood. Ophthalmology. 2012;119:21 26.e Norman RE, Flanagan JG, Rausch SM, et al. Dimensions of the human sclera: thickness measurement and regional changes with axial length. Exp Eye Res. 2010;90: Yamaoka A, Matsuo T, Shiraga F, Ohtsuki H. TIMP-1 production by human scleral fibroblast decreases in response to cyclic mechanical stretching. Ophthalmic Res. 2001;33: Cui W, Bryant MR, Sweet PM, McDonnell PJ. Changes in gene expression in response to mechanical strain in human scleral fibroblasts. Exp Eye Res. 2004;78: Shelton L, Rada JS. Effects of cyclic mechanical stretch on extracellular matrix synthesis by human scleral fibroblasts. Exp Eye Res. 2007;84: Girard MJ, Suh JK, Bottlang M, Burgoyne CF, Downs JC. Biomechanical changes in the sclera of monkey eyes exposed to chronic IOP elevations. Invest Ophthalmol Vis Sci. 2011; 52: Hood DC. Improving our understanding, and detection, of glaucomatous damage: an approach based upon optical coherence tomography (OCT). Prog Retin Eye Res. 2017;57: Curcio CA, Allen KA. Topography of ganglion cells in human retina. J Comp Neurol. 1990;300: Kerrigan-Baumrind LA, Quigley HA, Pease ME, Kerrigan DF, Mitchell RS. Number of ganglion cells in glaucoma eyes compared with threshold visual field tests in the same persons. Invest Ophthalmol Vis Sci. 2000;41: Mayama C, Suzuki Y, Araie M, et al. Myopia and advancedstage open-angle glaucoma. Ophthalmology. 2002;109: Kimura Y, Hangai M, Morooka S, et al. Retinal nerve fiber layer defects in highly myopic eyes with early glaucoma. Invest Ophthalmol Vis Sci. 2012;53: