Relationship Between Position of Peak Retinal Nerve Fiber Layer Thickness and Retinal Arteries on Sectoral Retinal Nerve Fiber Layer Thickness

Abstract

Purpose.: We determined the relationship between the position of the peak of the retinal nerve fiber layer (RNFL) thickness, and the retinal arteries, axial length (AL), and sectoral RNFL thickness in healthy eyes.

Methods.: A prospective, observational cross-sectional study (registration number, UMIN000006040) of 50 healthy right eyes (mean age 25.8 ± 3.7 years) was performed. The RNFL thickness was measured by optical coherence tomography in twelve 30° sectors (clock hours) around the optic disc. The RNFL nasal–superior–temporal–inferior–nasal curves and fundus photographs were used to measure the angles between the supratemporal and infratemporal peak RNFL positions (peak angle), and the retinal artery angle (artery angle), respectively. The relationships between the peak angle, artery angle, AL, and sectoral RNFL thickness were investigated by linear regression analyses.

Results.: The peak angles were highly correlated with the artery angle (r = 0.92, P < 0.001) and correlated negatively with the AL (r = −0.49, −0.38; P < 0.01). After excluding the effect of the AL, the peak and artery angles were correlated significantly with the sectoral RNFL thickness in 8 sectors. After excluding the effect of the peak angle, the AL was correlated significantly with the sectoral RNFL thicknesses in only one sector.

Conclusions.: The temporal RNFL thickness increased as the superior and inferior RNFL peaks, and retinal arteries shifted toward the fovea, whereas an inverse relationship was observed for the inferior and supranasal areas. The sectoral RNFL thickness is correlated better with the peak and artery angles than the axial length. ( https://upload.umin.ac.jp/cgi-open-bin/ctr/ctr.cgi?function=brows&action=brows&type=summary &recptno=R000007154&language=J number, UMIN000006040.)

Introduction

The Tajimi study in Japan found that the incidence of myopia in the Japanese population was the highest in the world, with an incidence of 41.8% for myopia > −0.5 diopters (D) and 5.5% for myopia > −6.0 D in individuals ≥ 40 years. ¹ The Tajimi study also found that myopia was a risk factor for primary open angle glaucoma (POAG) confirming other population-based studies that individuals with myopia have a 2- to 3-fold higher risk of glaucoma than nonmyopic individuals. ^1–3 Considering the increasing prevalence of myopia over the world, a correct diagnosis of glaucoma in myopic patients is becoming more important and necessary in ophthalmology. ^1,4,5  

However, it is difficult to assess the optic disc changes and retinal nerve fiber layer (RNFL) alterations accurately in myopic eyes, because the RNFL can be altered by the myopic changes in eyes without glaucoma. In addition, the effect of myopia alone on the peripapillary RNFL has not been determined definitively. 

Optical coherence tomography (OCT) is a noninvasive imaging method that can evaluate the morphology of the RNFL and optic disc with micrometer resolution. ^6,7 Several studies have documented that OCT can provide reliable measurements of the RNFL thickness. ^8–11 In addition, the recent advancement of OCT technology, for example, spectral-domain OCT, has allowed measurements that are significantly more accurate than the traditional time-domain OCT. ^12,13 However, the RNFL thickness measurements can be affected significantly by refractive errors, and the ability to detect glaucoma in highly myopic eyes by the RNFL thickness is inferior to that in emmetropic eyes. ¹⁴  

It has been reported that the thickness of the supratemporal and infratemporal RNFL bundles, and the position of the peak RNFL thickness are shifted in eyes with longer axial lengths (ALs). ^15,16 Thus, the supratemporal RNFL bundle is shifted closer to the fovea in highly myopic eyes than in low myopic eyes (Fig. 1). The shift in the RNFL bundle can be quantified by examining the position of the supratemporal and infratemporal peak RNFL thickness of the peripapillary RNFL thickness profile. ¹⁵ Because this shift may influence the peripapillary RNFL thickness, it is important to know the effect of the myopic changes on the peripapillary RNFL thickness for an accurate diagnosis of glaucoma. Hood et al. reported that the position of the supratemporal and infratemporal peak RNFL thickness was correlated significantly with the location of the main supratemporal and infratemporal arteries and veins. ^17,18 In addition, the location of the vessels influenced the shape of the OCT-determined RNFL thickness profile. Thus, the location of the vessels also may influence the peripapillary RNFL thickness. 

Figure 1

 

Red-free photographs of the fundus and retinal nerve fiber trajectory of a highly myopic eye (A) and low myopic eye (B).

Several studies have examined the association between the sectoral RNFL thickness and AL of the eye. ^15,19,20 However, to the best of our knowledge, no reports focused on the effect of a myopic shift in the peak RNFL position or vessel position and the sectoral RNFL thickness obtained by OCT. Moreover, the shift in the RNFL bundles in eyes with an elongation of the AL was not considered for the normative data embedded in SD-OCT instruments. 

Thus, the purpose of our study was to investigate the relationship between the position of the RNFL peak thickness and the position of the retinal vessels, the AL, and the sectoral RNFL thicknesses in young, healthy eyes of Japanese individuals. We determined what factors affected the sectoral RNFL thickness. 

Methods

All of the procedures used conformed to the tenets of the Declaration of Helsinki. A written informed consent was obtained from all 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 in the public domain 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. Originally, we studied 50 eyes of 50 individuals who were enrolled between November 1, 2010 and July 30, 2011. The primary outcome measures were the position of peak angle of the RNFL thickness, the sectoral RNFL thicknesses, and the AL. Volunteers with no known eye diseases as determined by examining their medical history were studied, and only the data from the right eyes were analyzed. The eligibility criteria were: age ≥ 20 but <40 years; eyes normal by slit–lamp biomicroscopy, ophthalmoscopy, and OCT; best corrected visual acuity (BCVA) was ≥0.1 logarithm of the minimum angle of resolution (logMAR) units; and the IOP was ≤21 mm Hg. The exclusion criteria were: eyes with known ocular diseases, such as glaucoma, staphyloma, and optic disc anomaly; known systemic diseases, such as hypertension and diabetes; presence of visual field defects; and history of refractive or intraocular surgery. None of the eyes was excluded because of poor OCT image quality caused by poor fixation. 

Measurement of AL and Refractive Error

All eyes had a standard ocular examination, including slit-lamp biomicroscopy of the anterior segment, ophthalmoscopy of the ocular fundus, IOP measured with a pneumotonometer (CT-80; Topcon, Tokyo, Japan), and the AL measured with the AL-2000 ultrasound instrument (TOMEY Corporation, Nagoya, Japan). The refractive error (spherical equivalent) was measured with the Topcon KR8800 autorefractometer/keratometer. 

Determination of Thickness of RNFL and Position of Peak Thickness and Retinal Vessels

All eyes were examined by a single examiner (TaY). The RNFL thickness was measured with the TOPCON 3D OCT-1000 MARK II using the RNFL 3.4 mm circle scan. In this protocol, 1024 A-scans/circle and 16 overlapping B-scans/image, and direct B-scan observations were made. The OCT images and color fundus photographs were taken at the same time. The OCT optical system detected the edge of the optic disc of the fundus image, and the scan circle was centered automatically on the optic disc just before taking the OCT image. To exclude the effects of errors in the scan circle centration, one examiner (YK) checked that the center of the scan circle was located at the center of the optic disc. 

The average thickness of the entire RNFL circle and at each clock hour was determined. Most OCT instruments display the RNFL profile in the following order: temporal, superior, nasal, inferior, and temporal, that is, the TSNIT order. However, the TOPCON 3D OCT-1000 MARK II displays the RNFL profile as the nasal, superior, temporal, inferior, and nasal, or the NSTIN order. The NSTIN thickness curves were used to measure the angle between supratemporal and infratemporal peaks of the RNFL. We determined this peak angle in the NSTIN thickness profile of the RNFL thickness analyses. The distance between the peak RNFL thickness for the supratemporal and infratemporal RNFL peaks was determined by dragging a vertical line in the profile graph in the Photoshop CS5 program (Adobe Systems, San Jose, CA). Then, the distance between supratemporal and infratemporal RNFL peaks (X) was converted to an angular value by dividing by the entire distance (Y) and multiplying by 360 (Fig. 2A). 

Figure 2

 

Measurement of the peak angle using the NSTIN profile (A). Measurement of the artery angle using the fundus photograph (B).

In the TOPCON 3D OCT-1000 MARK II, the color fundus photographs and OCT images were taken at the same time, and the scan circle was overlaid as a green circle in the photographs. Using the center of the green circle and points where the green circle and the center of supratemporal/infratemporal major retinal artery intersected (white dots), the angle between the supratemporal and infratemporal major retinal artery was measured (Fig. 2B). The angle between the supratemporal and infratemporal major retinal vein was measured by the same method. We named them the artery angle and the vein angle, respectively. 

Statistical Analyses

All statistical analyses were performed with the SPSS statistics 19 for Windows (SPSS Inc., Somers, NY). The relationships between the peak angle, and the AL and refractive error were investigated by linear regression analyses. The relationships between the sectoral RNFL thickness and the peak angle/artery angle/vein angle/AL were investigated using Pearson's correlation analysis. Partial correlations were calculated between the sectoral RNFL thickness and the peak angle/artery angle/AL because these values were correlated. 

Results

We screened 56 Japanese volunteers for this study. Three eyes were excluded due to ocular diseases; two cases of superior segmental optic hypoplasia (SSOH) and one case of post laser-assisted in situ keratomileusis. Three other eyes were excluded because of the difficulty in identifying the position of the peak RNFL thickness. As a result, the right eyes of 50 individuals (33 men and 17 women) were used for the analyses. 

The demographic information of the patients is presented in Table 1. The mean ± SD of the age was 25.5 ± 1.3 years, and the mean refractive error (spherical equivalent) was −4.31 ± 3.14 D. The mean AL was 25.5 ± 1.3 mm. The mean artery angle was 136.9 ± 22.1°, the mean vein angle was 140.6 ± 19.9°, and the mean peak angle was 125.3 ± 20.9°. The refractive error and AL were correlated significantly and negatively (R = −0.81, P < 0.001; Fig. 3A). 

Figure 3

 

Scatterplot of refractive error (spherical equivalent, [A]) and peak angle (B) against the AL. Scatterplot of the peak angle and AL against the artery angle (C, D) and the vein angle (E, F).

Table 1

 

Participants Data

	Mean ± SD	Range
Age, y	25.5 ± 3.7	22–38
Sex, M/F	33/17
Spherical equivalent, D	−4.31 ± 3.14	−10.25–4.50
AL, mm	25.5 ± 1.3	22.3–28.0
Artery angle, degrees	136.9 ± 22.1	88.0–186.0
Vein angle, degrees	140.6 ± 19.9	96.0–174.0
Peak angle, degrees	125.3 ± 20.9	77.3–165.1

Pearson's Correlation Coefficients Between Peak Angle, Artery Angle, Vein Angle, and AL

The peak angle of the RNFL thickness was correlated significantly and negatively with the AL (r = −0.49, P < 0.001). The peak angle decreased by 7.9°/1 mm increase in the AL (Fig. 3B). The artery angle was correlated significantly and positively with the peak angle (r = 0.92, P < 0.001; Fig. 3C). The artery angle was correlated significantly and negatively with the AL (r = −0.38, P = 0.006), and the artery angle decreased by 6.5°/1 mm increase in the AL (Fig. 3D). The vein angle was correlated significantly with the peak angle (r = 0.72, P < 0.001; Fig. 3E). The vein angle was correlated significantly and negatively with the AL (r = −0.36, P = 0.01), and the vein angle decreased by 5.5°/1 mm increase in the AL (Fig. 3F). The peak angle was correlated better with the artery angle than the vein angle. Therefore, the artery angle was used for the later analyses. 

Pearson's Correlation Coefficient and Partial Correlation Coefficient Between Peak Angle/Artery Angle and RNFL Thickness

The peak angle of the RNFL thickness was correlated significantly and positively with the sectoral RNFL thicknesses in sectors 1, 2, 4, 5, 6, and 12 (r = 0.36–0.71; P < 0.05), and significantly negatively associated in sectors 8, 9, 10, and 11 (r = −0.33 to −0.80, P < 0.05). The AL was correlated significantly and negatively with the sectoral RNFL thicknesses in sectors 1, 2, 3, 4, 5, 6, and 12 (r = −0.31 to −0.54, P < 0.05), and correlated significantly and positively in sectors 8, 9, and 10 (r = −0.34 to −0.36, P < 0.05). The entire RNFL thickness was not correlated significantly with the peak angle or AL. 

After excluding the effect of the AL, partial correlation analysis showed that the peak angle was correlated significantly and positively with the sectoral RNFL thicknesses in sectors 1, 2, 5, and 6 (r = 0.34–0.62, P < 0.05), and correlated significantly and negatively in sectors 8, 9, 10, and 11 (r = −0.31 to −0.77, P < 0.05). In contrast, after excluding the effect of the peak angle, the AL was associated significantly with the sectoral RNFL thicknesses in only sector 4 (r = −0.44, P = 0.001; Table 2). 

Table 2. 

 

The RNFLT Values, and Pearson's and Partial Correlation Coefficient Between Peak Angle/AL and an Independent Variable

Location	Sector	RNFLT		Pearson's Correlation Coefficient				Partial Correlation Coefficient
				RNFLT vs. Peak Angle		RNFLT vs. AL		RNFLT vs. Peak Angle, Control Variable = AL		RNFLT vs. AL, Control Variable = Peak Angle
		Mean ± SD	Range	R	P Value	R	P Value	R	P Value	R	P Value
Superior	1	126.6 ± 18.3	91–168	0.66	<0.001	−0.41	0.003	0.58	<0.001	−0.14	0.36
Nasal	2	94.2 ± 18.7	51–125	0.48	<0.001	−0.43	0.002	0.34	0.02	−0.25	0.08
Nasal	3	75.2 ± 14.9	44–116	0.18	0.21	−0.31	0.03	0.03	0.82	−0.26	0.07
Nasal	4	77.8 ± 16.6	45–118	0.36	0.01	−0.54	<0.001	0.13	0.39	−0.44	0.001
Inferior	5	100.9 ± 19.9	56–159	0.58	<0.001	−0.44	0.001	0.46	0.001	−0.22	0.13
Inferior	6	120.7 ± 32.4	66–218	0.71	<0.001	−0.50	<0.001	0.62	<0.001	−0.25	0.08
Inferior	7	148.1 ± 18.8	110–203	0.02	0.91	0.19	0.18	0.13	0.37	0.23	0.11
Temporal	8	102.3 ± 26.6	69–191	−0.80	<0.001	0.36	0.01	−0.77	<0.001	−0.06	0.67
Temporal	9	78.7 ± 17.1	54–130	−0.70	<0.001	0.34	0.02	−0.65	<0.001	−0.01	0.95
Temporal	10	112.2 ± 26.2	76–184	−0.70	<0.001	0.35	0.01	−0.64	<0.001	0.02	0.92
Superior	11	153.8 ± 22.3	106–198	−0.33	0.02	0.14	0.35	−0.31	0.03	−0.35	0.81
Superior	12	133.7 ± 22.8	86–188	0.38	0.007	−0.37	0.008	0.24	0.09	−0.23	0.12
Total	Whole	110.3 ± 10.0	89–133	0.12	0.41	−0.26	0.07	−0.01	0.95	−0.24	0.11

 RNFLT, RNFL thickness.

A similar tendency was seen in the artery angle and AL. The artery angle was correlated significantly and positively with the sectoral RNFL thicknesses in sectors 1, 2, 4, 5, 6, and 12 (r = 0.34–0.64, P < 0.05), and correlated significantly and negatively in sectors 8, 9, 10, and 11 (r = −0.37 to −0.78, P < 0.05). 

After excluding the effect of the AL, partial correlation analyses showed that the artery angle was correlated significantly and positively with the sectoral RNFL thicknesses in sectors 1, 2, 5, and 6 (r = 0.36–0.58, P < 0.05), and correlated significantly and negatively in sectors 8, 9, 10, and 11 (r = −0.34 to −0.74, P < 0.05). In contrast, after excluding the effect of the artery angle, the AL was correlated significantly and negatively with the sectoral RNFL thicknesses in sectors 2, 4, 5, and 6 (r = −0.30 to −0.46, P < 0.05; Table 3). 

Table 3. 

 

Pearson's and Partial Correlation Coefficient Between Artery Angle/AL and RNFLT Values

Location	Sector	Pearson's Correlation Coefficient		Partial Correlation Coefficient
		RNFLT vs. Artery Angle		RNFLT vs. Artery Angle, Control Variable = AL		RNFLT vs. AL , Control Variable = Artery Angle
		R	P Value	R	P Value	R	P Value
Superior	1	0.64	<0.001	0.58	<0.001	−0.23	0.11
Nasal	2	0.47	<0.001	0.36	0.01	−0.30	0.04
Nasal	3	0.17	0.25	0.05	0.72	−0.27	0.06
Nasal	4	0.36	0.01	0.19	0.18	−0.46	0.001
Inferior	5	0.53	<0.001	0.44	0.002	−0.30	0.04
Inferior	6	0.63	<0.001	0.55	<0.001	−0.37	0.01
Inferior	7	−0.05	0.72	0.02	0.87	0.19	0.20
Temporal	8	−0.78	<0.001	−0.74	<0.001	0.11	0.47
Temporal	9	−0.69	<0.001	−0.65	<0.001	0.11	0.46
Temporal	10	−0.68	<0.001	−0.63	<0.001	0.13	0.37
Superior	11	−0.37	0.01	−0.34	0.02	−0.01	0.97
Superior	12	0.34	0.02	0.23	0.11	−0.27	0.06
Total	Whole	0.07	0.63	−0.03	0.82	−0.25	0.08

Discussion

There were three main findings in our study. First, the supratemporal and infratemporal peaks of the RNFL thickness tended to shift toward the fovea as the AL increased. This is consistent with the report by Yoo et al. ¹⁵ When the AL increases, the RNFL bundles also are displaced toward the fovea relative to longer eyes (Fig. 1). These changes can be detected as a straightening and a shift toward the fovea of the RNFL bundles in the fundus photograph (two dimensions). This shift in the RNFL bundle may account for the RNFL defect often appearing at the paracentral area in the early stages of myopic glaucoma. 

Second, the position of the peak angle of the RNFL thickness was correlated better with the artery angle than with the vein angle with a high correlation coefficient of 0.92. This is consistent with the report by Hood et al. ^17,18 Our results indicated that the artery angle can be used as an alternative to the peak angle for the RNFL thickness assessments. In glaucomatous eyes, it is difficult to find the peak angle because the RNFL thickness is decreased. However, our results indicated that the location of peak angle can be determined by referring to the artery angle. This may provide important information in evaluating the RNFL thickness. When the artery angle differs from the peak angle in a glaucomatous eye, it is possible that the peak of the RNFL thickness might have been altered by the reduction of actual RNFL bundle thickness by glaucoma. This could be an important marker for evaluating the RNFL thickness in the early and middle stages of glaucoma. In addition, the artery angle can be obtained easily from the fundus photograph so that a shift of the RNFL thicknesses can be estimated from the fundus photograph without measuring the AL. Further longitudinal study of glaucomatous myopic eyes is required to establish these speculations. 

Third, we found significant correlations between the peak RNFL thickness angle and the temporal sectoral RNFL thickness (−0.70 to −0.80), and the supranasal and inferior sectoral RNFL thicknesses (0.38–0.71), which have not been reported to our knowledge. The R values between the peak angle and the sectoral RNFL thickness were higher than that between the AL and the sectoral RNFL thickness. In addition, the R value between the AL and the peak angle was not high (R = −0.49). These findings suggested that the AL and peak angle may influence the sectoral RNFL thickness independently. 

Therefore, we examined whether the peak angle or the AL was correlated significantly with the sectoral RNFL thickness using partial correlations. The peak angle, AL, and sectoral RNFL thicknesses were correlated significantly with each other. In such cases, we can determine the relationship between two variables while excluding the effect of a third variable using partial correlation analyses. After excluding the effect of the peak angle, the only significant correlation between the sectoral RNFL thickness and AL was for sector 4 (nasal). However, after excluding the effect of AL, there were significant correlations between the peak angle and temporal sectoral RNFL thickness (r = −0.77 to −0.64), and the supranasal and inferior sectoral RNFL thicknesses (r = 0.34–0.62, Fig. 4). A similar tendency was seen for the artery angle and AL (Fig. 5). These findings suggested that the peak and artery angles are factors that can be important for predicting the RNFL thickness after excluding the effect of the AL. 

Figure 4

 

Pearson's and partial correlation coefficients between the sectoral RNFL thickness, and the AL or peak angle. The significant correlated sectors are shown in gray.

Figure 5

 

Pearson's and partial correlation coefficient between the sectoral RNFL thickness, and the AL or artery angle. The significant correlated sectors are shown in gray.

The reason why the sectoral RNFL thickness is correlated better with the peak and artery angles than the AL might be because of the presence of paradoxical eyes. A paradoxical eye is one with a short AL with myopic fundus changes, for example, conus, elliptic optic disc, and smaller peak angles, or one with a longer AL with few myopic fundus changes. An earlier study showed that there were large individual variations in the AL at birth. ²¹ Thus, a longer AL does not necessarily mean that the eye is elongated. More specifically, even though the two eyes have the same AL as adults, if the AL was different at birth, the degree of elongation must have been different between these eyes during the growth period. This may affect the peak or artery angle. These findings suggested that the peak and artery angles should be analyzed independently in the evaluation of the effect of the AL on the sectoral RNFL thicknesses. 

Nevertheless, our findings should help clinicians understand the pattern of regional variations in the RNFL thickness in myopic eyes. The normative data embedded in the current OCT instruments do not include data of highly myopic eyes. For example, the mean of the refractive errors of the TOPCON 3D OCT-1000 Mark II is −0.66 ± +1.70 D, with a range of −5.75 to +2.88 D (Topcon, personal communication, 2012). The RNFL profile of a typical highly myopic eye is shown in Figure 6. The temporal RNFL was thicker than the normal range of the normative database, and the supranasal and inferior RNFLs were thinner than the normal range of the normative database because the normative database did not include highly myopic eyes. These normative data might cause a misdiagnosis or an underestimation of the changes. Thus in the early stage of glaucoma, the RNFL defects are more likely to be detected in the infratemporal and supratemporal sectors. ^22–24 However, in moderate to highly myopic eyes, the RNFL of the temporal quadrant is thicker than that in the normative database. Even though the actual RNFL thickness of the temporal quadrant is decreased in myopic glaucomatous eyes, the values still may be within the normal range of the embedded data. On the other hand, the RNFL of the inferior and supranasal areas in moderate to high myopic eyes is thinner than that of the normative database. This false positive may lead to a wrong diagnosis. Therefore, care must be taken in interpreting the RNFL thicknesses of highly myopic eyes. 

Figure 6

 

Peak angle is shown (white two-way arrow) in a fundus photograph (top). Bottom: Supratemporal and infratemporal peak RNFL position of typical high myopic eye (black arrows) and inbuilt normative database (green arrows).

It is widely accepted that the average circumferential RNFL becomes thinner as the AL elongates. ^19,20,25 However, no significant correlation was found between the RNFL thickness and AL. This could be because our cases were skewed toward myopia, that is, the average refractive error was −4.31 ± 3.14 D, compared to that of earlier studies. ^19,20,25 Our cases were volunteer Japanese students who are known to belong to the most myopic group in the world. Thus, our results might not necessarily hold for nonmyopic populations. An epidemiologic study should help generalize the present results to all populations. 

Another issue is the size of the scan circle for the RNFL thickness. The scan circle is projected larger than the actual circle after an elongation of the AL by the magnification effect. ²⁶ This might affect the RNFL thickness, peak angle, and artery angle. Unfortunately, it may not be possible to obtain the absolute RNFL thickness values corrected for AL for our OCT instrument because the distance effect is not known. However, the artery angle can be re-evaluated by adjusting the size of the scan circle, and the magnification effect of the scan circle can be calculated using Bennett's formula. ^26,27 Factor p is 2.74 in the fundus photographs of TOPCON 3D OCT-1000 Mark II with a 45° view (Topcon, personal communication, 2012). Based on Bennett's formula, the change in image size is approximately 3.58%/mm of AL. Using this factor, scan circles were rescaled to a reference AL of 24 mm, so that the vessel angles were measured at the same distance (in mm) from the disc center. Using this adjusted scan circle, we re-examined every artery angle and vein angle, and found that there were minor differences between the present (mean adjusted artery angle 138.1°, mean adjusted vein angle 141.8°) and the previous results (mean artery angle 136.9°, mean vein angle 140.6°). The difference was 1.2° for the artery and vein angles. The adjusted artery angle was correlated significantly with the AL with a coefficient of correlation of −0.34 and a P value of 0.015. Additionally, Pearson's correlation and partial correlation analysis showed there were no significant differences in the sectors of the adjusted artery angle and that of the original artery angle except in sector 12. These changes did not alter the major conclusions of our study. 

There are several limitations in this study. Many factors affect the RNFL thickness, for example, the size, shape, and torsion of the eye; tilt of the optic nerve head; and the relative location of the fovea relative to the optic disc. ²⁸ None of these was considered in our analyses. Only the effect of the eye torsion could be minimized by use of the angel between the supra- and infra-peak RNFL thickness or vessels. Many optical factors also can affect the RNFL thickness, for example, scan circle diameter, ²⁹ anterior segment power, ³⁰ and the scan angle of the optic nerve head. ³¹ We only considered the magnification effect in measuring the angle between vessels. How these other factors affect the relationship between sectoral RNFL thickness and AL or artery angle must be determined in future studies. 

In summary, our results indicated that the peak angles of the RNFL thicknesses and the artery angle are the major factors that affect the peripapillary RNFL thickness measured by OCT. The sectoral RNFL thickness is correlated better with the peak and artery angles than the AL. A shift in the RNFL peak may lead to incorrect interpretations of OCT values depending on the peripapillary location of the area of concern. The peak angle or the artery angle should be taken into account when determining the embedded RNFL thickness normative database for the OCT instrument to ensure more accurate detection of RNFL abnormalities. At present, special cautions are needed in the determination of RNFL thickness abnormalities using the current inbuilt RNFL normative database especially for moderate to high myopic eyes. 

Acknowledgments

Topcon Co., Ltd., Tokyo, Japan, provided technical advice. 

Supported in part by JSPS KAKENHI Grant Number 23791995. 

Disclosure: T. Yamashita, None; R. Asaoka, None; M. Tanaka, None; Y. Kii, None; T. Yamashita, None; K. Nakao, None; T. Sakamoto, None