ORIGINAL STUDY
Correlations Between Retinal Nerve Fiber Layer Thickness and Axial Length, Peripapillary Retinal Tilt, Optic Disc Size, and Retinal Artery Position in Healthy Eyes Takehiro Yamashita, MD, PhD, Taiji Sakamoto, MD, PhD, Naoya Yoshihara, MD, Hiroto Terasaki, MD, PhD, Minoru Tanaka, MD, Yuya Kii, MD, PhD, and Kumiko Nakao, MD, PhD
Purpose: To determine the correlations between the retinal nerve fiber layer thickness (RNFLT) and the axial length, peripapillary retinal tilt (PRT), and optic disc size, and retinal artery position. Methods: A prospective, observational cross-sectional study of 119 healthy right eyes of 119 volunteers. All participants underwent comprehensive ophthalmologic examinations including peripapillary RNFLT imaging and measurements of the axial length. The RNFLT was determined by the TOPCON 3D OCT-1000, MARK II. The RNFLT in a 3.4 mm circular scan was divided into 12 clock-hour sectors and 4 quadrant sectors around the optic disc. The PRT was assessed using the RNFLT B-scan images. The angle between the supra-temporal and infra-temporal retinal arteries was determined in the color fundus photographs. The correlations between the sectorial RNFLTs and the axial length, PRT, optic disc size, and artery angles were determined by simple and multiple regression analyses. Results: Multiple regression analyses showed that the nasal and inferior quadrants and the whole RNFLT were significantly and negatively correlated with the axial length (standardized coefficient (SC) = 0.39 to 0.30, P < 0.05). The PRT was significantly and positively associated with all of the quadrants and the whole RNFLT (SC = 0.22 to 0.45, P < 0.05). The retinal artery angle was significantly and negatively associated with the temporal RNFLT and positively associated with inferior RNFLT (SC = 0.49 to 0.31, P < 0.05). The optic disc size was significantly and positively associated with the superior and nasal quadrants, and the whole RNFLT (SC = 0.20 to 0.27, P < 0.05). Conclusions: The axial length, PRT, optic disc size, retinal artery angle can affect the peripapillary RNFLT. These variables should be considered when assessing the peripapillary RNFLT. Key Words: retinal nerve fiber layer thickness, axial length, peripapillary retinal tilt, retinal artery position, optic disc size
(J Glaucoma 2017;26:34–40)
Received for publication November 26, 2015; accepted August 22, 2016. From the Department of Ophthalmology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan. Disclosure: The authors declare no conflict of interest. Reprints: Taiji Sakamoto, MD, PhD, Department of Ophthalmology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, 890-8520, Japan (e-mail: tsakamot@m3. kufm.kagoshima-u.ac.jp). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.glaucomajournal.com. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/IJG.0000000000000550
I
t is widely accepted that myopia is a risk factor for primary open angle glaucoma.1–3 Considering the increasing prevalence of myopia worldwide, a correct diagnosis of glaucoma in myopic patients is becoming more important and necessary in ophthalmology.1,4,5 To detect the early stages of glaucoma in myopic eyes, it is necessary to assess the changes in the optic disc and alterations of the retinal nerve fiber layer thickness (RNFLT) accurately. However, it is difficult to make these assessments in myopic eyes because the RNFLT can be altered by the myopic changes in eyes without glaucoma. In addition, the effect of myopia without glaucoma on the peripapillary RNFLT has not been definitively determined. Optical coherence tomography (OCT) is a noninvasive imaging technique that can obtain images which can be used to evaluate the morphology and thickness of the RNFL and optic disc with micrometer resolution.6,7 Several studies have documented that reliable measurements of the RNFLT can be made from the OCT images.8–13 However, the RNFLT measurements can be significantly affected by the refractive errors, and thus the ability to detect glaucoma in highly myopic eyes by the RNFLT is poorer than that in emmetropic eyes.14 It is well recognized that the peripapillary RNFLT is affected by the axial length,15–18 the peak position of the RNFLT,15,18,19 position of the retinal vessel,18,20,21 and the tilt and size of the optic disc.22–24 In an earlier study, we reported that a mathematical method can be used to determine the angle of the retinal vessels, and we found that both the axial length and the peak artery angle influenced the sectoral RNFLT.18 We also developed another mathematical technique, a sine-curve analysis technique, to measure the degree of the peripapillary retinal tilt (PRT).25 Because these techniques allowed us to express each variable as a numerical value, it was possible to analyze the interrelationships among these factors quantitatively. If these relationships are valid, then the values may be suitable for determining how they are associated with the RFNLT. Thus, the purpose of this study was to determine the relationships between the sectoral and whole RNFLT and the peak angle, artery angle, PRT, optic disc size, and the axial length of healthy eyes of normal individuals. Because these variables were correlated to each other, we also determined which factors affected the sectoral and whole RNFLT using multiple regression analyses.
METHODS All of the procedures used conformed to the tenets of the Declaration of Helsinki. A written informed consent
34 | www.glaucomajournal.com J Glaucoma Volume 26, Number 1, January 2017 Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Glaucoma
Volume 26, Number 1, January 2017
was obtained from all of the subjects after an explanation of the procedures to be used. The study was approved by the Ethics Committee of Kagoshima University Hospital, and it was registered with the University Hospital Medical Network (UMIN)-clinical trials registry. The registration title was, “Morphological analysis of the optic disc and the retinal nerve fiber in myopic eyes” and the registration number was UMIN000006040. A detailed protocol is available at: https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr. cgi?function = brows&action = brows&type = summary& recptno = R000007154&language = J.
Subjects This was a cross-sectional, prospective, observational study. We initially examined 133 eyes of 133 volunteers who were enrolled between November 1, 2010 and February 29, 2012. The volunteers had no known eye diseases as determined by examining their medical charts, and the data from only the right eyes were analyzed. The eligibility criteria were: age 20 years or above but 40 years or below; eyes normal by slit-lamp biomicroscopy, ophthalmoscopy, and OCT; best-corrected visual acuity of r0.1 logarithm of the minimum angle of resolution (logMAR) units; and intraocular pressure (IOP) r21 mmHg. The exclusion criteria were eyes with known ocular diseases such as glaucoma, presence of a staphyloma, and optic disc anomalies; presence of visual field defects; and prior refractive or intraocular surgery. The presence of a staphyloma was determined by B-mode echo. None of the eyes was excluded because of poor OCT image quality caused by poor fixation.
Measurement of Axial Length and Refractive Error All eyes had a standard ocular examination including; slit-lamp biomicroscopy of the anterior segment, ophthalmoscopy of the ocular fundus, pneumo-tonometric (CT-80; Topcon, Tokyo, Japan) measurements of the IOP, and AL-2000 ultrasonographic (Tomey, Japan) measurements of the axial length. The refractive error (spherical equivalent) was measured with the Topcon KR8800 autorefractometer/keratometer.
Determination of Thickness of RNFL, Peak Angle, Artery Angle, PRT, and Optic Disc Size All eyes were examined by a single examiner (T.Y.). The RNFLT was measured with the TOPCON 3D OCT1000 MARK II using the RNFL 3.4 mm circle scan. In this
FIGURE 1. The retinal nerve fiber layer thickness of the whole, 4 quadrants, and the 12 clock hours were averaged. I indicates inferior; N, nasal; S, superior; T, temporal; W, whole.
Copyright
r
Association of PRT and RNFLT in Healthy Eyes
protocol, 1024 A-scans/circle, 16 overlapping B-scans/ image, and direct B-scan observations were made. The OCT images and the color fundus photographs were recorded at the same time. The optical system of the OCT instrument detected the edge of the optic disc in the fundus image, and the scan circle was centered automatically on the optic disc just before the OCT image was recorded. To exclude the effects of errors in the scan circle centration, 1 examiner (Y.K.) checked offline that the center of the scan circle was located at the center of the optic disc. The average RNFLT of the whole, each of the 4 quadrants, and each of the clock-hour sectors were measured (Fig. 1). The temporal-superior-nasal-inferiortemporal thickness curves were used to measure the angle between supra-temporal and infra-temporal peaks of the RNFL. We determined this peak angle in the temporalsuperior-nasal-inferior-temporal thickness profile of the RNFLT analyses. The distance between the peak RNFLTs for the supra-temporal and infra-temporal RNFL peaks was determined by dragging a vertical line in the profile graph in the Photoshop CS5 program. Then, the distance between supra-temporal and infra-temporal RNFL peaks (Supplementary Figs. 1A, X, Supplemental Digital Content 1, http://links.lww.com/IJG/A92) was converted to an angular value by dividing it by the entire distance (Supplementary Figs. 1A, Y, Supplemental Digital Content 1, http://links.lww.com/IJG/A92), then multiplied by 360.15,18 Color fundus photographs and OCT images were taken at the same time with the TOPCON 3D OCT-1000 MARK II. The scan circle was placed over the optic disc in the photographs as a green circle. The points where the green circle intersected the supra-temporal and infratemporal major retinal arteries were determined, and the angle between supra-temporal and infra-temporal major retinal artery was measured (Supplementary Fig. 1B, Supplemental Digital Content 1, http://links.lww.com/IJG/ A92). We named this angle the artery angle.18 The course of the retinal pigment epithelium was marked on the B-scan images manually. The pixel coordinates of each mark were determined automatically using the Image J program [Image J version 1.47, National Institutes of Health, Bethesda, MD; http://imagej.nih.gov/ ij/ (in the public domain)]. The “x” and “y” coordinates of each pixel were converted to a new set of “x” and “y” coordinates with the center of the curve. Finally, the converted data were fit to a sine-curve equation (y = a sin(b x-c)) with the curve fitting program of Image J. The “a,” “b,” and “c” are constants calculated by the least squares method of the curve fitting program of Image J. The constant “a” is the amplitude of the sine curve, and a larger amplitude “a” will make the amplitude of the curve larger and make the arms of the PRT larger. Because the eyes were relatively stationary throughout the measurements, the retinal plane of macular area was fixed at an orientation approximately perpendicular to the reference light axis. Then, the PRT examined was represented by the angle between the optic axis and the PRT. The amplitude of the sine curve, “a,” was considered to reflect the degree of the PRT relative to the optical axis.25 The details of the measurement of the peak angle, the artery angle, and amplitude of the sine curve were described in more detail in earlier publications.18,25 We measured the size of the optic disc in the color fundus photographs by Image J after adjusting for the magnification effect.18,26,27
2016 Wolters Kluwer Health, Inc. All rights reserved.
www.glaucomajournal.com |
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
35
Yamashita et al
J Glaucoma
TABLE 1. Participants Date Age (y) Sex (male/female) Spherical equivalent (D) Axial length (mm) Peak angle (deg.) Artery angle (deg.) Amplitude (pixels)
Mean ± SD
Range
25.9 ± 4.0 81/38 4.68 ± 3.36 25.4 ± 1.4 124.2 ± 20.5 128.8 ± 23.1 37.4 ± 17.2
22-40 14.25-0.50 22.4-30.4 74.4-193.9 77.9-193.9 8.8-80.8
Volume 26, Number 1, January 2017
The axial length was significantly and negatively correlated with the peak angle (r = 0.31, P = 0.001) and the artery angle (r = 0.41, P < 0.001). The peak angle was significantly correlated with the artery angle (r = 0.79, P < 0.001). Because peak angle and artery angle were highly and significantly correlated, they were analyzed separately to exclude the effects of multicollinearity. Specifically, the correlation of either peak angle or artery angle to the other factors was analyzed. The other factors included the RNFLT, axial length, optic disc size, and PRT. The average RNFLT of the whole, 4 quadrants, and the 12 clock-hour sectors are presented in Table 2.
Statistical Analyses All statistical analyses were performed with the SPSS statistics 19 for Windows (SPSS Inc., IBM, Somers, New York). The relationships between the whole or sectoral RNFLTs and the axial length, peak angle, artery angle, PRT, and optic disc size were determined by the Pearson correlation analyses. Multiple regression analyses were calculated for the whole and sectoral RNFLTs and the same factors because each of these factors were significantly correlated with the RNFLT.
RESULTS In total, 133 Japanese volunteers were examined. Seven eyes were excluded because of ocular diseases or prior ocular surgery, 3 eyes because of superior segmental optic hypoplasia, 1 case because of glaucoma, and 3 cases because of prior laser-assisted in situ keratomileusis. Seven other eyes were excluded because of difficulty in identifying the position of the peak of the RNFLT. In the end, the right eyes of 119 individuals (81 men and 38 women) were studied. The demographic information of the 119 subjects is presented in Table 1. The mean ± SD of the age was 25.9 ± 4.0 years, and the mean refractive error (spherical equivalent) was 4.68 ± 3.36 D. The mean axial length was 25.4 ± 1.4 mm. The mean peak angle was 124.2 ± 20.5 degrees, the mean artery angle was 128.8 ± 23.1 degrees, the mean PRT was 37.4 ± 17.2 pixels, and the mean optic disc size was 2006.9 ± 474.7 pixels.
TABLE 2. Retinal Nerve Fiber Layer Thickness Values
Locations
Sector
Mean ± SD
Range
Superior Nasal Nasal Nasal Inferior Inferior Inferior Temporal Temporal Temporal Superior Superior Superior Nasal Inferior Temporal Total
1 2 3 4 5 6 7 8 9 10 11 12 11, 12, 1 2, 3, 4 5, 6, 7 8, 9, 10 Whole
125.0 ± 18.5 97.6 ± 18.6 80.2 ± 14.7 80.2 ± 16.1 104.9 ± 19.8 128.7 ± 30.9 151.6 ± 20.3 101.7 ± 24.8 81.2 ± 15.5 114.3 ± 23.7 153.8 ± 22.2 130.5 ± 23.4 136.4 ± 15.9 85.9 ± 14.7 128.3 ± 19.0 99.0 ± 19.8 112.4 ± 10.6
71-183 51-144 44-122 45-125 56-159 66-218 104-205 65-191 54-130 76-184 95-215 86-206 97-182 53-121 80-169 67-165 89-136
The Pearson Correlation Coefficients and Standardized Coefficients (SCs) by Multiple Regression Analyses for Axial Length, PRT, Optic Disc Size, Peak Angle, Artery Angle, and RNFLT The Pearson correlation coefficients and the SCs of multiple regression for the peak angle, axial length, PRT, optic disc size, and the RNFLTs are shown in Fig. 2 and Supplementary Table 1 (Supplemental Digital Content 2, http://links.lww.com/IJG/A93). For both types of regression analyses, the whole RNFLT was significantly and negatively correlated with the axial length (r = 0.32 and 0.33, P < 0.05), significantly and positively correlated with the PRT (r = 0.32 and 0.41, P < 0.05), but not significantly correlated with the peak angle (r = 0.08 and 0.12, PZ0.05). The optic disc size was not significantly associated with the whole RNFLT in the Pearson correlation analysis (r = 0.11, PZ0.05), but significantly and positively associated with the whole RNFLT in the multiple regression analysis (r = 0.27, P < 0.05). In the Pearson correlation analysis, the temporal RNFLT was significantly and positively correlated with the axial length (r = 0.19 to 0.31, P < 0.05), and the nasal and inferior RNFLTs were significantly and negatively correlated with the axial length (r = 0.31 to 0.56, P < 0.05). Multiple regression analyses showed that the temporal to superior RNFLT was not significantly associated with the axial length (r = 0.06 to 0.15, PZ0.05), but the RNFLTs of the nasal and inferior quadrants were significantly associated with the axial length (r = 0.24 to 0.44, P < 0.05). The Pearson correlation analysis showed that the RNFLTs of the temporal and the 3 o’clock sectors were significantly and positively correlated with the PRT (r = 0.32 to 0.61, P < 0.05). Multiple regression analyses showed that the correlation between the temporal quadrant and 3 o’clock sector and the PRT were weakly but significantly correlated (r = 0.17 to 0.44, P < 0.05). The Pearson correlation analysis showed that the RNFLT of the temporal quadrant was significantly and negatively correlated with peak angle (r = 0.25 to 0.78, P < 0.05), and the RNFLT of the superior and inferior quadrants were significantly and positively correlated with the peak angle (r = 0.21 to 0.56, P < 0.05). A similar tendency was seen in the multiple regression analysis, but the coefficients of correlation were lower. The Pearson correlation analysis showed that the RNFLT of the temporal quadrant was significantly and negatively correlated with optic disc size (r = 0.24 to 0.24, P < 0.05), and the RNFLT of the other quadrants were significantly and positively correlated with the optic disc size (r = 0.21 to 0.33, P < 0.05). However, multiple regression analyses showed that the optic disc size was not significantly correlated with the temporal and
36 | www.glaucomajournal.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Glaucoma
Volume 26, Number 1, January 2017
Association of PRT and RNFLT in Healthy Eyes
FIGURE 2. The Pearson correlation coefficients and standardized coefficients of multiple regression for the axial length, peripapillary retinal tilt (PRT), optic disc size, peak angle, and retinal nerve fiber layer thickness (RNFLT). The significant correlated sectors are shown in gray.
inferior quadrants RNFLT (r = 0.05 to 0.10, PZ0.05) and significantly positively associated with whole, superior, and nasal quadrant RNFLT (r = 0.20 to 0.27, P < 0.05). We also investigated these relationships by changing the peak angle to the artery angle. The Pearson correlation coefficients and SCs of multiple regression correlation for the artery angle, axial length, PRT and optic disc size, and RNFLT are shown in Figure 3 and Supplementary Table 2 (Supplemental Digital Content 3, http://links.lww.com/ IJG/A94). A similar tendency was seen in the Pearson correlation coefficients and standardized correlation coefficients when the peak angle was changed to the artery angle.
DISCUSSION The multiple regression analyses showed that the axial length was significantly and negatively associated with the RNFLT of the nasal and inferior quadrants and also with the whole RNFLT. However, these correlations may be caused by the magnification effect or retinal thinning caused by an elongation of the axial length. Kim et al17 examined the association between the axial length and RNFLT in low, moderate, and high myopic eyes. They reported that the average global RNFL was significantly thinner in the high myopic eyes than in the low myopic eyes. For the quadrants, the RNFL was thicker in the low myopic eyes than in the moderate and the high myopic eyes in the superior, nasal, and inferior quadrants. The temporal quadrant was thinner in the low myopia group than in the moderate and high myopia groups. Their results are similar to our results from the Pearson correlation analysis, that is, the longer axial length eyes had a thinner average RNFL in the superior, nasal, and inferior quadrants, and a thicker temporal quadrant. However, multiple regression analysis showed that the relationship between the axial length and Copyright
r
the temporal sectoral RNFLT was not significant. The significant correlation may have been caused by the effect of the PRT, and the peak angle or the artery angle. In addition, we did not adjust the angle for the axial length or refractive error which might have led to differences in the magnification. The scan circle is projected larger than the actual circle after an elongation of the axial length caused by the magnification effect.26,27 This might affect the RNFLT, PRT, peak angle, and artery angle. Unfortunately, it may not be possible to obtain the absolute RNFLT values corrected for the axial length for our OCT instrument because the effect of distance is not known. Further studies using OCT with a magnification effect correction system are needed. The PRT was significantly and positively associated with the RNFLT at the 3 o’clock sector, the temporal quadrant, and the whole region. In addition, the RNFLT of the 4 quadrants was positively associated with the PRT in the multiple regression analyses which may have been caused by the alterations of the angle of the measuring light beam (Fig. 4). However, the effect was small because the coefficient did not attain significant levels (r = 0.13 to 0.18, PZ0.05) except for the temporal quadrant (r = 0.32, P < 0.001). Hwang et al22 reported that the peripapillary RNFLT was associated with the degree of myopic PRT especially in the temporal area in healthy subjects. In our study, the RNFL of the temporal quadrant was also significantly thicker in eyes with a greater PRT in the multiple regression analysis. Theoretically, if the plane of the peripapillary retina is constant, the RNFLT of all quadrants should be affected equally by an alteration in the angle of the measuring light beam. However, if the tilt on the temporal side of the optic disc is larger than the other quadrants because of myopic changes, the temporal RNFLT should be most affected by the alteration of the angle of the measuring light beam. Most subjects (104/119, 87%) had a
2016 Wolters Kluwer Health, Inc. All rights reserved.
www.glaucomajournal.com |
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
37
Yamashita et al
J Glaucoma
Volume 26, Number 1, January 2017
FIGURE 3. The Pearson correlation coefficients and standardized coefficients of multiple regression for the axial length, peripapillary retinal tilt (PRT), optic disc size, artery angle, and retinal nerve fiber layer thickness (RNFLT). The significantly correlated sectors are shown in gray.
vertically long optic discs. Thus, the temporal tilt is likely to be largest, and this hypothesis may be correct. However, further studies on the shape of the eye as determined by magnetic resonance imaging is needed. The peak angle and the artery angle were significantly and negatively associated with the RNFLT of the temporal quadrant. In addition, the peak angle and the artery angle were significantly and positively associated with the RNFLT of the supra-nasal and inferior quadrants but not with whole RNFLT. The results of this and earlier studies showed that the supra-temporal and infra-temporal peaks of the RNFLT
FIGURE 4. The measuring light is almost vertical (upper red line) to the retinal plane in eye with small peripapillary retinal tilt (PRT) (A). However, the measuring light pass though obliquely (lower red line) against the retinal plane in eye with large PRT (B). Figure 4 can be viewed in color online at www.glaucomajournal.com.
tended to shift toward the fovea as the axial length increased (r = 0.31, P = 0.001).15,18,19 We also showed that the axial length and the peak angle affected the sectoral RNFLT independently.18 Even when the PRT and the optic disc tilt were included in the multiple regression analysis, the peak angle and artery angle influenced the sectoral RNFLT independently. However, the peak angle and artery angle affected only the distribution of the RNFLTs and did not affect the whole RNFLT. Representative images of an eye with a large PRT (A) and a narrow artery angle (B) are shown in Figure 5. The RNFLT between the supra-temporal peak and infratemporal peak is thicker than that of the normal range of thicknesses embedded in the instrument (Fig. 5C). Thus, even though the actual RNFL thickness of the supra-temporal to infra-temporal sector is decreased, the values may still be within the normal range of the embedded data.28 In this study and earlier studies, the optic disc size was not significantly associated with the whole RNFLT in the Pearson correlation analysis.23,24 However, the optic disc size was significantly associated with the whole RNFLT in multiple correlation analysis of this study. Thus, as the size of the optic disc increases, the measuring circle for the RNFLT gets closer to the edge of optic disc. Therefore, it is reasonable that the optic disc size was significantly associated with whole RNFLT. In the earlier study, the effects of such factors as the PRT, peak angle, and artery angle were not assessed. In contrast, we evaluated the peak angle and artery angle, which were significantly associated with the size of the optic disc (r = 0.44, 0.43, P < 0.001). These factors might have masked the actual effect of size of optic disc in the univariate analysis of the earlier studies. The overall r-square values determined by multiple regression analysis of the superior, nasal, inferior, and temporal quadrants and the axial length, PRT and optic disc size, and artery angle were 0.11, 0.25, 0.38, and 0.50, respectively
38 | www.glaucomajournal.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Glaucoma
Volume 26, Number 1, January 2017
Association of PRT and RNFLT in Healthy Eyes
FIGURE 5. Images of a representative eye with a large peripapillary retinal tilt (A) and a narrow artery angle (B). The RNFLT between the supra-temporal peak and infratemporal peak is thicker than that of the normal range of thicknesses embedded in the instrument (C). Figure 5 can be viewed in color online at www.glaucomajournal.com.
(Supplementary Table 2, Supplemental Digital Content 3, http://links.lww.com/IJG/A94). The temporal RNFLT was moderate associated with the axial length, PRT and optic disc size, and artery angle. However, the overall r-square value of the multiple regression analysis of the superior RNFLT was only 0.11. Other factors, such as the RNFLT at birth may influence the RNFLT of these sectors. There are several limitations in this study. There are many factors that can affect the RNFLT, for example, the size, shape, and torsion of the eye, optic disc tilt, and the relative location of the fovea relative to the optic disc.22,29–32 None of these were considered in our analyses. Only the effect of the eye torsion could be minimized by use of the angle between the supra and infra peak RNFLT or angles. There are also many optical factors that can affect the RNFLT, for example, scan circle diameter,33 anterior segment power,34 and the scan angle of the optic nerve head.35 How these other factors affect the relationship between sectoral RNFLT and axial length or artery angle must be determined in future studies. This study was designed to determine the effects of myopic factors on the RNFLT. To examine the effects of these myopic factors, it was first necessary to study eyes with a wide range of refractive errors to obtain baseline values. The Japanese population is known to be the most myopic population in the world. As a result, most of the subjects were myopic. This may have affected the results. In summary, our results indicate that the axial length, the PRT and optic disc size, the peak angle, and the artery Copyright
r
angle are the major factors that independently affect the peripapillary RNFLT measured by OCT. The peak angle, the artery angle, and the PRT can be calculated with only the RNFL circle scan. Therefore, these values may be suitable to adjust the RNFLT values. Our results indicate that the location of peak angle can be determined by referring to the artery angle. These factors should be taken into account when determining the embedded RNFLT normative database for the OCT instrument to ensure more accurate detection of RNFL abnormalities.35 At present, special cautions are needed in the determination of RNFLT abnormalities using the current inbuilt RNFL normative database especially for moderate to highly myopic eyes. ACKNOWLEDGMENT The authors thank Professor Duco Hamasaki of Bascom Palmer Eye Institute, University of Miami, Florida, for providing critical discussions and suggestions to our study and editing of the final manuscript. REFERENCES 1. Sawada A, Tomidokoro A, Araie M, et al. Refractive errors in an elderly Japanese population: the Tajimi study. Ophthalmology. 2008;115:363–370. 2. Mitchell P, Hourihan F, Sandbach J, et al. The relationship between glaucoma and myopia. The Blue Mountain Eye Study. Ophthalmology. 1999;106:2010–2015.
2016 Wolters Kluwer Health, Inc. All rights reserved.
www.glaucomajournal.com |
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
39
Yamashita et al
3. Iwase A, Suzuki Y, Araie M, et al. The prevalence of primary open-angle glaucoma in Japanese. The Tajimi Study. Ophthalmology. 2004;111:1641–1648. 4. Vitale S, Sperduto RD, Ferris FL 3rd. Increased prevalence of myopia in the United States between 1971-1972 and 1999-2004. Arch Ophthalmol. 2009;127:1632–1639. 5. Bloom RI, Friedman IB, Chuck RS. Increasing rates of myopia: the long view. Curr Opin Ophthalmol. 2010;21:247–248. 6. Blumenthal EZ, Williams JM, Weinreb RN, et al. Reproducibility of nerve fiber layer thickness measurements by use of optical coherence tomography. Ophthalmology. 2000;107:2278–2282. 7. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science. 1991;254:1178–1181. 8. Mistlberger A, Liebmann JM, Greenfield DS, et al. Heidelberg retina tomography and optical coherence tomography in normal, ocular-hypertensive, and glaucomatous eyes. Ophthalmology. 1999;106:2027–2032. 9. Zangwill LM, Williams J, Berry CC, et al. A comparison of optical coherence tomography and retinal nerve fiber layer photography for detection of nerve fiber layer damage in glaucoma. Ophthalmology. 2000;107:1309–1315. 10. Leung CK, Mohamed S, Leung KS, et al. Retinal nerve fiber layer measurements in myopia: an Optical Coherence Tomography Study. Invest Ophthalmol Vis Sci. 2006;47:5171–5176. 11. Hoh ST, Lim MC, Seah SK, et al. Peripapillary retinal nerve fiber layer thickness variations with myopia. Ophthalmology. 2006;113:773–777. 12. Leitgeb R, Hitzenberger CK, Fercher AF. Performance of Fourier-domain vs. time-domain optical coherence tomography. Opt Express. 2003;11:889–894. 13. Vizzeri G, Balasubramanian M, Bowd C, et al. Spectoral domain-optical coherence tomography to detect localized retinal nerve fiber layer defects in glaucomatous eyes. Opt Express. 2009;17:4004–4018. 14. Shoji T, Nagaoka Y, Sato H, et al. Impact of high myopia on the performance of SD-OCT parameters to detect glaucoma. Graefes Arch Clin Exp Ophthalmol. 2012;250:1843–1849. 15. Yoo YC, Lee CM, Park JH. Changes in peripapillary retinal nerve fiber layer distribution by axial length. Optom Vis Sci. 2012;89:4–11. 16. Rauscher FM, Sekhon N, Feuer WJ, et al. Myopia affects retinal nerve fiber layer measurements as determined by optical coherence tomography. J Glaucoma. 2009;18:501–505. 17. Kim MJ, Lee EJ, Kim TW. Peripapillary retinal nerve fibre layer thickness profile in subjects with myopia measured using the stratus optical coherence tomography. Br J Ophthalmol. 2010;94:115–120. 18. Yamashita T, Asaoka R, Tanaka M, et al. Relationship between position of peak retinal nerve fiber layer thickness and retinal arteries on sectoral retinal nerve fiber layer thickness. Invest Ophthalmol Vis Sci. 2013;54:5481–5488. 19. Hong SW, Ahn MD, Kang SH, et al. Analysis of peripapillary retinal nerve fiber distribution in normal young adults. Invest Ophthalmol Vis Sci. 2010;51:3515–3523. 20. Hood DC, Fortune B, Arthur SN, et al. Blood vessel contributions to retinal nerve fiber layer thickness profiles
J Glaucoma
21.
22.
23. 24.
25.
26. 27.
28. 29. 30.
31. 32.
33.
34. 35.
Volume 26, Number 1, January 2017
measured with optical coherence tomography. J Glaucoma. 2008;17:519–528. Hood DC, Salant JA, Arthur SN, et al. The location of the inferior and superior temporal blood vessels and interindividual variability of the retinal nerve fiber layer thickness. J Glaucoma. 2010;19:158–166. Hwang YH, Yoo C, Kim YY. Myopic optic disc tilt and the characteristics of peripapillary retinal nerve fiber layer thickness measured by spectral-domain optical coherence tomography. J Glaucoma. 2012;21:260–265. Bendschneider D, Tornow RP, Horn FK, et al. Retinal nerve fiber layer thickness in normals measured by spectral domain OCT. J Glaucoma. 2010;19:475–482. Mansoori T, Viswanath K, Balakrishna N. Correlation between peripapillary retinal nerve fiber layer thickness and optic nerve head parameters using spectral domain optical coherence tomography. J Glaucoma. 2010;19:604–608. Yamashita T, Sakamoto T, Yoshihara N, et al. Circumpapillary course of retinal pigment epithelium can be fit to sine wave and amplitude of sine wave is significantly correlated with ovality ratio of optic disc. PLoS One. 2015;10:e0122191. Garway-Heath DF, Rudnicka AR, Lowe T, et al. Measurement of optic disc size: equivalence of methods to correct for ocular magnification. Br J Ophthalmol. 1998;82:643–649. Bennett AG, Rudnicka AR, Edgar DF. Improvements on Littmann’s method of determining the size of retinal features by fundus photography. Graefes Arch Clin Exp Ophthalmol. 1994;232:361–367. Yamashita T, Kii Y, Tanaka M, et al. Relationship between supernormal sectors of retinal nerve fibre layer and axial length in normal eyes. Acta Ophthalmol. 2014;92:e481–e487. Garway-Heath DF, Poinoosawmy D, Fitzke FW, et al. Mapping the visual field to the optic disc in normal tension glaucoma eyes. Ophthalmology. 2000;100:1809–1815. Lamparter J, Russell RA, Zhu H, et al. The influence of intersubject variability in ocular anatomical variables on the mapping of retinal locations to the retinal nerve fiber layer and optic nerve head. Invest Ophthalmol Vis Sci. 2013;54: 6074–6082. Takasaki H, Higashide T, Takeda H, et al. Relationship between optic disc ovality and horizontal disc tilt in normal young subjects. Jpn J Ophthalmol. 2013;57:34–40. Hosseini H, Nassiri N, Azarbod P, et al. Measurement of the optic disc vertical tilt angle with spectral-domain optical coherence tomography and influencing factors. Am J Ophthalmol. 2013;156:737–744. Patel NB, Luo X, Wheat JL, et al. Retinal nerve fiber layer assessment: area versus thickness measurements from elliptical scans centered on the optic nerve. Invest Ophthalmol Vis Sci. 2011;52:2477–2489. Patel NB, Garcia B, Harwerth RS. Influence of anterior segment power on the scan path and RNFLT using SD-OCT. Invest Ophthalmol Vis Sci. 2012;53:5788–5798. Hong S, Kim CY, Seong GJ. Adjusted peripapillary retinal nerve fiber layer thickness measurements based on the optic nerve head scan angle. Invest Ophthalmol Vis Sci. 2010;51:4067–4074.
40 | www.glaucomajournal.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
Correlations Between Retinal Nerve Fiber Layer Thickness and Axial Length, Peripapillary Retinal Tilt, Optic Disc Size, and Retinal Artery Position in Healthy Eyes Takehiro Yamashita, MD, PhD, Taiji Sakamoto, MD, PhD, Naoya Yoshihara, MD, Hiroto Terasaki, MD, PhD, Minoru Tanaka, MD, Yuya Kii, MD, PhD, and Kumiko Nakao, MD, PhD
Purpose: To determine the correlations between the retinal nerve fiber layer thickness (RNFLT) and the axial length, peripapillary retinal tilt (PRT), and optic disc size, and retinal artery position. Methods: A prospective, observational cross-sectional study of 119 healthy right eyes of 119 volunteers. All participants underwent comprehensive ophthalmologic examinations including peripapillary RNFLT imaging and measurements of the axial length. The RNFLT was determined by the TOPCON 3D OCT-1000, MARK II. The RNFLT in a 3.4 mm circular scan was divided into 12 clock-hour sectors and 4 quadrant sectors around the optic disc. The PRT was assessed using the RNFLT B-scan images. The angle between the supra-temporal and infra-temporal retinal arteries was determined in the color fundus photographs. The correlations between the sectorial RNFLTs and the axial length, PRT, optic disc size, and artery angles were determined by simple and multiple regression analyses. Results: Multiple regression analyses showed that the nasal and inferior quadrants and the whole RNFLT were significantly and negatively correlated with the axial length (standardized coefficient (SC) = 0.39 to 0.30, P < 0.05). The PRT was significantly and positively associated with all of the quadrants and the whole RNFLT (SC = 0.22 to 0.45, P < 0.05). The retinal artery angle was significantly and negatively associated with the temporal RNFLT and positively associated with inferior RNFLT (SC = 0.49 to 0.31, P < 0.05). The optic disc size was significantly and positively associated with the superior and nasal quadrants, and the whole RNFLT (SC = 0.20 to 0.27, P < 0.05). Conclusions: The axial length, PRT, optic disc size, retinal artery angle can affect the peripapillary RNFLT. These variables should be considered when assessing the peripapillary RNFLT. Key Words: retinal nerve fiber layer thickness, axial length, peripapillary retinal tilt, retinal artery position, optic disc size
(J Glaucoma 2017;26:34–40)
Received for publication November 26, 2015; accepted August 22, 2016. From the Department of Ophthalmology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan. Disclosure: The authors declare no conflict of interest. Reprints: Taiji Sakamoto, MD, PhD, Department of Ophthalmology, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, 890-8520, Japan (e-mail: tsakamot@m3. kufm.kagoshima-u.ac.jp). Supplemental Digital Content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s Website, www.glaucomajournal.com. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/IJG.0000000000000550
I
t is widely accepted that myopia is a risk factor for primary open angle glaucoma.1–3 Considering the increasing prevalence of myopia worldwide, a correct diagnosis of glaucoma in myopic patients is becoming more important and necessary in ophthalmology.1,4,5 To detect the early stages of glaucoma in myopic eyes, it is necessary to assess the changes in the optic disc and alterations of the retinal nerve fiber layer thickness (RNFLT) accurately. However, it is difficult to make these assessments in myopic eyes because the RNFLT can be altered by the myopic changes in eyes without glaucoma. In addition, the effect of myopia without glaucoma on the peripapillary RNFLT has not been definitively determined. Optical coherence tomography (OCT) is a noninvasive imaging technique that can obtain images which can be used to evaluate the morphology and thickness of the RNFL and optic disc with micrometer resolution.6,7 Several studies have documented that reliable measurements of the RNFLT can be made from the OCT images.8–13 However, the RNFLT measurements can be significantly affected by the refractive errors, and thus the ability to detect glaucoma in highly myopic eyes by the RNFLT is poorer than that in emmetropic eyes.14 It is well recognized that the peripapillary RNFLT is affected by the axial length,15–18 the peak position of the RNFLT,15,18,19 position of the retinal vessel,18,20,21 and the tilt and size of the optic disc.22–24 In an earlier study, we reported that a mathematical method can be used to determine the angle of the retinal vessels, and we found that both the axial length and the peak artery angle influenced the sectoral RNFLT.18 We also developed another mathematical technique, a sine-curve analysis technique, to measure the degree of the peripapillary retinal tilt (PRT).25 Because these techniques allowed us to express each variable as a numerical value, it was possible to analyze the interrelationships among these factors quantitatively. If these relationships are valid, then the values may be suitable for determining how they are associated with the RFNLT. Thus, the purpose of this study was to determine the relationships between the sectoral and whole RNFLT and the peak angle, artery angle, PRT, optic disc size, and the axial length of healthy eyes of normal individuals. Because these variables were correlated to each other, we also determined which factors affected the sectoral and whole RNFLT using multiple regression analyses.
METHODS All of the procedures used conformed to the tenets of the Declaration of Helsinki. A written informed consent
34 | www.glaucomajournal.com J Glaucoma Volume 26, Number 1, January 2017 Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Glaucoma
Volume 26, Number 1, January 2017
was obtained from all of the subjects after an explanation of the procedures to be used. The study was approved by the Ethics Committee of Kagoshima University Hospital, and it was registered with the University Hospital Medical Network (UMIN)-clinical trials registry. The registration title was, “Morphological analysis of the optic disc and the retinal nerve fiber in myopic eyes” and the registration number was UMIN000006040. A detailed protocol is available at: https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr. cgi?function = brows&action = brows&type = summary& recptno = R000007154&language = J.
Subjects This was a cross-sectional, prospective, observational study. We initially examined 133 eyes of 133 volunteers who were enrolled between November 1, 2010 and February 29, 2012. The volunteers had no known eye diseases as determined by examining their medical charts, and the data from only the right eyes were analyzed. The eligibility criteria were: age 20 years or above but 40 years or below; eyes normal by slit-lamp biomicroscopy, ophthalmoscopy, and OCT; best-corrected visual acuity of r0.1 logarithm of the minimum angle of resolution (logMAR) units; and intraocular pressure (IOP) r21 mmHg. The exclusion criteria were eyes with known ocular diseases such as glaucoma, presence of a staphyloma, and optic disc anomalies; presence of visual field defects; and prior refractive or intraocular surgery. The presence of a staphyloma was determined by B-mode echo. None of the eyes was excluded because of poor OCT image quality caused by poor fixation.
Measurement of Axial Length and Refractive Error All eyes had a standard ocular examination including; slit-lamp biomicroscopy of the anterior segment, ophthalmoscopy of the ocular fundus, pneumo-tonometric (CT-80; Topcon, Tokyo, Japan) measurements of the IOP, and AL-2000 ultrasonographic (Tomey, Japan) measurements of the axial length. The refractive error (spherical equivalent) was measured with the Topcon KR8800 autorefractometer/keratometer.
Determination of Thickness of RNFL, Peak Angle, Artery Angle, PRT, and Optic Disc Size All eyes were examined by a single examiner (T.Y.). The RNFLT was measured with the TOPCON 3D OCT1000 MARK II using the RNFL 3.4 mm circle scan. In this
FIGURE 1. The retinal nerve fiber layer thickness of the whole, 4 quadrants, and the 12 clock hours were averaged. I indicates inferior; N, nasal; S, superior; T, temporal; W, whole.
Copyright
r
Association of PRT and RNFLT in Healthy Eyes
protocol, 1024 A-scans/circle, 16 overlapping B-scans/ image, and direct B-scan observations were made. The OCT images and the color fundus photographs were recorded at the same time. The optical system of the OCT instrument detected the edge of the optic disc in the fundus image, and the scan circle was centered automatically on the optic disc just before the OCT image was recorded. To exclude the effects of errors in the scan circle centration, 1 examiner (Y.K.) checked offline that the center of the scan circle was located at the center of the optic disc. The average RNFLT of the whole, each of the 4 quadrants, and each of the clock-hour sectors were measured (Fig. 1). The temporal-superior-nasal-inferiortemporal thickness curves were used to measure the angle between supra-temporal and infra-temporal peaks of the RNFL. We determined this peak angle in the temporalsuperior-nasal-inferior-temporal thickness profile of the RNFLT analyses. The distance between the peak RNFLTs for the supra-temporal and infra-temporal RNFL peaks was determined by dragging a vertical line in the profile graph in the Photoshop CS5 program. Then, the distance between supra-temporal and infra-temporal RNFL peaks (Supplementary Figs. 1A, X, Supplemental Digital Content 1, http://links.lww.com/IJG/A92) was converted to an angular value by dividing it by the entire distance (Supplementary Figs. 1A, Y, Supplemental Digital Content 1, http://links.lww.com/IJG/A92), then multiplied by 360.15,18 Color fundus photographs and OCT images were taken at the same time with the TOPCON 3D OCT-1000 MARK II. The scan circle was placed over the optic disc in the photographs as a green circle. The points where the green circle intersected the supra-temporal and infratemporal major retinal arteries were determined, and the angle between supra-temporal and infra-temporal major retinal artery was measured (Supplementary Fig. 1B, Supplemental Digital Content 1, http://links.lww.com/IJG/ A92). We named this angle the artery angle.18 The course of the retinal pigment epithelium was marked on the B-scan images manually. The pixel coordinates of each mark were determined automatically using the Image J program [Image J version 1.47, National Institutes of Health, Bethesda, MD; http://imagej.nih.gov/ ij/ (in the public domain)]. The “x” and “y” coordinates of each pixel were converted to a new set of “x” and “y” coordinates with the center of the curve. Finally, the converted data were fit to a sine-curve equation (y = a sin(b x-c)) with the curve fitting program of Image J. The “a,” “b,” and “c” are constants calculated by the least squares method of the curve fitting program of Image J. The constant “a” is the amplitude of the sine curve, and a larger amplitude “a” will make the amplitude of the curve larger and make the arms of the PRT larger. Because the eyes were relatively stationary throughout the measurements, the retinal plane of macular area was fixed at an orientation approximately perpendicular to the reference light axis. Then, the PRT examined was represented by the angle between the optic axis and the PRT. The amplitude of the sine curve, “a,” was considered to reflect the degree of the PRT relative to the optical axis.25 The details of the measurement of the peak angle, the artery angle, and amplitude of the sine curve were described in more detail in earlier publications.18,25 We measured the size of the optic disc in the color fundus photographs by Image J after adjusting for the magnification effect.18,26,27
2016 Wolters Kluwer Health, Inc. All rights reserved.
www.glaucomajournal.com |
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
35
Yamashita et al
J Glaucoma
TABLE 1. Participants Date Age (y) Sex (male/female) Spherical equivalent (D) Axial length (mm) Peak angle (deg.) Artery angle (deg.) Amplitude (pixels)
Mean ± SD
Range
25.9 ± 4.0 81/38 4.68 ± 3.36 25.4 ± 1.4 124.2 ± 20.5 128.8 ± 23.1 37.4 ± 17.2
22-40 14.25-0.50 22.4-30.4 74.4-193.9 77.9-193.9 8.8-80.8
Volume 26, Number 1, January 2017
The axial length was significantly and negatively correlated with the peak angle (r = 0.31, P = 0.001) and the artery angle (r = 0.41, P < 0.001). The peak angle was significantly correlated with the artery angle (r = 0.79, P < 0.001). Because peak angle and artery angle were highly and significantly correlated, they were analyzed separately to exclude the effects of multicollinearity. Specifically, the correlation of either peak angle or artery angle to the other factors was analyzed. The other factors included the RNFLT, axial length, optic disc size, and PRT. The average RNFLT of the whole, 4 quadrants, and the 12 clock-hour sectors are presented in Table 2.
Statistical Analyses All statistical analyses were performed with the SPSS statistics 19 for Windows (SPSS Inc., IBM, Somers, New York). The relationships between the whole or sectoral RNFLTs and the axial length, peak angle, artery angle, PRT, and optic disc size were determined by the Pearson correlation analyses. Multiple regression analyses were calculated for the whole and sectoral RNFLTs and the same factors because each of these factors were significantly correlated with the RNFLT.
RESULTS In total, 133 Japanese volunteers were examined. Seven eyes were excluded because of ocular diseases or prior ocular surgery, 3 eyes because of superior segmental optic hypoplasia, 1 case because of glaucoma, and 3 cases because of prior laser-assisted in situ keratomileusis. Seven other eyes were excluded because of difficulty in identifying the position of the peak of the RNFLT. In the end, the right eyes of 119 individuals (81 men and 38 women) were studied. The demographic information of the 119 subjects is presented in Table 1. The mean ± SD of the age was 25.9 ± 4.0 years, and the mean refractive error (spherical equivalent) was 4.68 ± 3.36 D. The mean axial length was 25.4 ± 1.4 mm. The mean peak angle was 124.2 ± 20.5 degrees, the mean artery angle was 128.8 ± 23.1 degrees, the mean PRT was 37.4 ± 17.2 pixels, and the mean optic disc size was 2006.9 ± 474.7 pixels.
TABLE 2. Retinal Nerve Fiber Layer Thickness Values
Locations
Sector
Mean ± SD
Range
Superior Nasal Nasal Nasal Inferior Inferior Inferior Temporal Temporal Temporal Superior Superior Superior Nasal Inferior Temporal Total
1 2 3 4 5 6 7 8 9 10 11 12 11, 12, 1 2, 3, 4 5, 6, 7 8, 9, 10 Whole
125.0 ± 18.5 97.6 ± 18.6 80.2 ± 14.7 80.2 ± 16.1 104.9 ± 19.8 128.7 ± 30.9 151.6 ± 20.3 101.7 ± 24.8 81.2 ± 15.5 114.3 ± 23.7 153.8 ± 22.2 130.5 ± 23.4 136.4 ± 15.9 85.9 ± 14.7 128.3 ± 19.0 99.0 ± 19.8 112.4 ± 10.6
71-183 51-144 44-122 45-125 56-159 66-218 104-205 65-191 54-130 76-184 95-215 86-206 97-182 53-121 80-169 67-165 89-136
The Pearson Correlation Coefficients and Standardized Coefficients (SCs) by Multiple Regression Analyses for Axial Length, PRT, Optic Disc Size, Peak Angle, Artery Angle, and RNFLT The Pearson correlation coefficients and the SCs of multiple regression for the peak angle, axial length, PRT, optic disc size, and the RNFLTs are shown in Fig. 2 and Supplementary Table 1 (Supplemental Digital Content 2, http://links.lww.com/IJG/A93). For both types of regression analyses, the whole RNFLT was significantly and negatively correlated with the axial length (r = 0.32 and 0.33, P < 0.05), significantly and positively correlated with the PRT (r = 0.32 and 0.41, P < 0.05), but not significantly correlated with the peak angle (r = 0.08 and 0.12, PZ0.05). The optic disc size was not significantly associated with the whole RNFLT in the Pearson correlation analysis (r = 0.11, PZ0.05), but significantly and positively associated with the whole RNFLT in the multiple regression analysis (r = 0.27, P < 0.05). In the Pearson correlation analysis, the temporal RNFLT was significantly and positively correlated with the axial length (r = 0.19 to 0.31, P < 0.05), and the nasal and inferior RNFLTs were significantly and negatively correlated with the axial length (r = 0.31 to 0.56, P < 0.05). Multiple regression analyses showed that the temporal to superior RNFLT was not significantly associated with the axial length (r = 0.06 to 0.15, PZ0.05), but the RNFLTs of the nasal and inferior quadrants were significantly associated with the axial length (r = 0.24 to 0.44, P < 0.05). The Pearson correlation analysis showed that the RNFLTs of the temporal and the 3 o’clock sectors were significantly and positively correlated with the PRT (r = 0.32 to 0.61, P < 0.05). Multiple regression analyses showed that the correlation between the temporal quadrant and 3 o’clock sector and the PRT were weakly but significantly correlated (r = 0.17 to 0.44, P < 0.05). The Pearson correlation analysis showed that the RNFLT of the temporal quadrant was significantly and negatively correlated with peak angle (r = 0.25 to 0.78, P < 0.05), and the RNFLT of the superior and inferior quadrants were significantly and positively correlated with the peak angle (r = 0.21 to 0.56, P < 0.05). A similar tendency was seen in the multiple regression analysis, but the coefficients of correlation were lower. The Pearson correlation analysis showed that the RNFLT of the temporal quadrant was significantly and negatively correlated with optic disc size (r = 0.24 to 0.24, P < 0.05), and the RNFLT of the other quadrants were significantly and positively correlated with the optic disc size (r = 0.21 to 0.33, P < 0.05). However, multiple regression analyses showed that the optic disc size was not significantly correlated with the temporal and
36 | www.glaucomajournal.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Glaucoma
Volume 26, Number 1, January 2017
Association of PRT and RNFLT in Healthy Eyes
FIGURE 2. The Pearson correlation coefficients and standardized coefficients of multiple regression for the axial length, peripapillary retinal tilt (PRT), optic disc size, peak angle, and retinal nerve fiber layer thickness (RNFLT). The significant correlated sectors are shown in gray.
inferior quadrants RNFLT (r = 0.05 to 0.10, PZ0.05) and significantly positively associated with whole, superior, and nasal quadrant RNFLT (r = 0.20 to 0.27, P < 0.05). We also investigated these relationships by changing the peak angle to the artery angle. The Pearson correlation coefficients and SCs of multiple regression correlation for the artery angle, axial length, PRT and optic disc size, and RNFLT are shown in Figure 3 and Supplementary Table 2 (Supplemental Digital Content 3, http://links.lww.com/ IJG/A94). A similar tendency was seen in the Pearson correlation coefficients and standardized correlation coefficients when the peak angle was changed to the artery angle.
DISCUSSION The multiple regression analyses showed that the axial length was significantly and negatively associated with the RNFLT of the nasal and inferior quadrants and also with the whole RNFLT. However, these correlations may be caused by the magnification effect or retinal thinning caused by an elongation of the axial length. Kim et al17 examined the association between the axial length and RNFLT in low, moderate, and high myopic eyes. They reported that the average global RNFL was significantly thinner in the high myopic eyes than in the low myopic eyes. For the quadrants, the RNFL was thicker in the low myopic eyes than in the moderate and the high myopic eyes in the superior, nasal, and inferior quadrants. The temporal quadrant was thinner in the low myopia group than in the moderate and high myopia groups. Their results are similar to our results from the Pearson correlation analysis, that is, the longer axial length eyes had a thinner average RNFL in the superior, nasal, and inferior quadrants, and a thicker temporal quadrant. However, multiple regression analysis showed that the relationship between the axial length and Copyright
r
the temporal sectoral RNFLT was not significant. The significant correlation may have been caused by the effect of the PRT, and the peak angle or the artery angle. In addition, we did not adjust the angle for the axial length or refractive error which might have led to differences in the magnification. The scan circle is projected larger than the actual circle after an elongation of the axial length caused by the magnification effect.26,27 This might affect the RNFLT, PRT, peak angle, and artery angle. Unfortunately, it may not be possible to obtain the absolute RNFLT values corrected for the axial length for our OCT instrument because the effect of distance is not known. Further studies using OCT with a magnification effect correction system are needed. The PRT was significantly and positively associated with the RNFLT at the 3 o’clock sector, the temporal quadrant, and the whole region. In addition, the RNFLT of the 4 quadrants was positively associated with the PRT in the multiple regression analyses which may have been caused by the alterations of the angle of the measuring light beam (Fig. 4). However, the effect was small because the coefficient did not attain significant levels (r = 0.13 to 0.18, PZ0.05) except for the temporal quadrant (r = 0.32, P < 0.001). Hwang et al22 reported that the peripapillary RNFLT was associated with the degree of myopic PRT especially in the temporal area in healthy subjects. In our study, the RNFL of the temporal quadrant was also significantly thicker in eyes with a greater PRT in the multiple regression analysis. Theoretically, if the plane of the peripapillary retina is constant, the RNFLT of all quadrants should be affected equally by an alteration in the angle of the measuring light beam. However, if the tilt on the temporal side of the optic disc is larger than the other quadrants because of myopic changes, the temporal RNFLT should be most affected by the alteration of the angle of the measuring light beam. Most subjects (104/119, 87%) had a
2016 Wolters Kluwer Health, Inc. All rights reserved.
www.glaucomajournal.com |
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
37
Yamashita et al
J Glaucoma
Volume 26, Number 1, January 2017
FIGURE 3. The Pearson correlation coefficients and standardized coefficients of multiple regression for the axial length, peripapillary retinal tilt (PRT), optic disc size, artery angle, and retinal nerve fiber layer thickness (RNFLT). The significantly correlated sectors are shown in gray.
vertically long optic discs. Thus, the temporal tilt is likely to be largest, and this hypothesis may be correct. However, further studies on the shape of the eye as determined by magnetic resonance imaging is needed. The peak angle and the artery angle were significantly and negatively associated with the RNFLT of the temporal quadrant. In addition, the peak angle and the artery angle were significantly and positively associated with the RNFLT of the supra-nasal and inferior quadrants but not with whole RNFLT. The results of this and earlier studies showed that the supra-temporal and infra-temporal peaks of the RNFLT
FIGURE 4. The measuring light is almost vertical (upper red line) to the retinal plane in eye with small peripapillary retinal tilt (PRT) (A). However, the measuring light pass though obliquely (lower red line) against the retinal plane in eye with large PRT (B). Figure 4 can be viewed in color online at www.glaucomajournal.com.
tended to shift toward the fovea as the axial length increased (r = 0.31, P = 0.001).15,18,19 We also showed that the axial length and the peak angle affected the sectoral RNFLT independently.18 Even when the PRT and the optic disc tilt were included in the multiple regression analysis, the peak angle and artery angle influenced the sectoral RNFLT independently. However, the peak angle and artery angle affected only the distribution of the RNFLTs and did not affect the whole RNFLT. Representative images of an eye with a large PRT (A) and a narrow artery angle (B) are shown in Figure 5. The RNFLT between the supra-temporal peak and infratemporal peak is thicker than that of the normal range of thicknesses embedded in the instrument (Fig. 5C). Thus, even though the actual RNFL thickness of the supra-temporal to infra-temporal sector is decreased, the values may still be within the normal range of the embedded data.28 In this study and earlier studies, the optic disc size was not significantly associated with the whole RNFLT in the Pearson correlation analysis.23,24 However, the optic disc size was significantly associated with the whole RNFLT in multiple correlation analysis of this study. Thus, as the size of the optic disc increases, the measuring circle for the RNFLT gets closer to the edge of optic disc. Therefore, it is reasonable that the optic disc size was significantly associated with whole RNFLT. In the earlier study, the effects of such factors as the PRT, peak angle, and artery angle were not assessed. In contrast, we evaluated the peak angle and artery angle, which were significantly associated with the size of the optic disc (r = 0.44, 0.43, P < 0.001). These factors might have masked the actual effect of size of optic disc in the univariate analysis of the earlier studies. The overall r-square values determined by multiple regression analysis of the superior, nasal, inferior, and temporal quadrants and the axial length, PRT and optic disc size, and artery angle were 0.11, 0.25, 0.38, and 0.50, respectively
38 | www.glaucomajournal.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
J Glaucoma
Volume 26, Number 1, January 2017
Association of PRT and RNFLT in Healthy Eyes
FIGURE 5. Images of a representative eye with a large peripapillary retinal tilt (A) and a narrow artery angle (B). The RNFLT between the supra-temporal peak and infratemporal peak is thicker than that of the normal range of thicknesses embedded in the instrument (C). Figure 5 can be viewed in color online at www.glaucomajournal.com.
(Supplementary Table 2, Supplemental Digital Content 3, http://links.lww.com/IJG/A94). The temporal RNFLT was moderate associated with the axial length, PRT and optic disc size, and artery angle. However, the overall r-square value of the multiple regression analysis of the superior RNFLT was only 0.11. Other factors, such as the RNFLT at birth may influence the RNFLT of these sectors. There are several limitations in this study. There are many factors that can affect the RNFLT, for example, the size, shape, and torsion of the eye, optic disc tilt, and the relative location of the fovea relative to the optic disc.22,29–32 None of these were considered in our analyses. Only the effect of the eye torsion could be minimized by use of the angle between the supra and infra peak RNFLT or angles. There are also many optical factors that can affect the RNFLT, for example, scan circle diameter,33 anterior segment power,34 and the scan angle of the optic nerve head.35 How these other factors affect the relationship between sectoral RNFLT and axial length or artery angle must be determined in future studies. This study was designed to determine the effects of myopic factors on the RNFLT. To examine the effects of these myopic factors, it was first necessary to study eyes with a wide range of refractive errors to obtain baseline values. The Japanese population is known to be the most myopic population in the world. As a result, most of the subjects were myopic. This may have affected the results. In summary, our results indicate that the axial length, the PRT and optic disc size, the peak angle, and the artery Copyright
r
angle are the major factors that independently affect the peripapillary RNFLT measured by OCT. The peak angle, the artery angle, and the PRT can be calculated with only the RNFL circle scan. Therefore, these values may be suitable to adjust the RNFLT values. Our results indicate that the location of peak angle can be determined by referring to the artery angle. These factors should be taken into account when determining the embedded RNFLT normative database for the OCT instrument to ensure more accurate detection of RNFL abnormalities.35 At present, special cautions are needed in the determination of RNFLT abnormalities using the current inbuilt RNFL normative database especially for moderate to highly myopic eyes. ACKNOWLEDGMENT The authors thank Professor Duco Hamasaki of Bascom Palmer Eye Institute, University of Miami, Florida, for providing critical discussions and suggestions to our study and editing of the final manuscript. REFERENCES 1. Sawada A, Tomidokoro A, Araie M, et al. Refractive errors in an elderly Japanese population: the Tajimi study. Ophthalmology. 2008;115:363–370. 2. Mitchell P, Hourihan F, Sandbach J, et al. The relationship between glaucoma and myopia. The Blue Mountain Eye Study. Ophthalmology. 1999;106:2010–2015.
2016 Wolters Kluwer Health, Inc. All rights reserved.
www.glaucomajournal.com |
Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
39
Yamashita et al
3. Iwase A, Suzuki Y, Araie M, et al. The prevalence of primary open-angle glaucoma in Japanese. The Tajimi Study. Ophthalmology. 2004;111:1641–1648. 4. Vitale S, Sperduto RD, Ferris FL 3rd. Increased prevalence of myopia in the United States between 1971-1972 and 1999-2004. Arch Ophthalmol. 2009;127:1632–1639. 5. Bloom RI, Friedman IB, Chuck RS. Increasing rates of myopia: the long view. Curr Opin Ophthalmol. 2010;21:247–248. 6. Blumenthal EZ, Williams JM, Weinreb RN, et al. Reproducibility of nerve fiber layer thickness measurements by use of optical coherence tomography. Ophthalmology. 2000;107:2278–2282. 7. Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science. 1991;254:1178–1181. 8. Mistlberger A, Liebmann JM, Greenfield DS, et al. Heidelberg retina tomography and optical coherence tomography in normal, ocular-hypertensive, and glaucomatous eyes. Ophthalmology. 1999;106:2027–2032. 9. Zangwill LM, Williams J, Berry CC, et al. A comparison of optical coherence tomography and retinal nerve fiber layer photography for detection of nerve fiber layer damage in glaucoma. Ophthalmology. 2000;107:1309–1315. 10. Leung CK, Mohamed S, Leung KS, et al. Retinal nerve fiber layer measurements in myopia: an Optical Coherence Tomography Study. Invest Ophthalmol Vis Sci. 2006;47:5171–5176. 11. Hoh ST, Lim MC, Seah SK, et al. Peripapillary retinal nerve fiber layer thickness variations with myopia. Ophthalmology. 2006;113:773–777. 12. Leitgeb R, Hitzenberger CK, Fercher AF. Performance of Fourier-domain vs. time-domain optical coherence tomography. Opt Express. 2003;11:889–894. 13. Vizzeri G, Balasubramanian M, Bowd C, et al. Spectoral domain-optical coherence tomography to detect localized retinal nerve fiber layer defects in glaucomatous eyes. Opt Express. 2009;17:4004–4018. 14. Shoji T, Nagaoka Y, Sato H, et al. Impact of high myopia on the performance of SD-OCT parameters to detect glaucoma. Graefes Arch Clin Exp Ophthalmol. 2012;250:1843–1849. 15. Yoo YC, Lee CM, Park JH. Changes in peripapillary retinal nerve fiber layer distribution by axial length. Optom Vis Sci. 2012;89:4–11. 16. Rauscher FM, Sekhon N, Feuer WJ, et al. Myopia affects retinal nerve fiber layer measurements as determined by optical coherence tomography. J Glaucoma. 2009;18:501–505. 17. Kim MJ, Lee EJ, Kim TW. Peripapillary retinal nerve fibre layer thickness profile in subjects with myopia measured using the stratus optical coherence tomography. Br J Ophthalmol. 2010;94:115–120. 18. Yamashita T, Asaoka R, Tanaka M, et al. Relationship between position of peak retinal nerve fiber layer thickness and retinal arteries on sectoral retinal nerve fiber layer thickness. Invest Ophthalmol Vis Sci. 2013;54:5481–5488. 19. Hong SW, Ahn MD, Kang SH, et al. Analysis of peripapillary retinal nerve fiber distribution in normal young adults. Invest Ophthalmol Vis Sci. 2010;51:3515–3523. 20. Hood DC, Fortune B, Arthur SN, et al. Blood vessel contributions to retinal nerve fiber layer thickness profiles
J Glaucoma
21.
22.
23. 24.
25.
26. 27.
28. 29. 30.
31. 32.
33.
34. 35.
Volume 26, Number 1, January 2017
measured with optical coherence tomography. J Glaucoma. 2008;17:519–528. Hood DC, Salant JA, Arthur SN, et al. The location of the inferior and superior temporal blood vessels and interindividual variability of the retinal nerve fiber layer thickness. J Glaucoma. 2010;19:158–166. Hwang YH, Yoo C, Kim YY. Myopic optic disc tilt and the characteristics of peripapillary retinal nerve fiber layer thickness measured by spectral-domain optical coherence tomography. J Glaucoma. 2012;21:260–265. Bendschneider D, Tornow RP, Horn FK, et al. Retinal nerve fiber layer thickness in normals measured by spectral domain OCT. J Glaucoma. 2010;19:475–482. Mansoori T, Viswanath K, Balakrishna N. Correlation between peripapillary retinal nerve fiber layer thickness and optic nerve head parameters using spectral domain optical coherence tomography. J Glaucoma. 2010;19:604–608. Yamashita T, Sakamoto T, Yoshihara N, et al. Circumpapillary course of retinal pigment epithelium can be fit to sine wave and amplitude of sine wave is significantly correlated with ovality ratio of optic disc. PLoS One. 2015;10:e0122191. Garway-Heath DF, Rudnicka AR, Lowe T, et al. Measurement of optic disc size: equivalence of methods to correct for ocular magnification. Br J Ophthalmol. 1998;82:643–649. Bennett AG, Rudnicka AR, Edgar DF. Improvements on Littmann’s method of determining the size of retinal features by fundus photography. Graefes Arch Clin Exp Ophthalmol. 1994;232:361–367. Yamashita T, Kii Y, Tanaka M, et al. Relationship between supernormal sectors of retinal nerve fibre layer and axial length in normal eyes. Acta Ophthalmol. 2014;92:e481–e487. Garway-Heath DF, Poinoosawmy D, Fitzke FW, et al. Mapping the visual field to the optic disc in normal tension glaucoma eyes. Ophthalmology. 2000;100:1809–1815. Lamparter J, Russell RA, Zhu H, et al. The influence of intersubject variability in ocular anatomical variables on the mapping of retinal locations to the retinal nerve fiber layer and optic nerve head. Invest Ophthalmol Vis Sci. 2013;54: 6074–6082. Takasaki H, Higashide T, Takeda H, et al. Relationship between optic disc ovality and horizontal disc tilt in normal young subjects. Jpn J Ophthalmol. 2013;57:34–40. Hosseini H, Nassiri N, Azarbod P, et al. Measurement of the optic disc vertical tilt angle with spectral-domain optical coherence tomography and influencing factors. Am J Ophthalmol. 2013;156:737–744. Patel NB, Luo X, Wheat JL, et al. Retinal nerve fiber layer assessment: area versus thickness measurements from elliptical scans centered on the optic nerve. Invest Ophthalmol Vis Sci. 2011;52:2477–2489. Patel NB, Garcia B, Harwerth RS. Influence of anterior segment power on the scan path and RNFLT using SD-OCT. Invest Ophthalmol Vis Sci. 2012;53:5788–5798. Hong S, Kim CY, Seong GJ. Adjusted peripapillary retinal nerve fiber layer thickness measurements based on the optic nerve head scan angle. Invest Ophthalmol Vis Sci. 2010;51:4067–4074.
40 | www.glaucomajournal.com Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved. Copyright r 2016 Wolters Kluwer Health, Inc. All rights reserved.
