Worldwide, myopia is among the most common ocular disorder with high prevalence in Asian population [
Optical coherence tomography angiography (OCTA) is a novel noninvasive technology that provides fast, depth-resolved visualization of the retinal and choroidal microvasculature [
However, it remains unknown whether there are significant sectional correlations between regional RNFL thickness, retinal vessel distribution, and eye structural parameters, since the previous OCTA studies about myopia only assessed the association between vascular and RNFL structure in an overall way. Because different eye diseases such as myopia and glaucoma often originate and exaggerate regionally with different patterns [
This cross-sectional, observational study was conducted in the Department of Ophthalmology, the Second Xiangya Hospital, Central South University. The study population consisted of healthy and myopia subjects recruited from October 2016 to January 2017. This study was conducted in accordance with the tenets of the Declaration of Helsinki (1964) and fully approved by the ethics committee of the Second Xiangya Hospital, Central South University. Informed consent as to the scientific objectives and process of the study was obtained from each subject.
Inclusion criteria were age of 18 years or more and presence of myopia in the studied eyes, as assessed by the refraction error and axial length. For each participant, one or both eyes meeting the criteria were included in the study. Participants had no known eye diseases as determined by full ophthalmic examinations.
The studied eyes were assigned to one of the four groups according to refraction: emmetropia (EM; mean spherical equivalent (MSE) 0.50 D to −0.50 D), mild myopia (MIM; MSE −0.75 D to −2.75 D), moderate myopia (MOM; MSE −3.00 D to −5.75 D), and high myopia (HM; MSE ≤ −6.00 D).
Exclusion criteria were any history of prior vitreous or retinal surgery or evidence of retinal diseases (other than myopic degeneration, such as age-related macular degeneration, macular hole, and foveal hypoplasia) affecting the retinal or choroidal vasculature by history or examination, presence of media opacities preventing reliable retinal thickness which prevents good visualization of the retinal structure, having systemic diseases such as glaucoma or diabetes mellitus which might affect the ocular circulation, and medication usage within 2 weeks of measurements.
All participants in the study underwent comprehensive ophthalmologic examination including spherical equivalent (SE) refraction measurement with an autorefractometer (KR-8800, Topcon, Tokyo, Japan), slit-lamp biomicroscopy, and fundus examination. An IOL Master (Carl Zeiss Inc., Jena, Germany) keratometer was used for measuring anterior chamber depth,
OCTA images were acquired with AngioVue (Optovue RTVue XR Avanti, Optovue Inc., Fremont, CA, USA) using automated segmentation algorithms. The system has an A-scan rate of 70 kHz scans per second, with a light source centered on 840 nm and a bandwidth of 45 nm. A 6 mm × 6 mm (36 mm2) area OCTA acquisition centered on the optic disc was performed, to record the overall and regional retinal RNFL thickness and vessel density. Three-dimensional (3D) OCTA scans were acquired by using two repeated B-scans at 304 raster positions, with each B-scan consisting of 304 A-scans. With a B-scan frame rate of 210 frames per second, each OCTA volume scan can be acquired in approximately 3 s. Two volumetric raster scans, including one horizontal priority (x-fast) and one vertical priority (y-fast), were obtained consecutively. The volumetric scans were processed by the split-spectrum amplitude-decorrelation angiography (SSADA) algorithm. All scans were reviewed by an examiner to ensure correct imaging and sufficient scan quality. Signal strength index (SSI) <45 and severe artifacts (double vessel pattern, loss of fixation, motion artifacts, and/or segmentation error resulting in poor outlining of vascular networks) in scans were the OCTA exclusion criteria. All of this processing can be achieved using the contained software (Optovue Inc., software V.2014.2.0.65).
Peripapillary retinal vessel density was imaged by OCTA with lateral fixation position of the subjects and quantified with the Angio Retina program. Measurements were automatically performed at four peripapillary quadrants: superior (S), inferior (I), nasal (N), and temporal (T) (Figure
Peripapillary RNFL thickness and vessel density at quadrants were identified by OCTA. The boundaries of segmentation are indicated by the red lines ((a) peripapillary RNFL thickness) or blue lines ((b) peripapillary vessel density). Peripapillary quadrants are identified as 6.0 mm radial scans 90° apart, around the central point of the optic disc. S, superior, radial scan from 315° to 45°; N, nasal, from 45° to 135°; I, inferior, from 135° to 225°; T, temporal, from 225° to 315°.
Raw data results were processed by statistical analysis software (IBM SPSS version 22.0 for Mac, SPSS Inc., Chicago, IL, USA). The Spearman rank correlation analysis was performed to determine the relationships between the overall and regional RNFL thickness, vessel density, and myopia-related eye structural parameters. Qualitative variables are presented as numbers and percentages. Quantitative variables are presented as means and standard deviations. Nonparametric ANOVA comparing the 4 tested eye groups used the nonparametric Kruskal-Wallis test. Qualitative variable (gender) was compared by the Fisher exact test.
A total of 56 healthy subjects (98 eyes) were included in this cross-sectional OCTA study. Demographic characteristics and eye structural measurements are shown in Table
Demographic and ocular characteristics of participants between the refractive groups.
Measurements | EM ( |
MIM ( |
MOM ( |
HM ( |
|
---|---|---|---|---|---|
Gender (male : female) | 8 : 9 | 9 : 18 | 16 : 17 | 9 : 12 | 0.671 |
Laterality (OD : OS) | 9 : 8 | 11 : 16 | 18 : 15 | 10 : 11 | 0.565 |
Age (years) | 23.9 ± 2.7 | 25.0 ± 3.3 | 22.6 ± 2.1 | 23.5 ± 4.9 | 0.113 |
MSE (dioptres) | 0.2853 ± 0.65 | −1.7893 ± 0.74 | −4.6379 ± 0.81 | −7.5976 ± 1.57 |
|
Axial length (mm) | 23.2 ± 0.57 | 24.2 ± 0.92 | 25.2 ± 0.84 | 26.5 ± 0.68 |
|
RNFL thickness ( |
|||||
Average | 109.8 ± 5.2 | 110.3 ± 6.2 | 105.5 ± 6.3 | 101.0 ± 8.9 |
|
S | 137.5 ± 10.7 | 140.0 ± 11.2 | 131.9 ± 13.9 | 124.2 ± 14.8 |
|
N | 81.4 ± 5.8 | 83.2 ± 8.7 | 79.4 ± 14.0 | 76.0 ± 16.5 | 0.211 |
I | 139.4 ± 10.4 | 140.3 ± 14.1 | 132.1 ± 11.3 | 123.1 ± 14.9 |
|
T | 80.4 ± 7.1 | 77.7 ± 12.0 | 79.1 ± 11.3 | 78.3 ± 16.2 | 0.919 |
Vessel density (%) | |||||
Average | 59.9 ± 3.5 | 61.3 ± 2.7 | 59.9 ± 3.9 | 58.7 ± 3.6 |
|
S | 59.7 ± 6.2 | 62.8 ± 3.7 | 62.4 ± 3.9 | 59.7 ± 6.3 |
|
N | 57.8 ± 3.8 | 57.7 ± 3.7 | 55.7 ± 4.4 | 53.9 ± 4.1 |
|
I | 60.9 ± 3.1 | 63.0 ± 3.0 | 61.5 ± 5.4 | 61.8 ± 3.7 | 0.206 |
T | 58.5 ± 4.9 | 61.7 ± 4.0 | 60.0 ± 5.2 | 59.3 ± 3.0 |
|
Numbers appear as mean ± SD. I: inferior quadrant; MSE: mean spherical equivalent; N: nasal quadrant; RNFL: retinal nerve fibre layer; S: superior quadrant; T: temporal quadrant
Generally, significant differences among the four groups were found at the superior and inferior quadrants of the RNFL thickness in the peripapillary area according to our OCTA results. At the inferior quadrant, the peripapillary RNFL thickness decreased as the myopia grew. The measurement was 139.4 ± 10.4
Overall, the highly myopic eyes had a lower peripapillary vessel density, as shown in Table
The results of the Spearman correlation analysis of the RNFL thickness at quadrants, AL, and SE are shown in Table
Correlation between overall and regional peripapillary RNFL thickness and myopic measurements.
AL | SE | |
---|---|---|
Average |
|
|
S |
|
|
N | −0.145 (0.154) |
|
I |
|
|
T | −0.007 (0.942) | −0.026 (0.798) |
Numbers appear as correlation coefficient (
Scatterplots illustrating the linear (black line) associations between axial length (mm) and OCTA peripapillary quadrant (superior, nasal, inferior, and temporal) RNFL thickness (
Scatterplots illustrating the linear (black line) associations between spherical equivalent (D) and optical coherence tomography angiography (OCTA) peripapillary quadrant (superior, nasal, inferior, and temporal) RNFL thickness (
Generally, there were significantly negative correlations between peripapillary vessel density and AL (
Correlation between overall and regional peripapillary vessel density and myopic measurements.
AL | SE | |
---|---|---|
Average |
|
0.152 (0.135) |
S | −0.121 (0.236) | 0.054 (0.600) |
N | −0.053 (0.607) | −0.047 (0.646) |
I |
|
|
T | −0.117 (0.252) | 0.093 (0.360) |
Numbers appear as correlation coefficient (
Scatterplots illustrating the linear (black line) associations between axial length (mm) and optical coherence tomography angiography (OCTA) peripapillary quadrant (superior, nasal, inferior, and temporal) vessel density (%) measurement of the studied eyes.
Scatterplots illustrating the linear (black line) associations between spherical equivalent (D) and optical coherence tomography angiography (OCTA) peripapillary quadrant (superior, nasal, inferior, and temporal) vessel density (%) measurement of the studied eyes.
In the present study, the graphical relationship between peripapillary RNFL and vasculatures with the stages of myopia has been quantified with the OCTA technique. It was shown that as myopia increases, the reduction of peripapillary RNFL thickness mainly occurred at the superior and inferior quadrants of the retina. Characterization of RNFL thickness in the myopic eyes with OCT-based techniques has been described in the literature (Table
Comparison of OCT-based RNFL thickness studies in the myopic eyes.
Division | Method | Conclusion | Reference |
---|---|---|---|
None | OCTA | Peripapillary RNFL thickness reduced significantly in high myopia compared to mild myopia. | Wang et al. 2016 [ |
12 o’clock | OCT | Peripapillary RNFL thickness was thinner at 1, 7, and 12 o’clock sectors in the highly myopic eyes than in the mild myopic eyes. | Leung et al. 2006 [ |
12 o’clock | Cirrus HD-OCT | RNFL thickness of the 1, 2, 5, 6, and 12 o’clock sectors was significantly thinner in moderate to high myopia than in mild myopia. | Seo et al. 2017 [ |
None | OCT-1 | Mean peripapillary RNFL thickness did not vary with myopic SE or axial length for any OCT scan diameter investigated. | Hoh et al. 2006 [ |
12 o’clock | Stratus OCT | The RNFL thickness in high myopia decreased significantly at 1, 5, 6, 7, 8, 9, 10, 11, and 12 o’clock. | Efendieva 2014 [ |
Quadrant/12 o’clock | Cirrus HD-OCT | Average and temporal RNFLs increased significantly as the AL increased. | Choi et al. 2014 [ |
Quadrant | SD-OCT | Global and the temporal RNFL were thicker in the myopia group. | AttaAllah et al. 2017 [ |
Six sectors | SD-OCT | RNFL thickness in children was not affected by myopia. | Goh et al. 2017 [ |
On the other hand, the study showed that only at the nasal quadrant, there was a statistically negative correlation between the peripapillary vessel density and AL in the studied groups, indicating that the elongation of the globe affected nasal retinal vasculature significantly. A wide range of previous studies has already shown a decreased peripapillary perfusion in the greater myopic eyes. Wang and colleagues in their vascular-related OCTA research showed a lower peripapillary retinal perfusion measurement, including retinal flow index and vessel density, with respect to emmetropic eyes in Chinese population [
With OCTA, Fan et al. studied 91 eyes from emmetropia to high myopia, including pathologic myopic eyes with peripheral retinal degeneration, and revealed that superficial and deep vascular density in the macula negatively correlated with AL and the degree of myopia, but positively correlated with ganglion cell complex (GCC) thickness [
Comparison of OCTA-based fundus vasculature studies in the myopic eyes.
Division | Method | Conclusion | Reference |
---|---|---|---|
None | OCTA | Peripapillary retinal perfusion (flow index and vessel density) was lower in higher myopia than emmetropia. | Wang et al. 2016 [ |
None | OCTA | Peripapillary retinal perfusion (flow index and vessel density) reduced in high myopia with tessellated fundus with respect to the control eyes. | Wang et al. 2016 [ |
Six sectors | OCTA | Peripapillary, IT, and ST vessel density in the highly myopic eyes with PICC was lower than the density in those without. | Chen et al. 2017 [ |
None | OCTA | Macular but not optic disc superficial and deep vascular density reduced as AL increased in the myopic eyes. | Fan et al. 2017 [ |
None | OCTA | Retinal capillary density reduced while CC flow deficit increased in greater myopia. | Al-Sheikh et al. 2017 [ |
None | OCTA | Retinal microvascular network alterations in the highly myopic eyes correlates with axial length elongation. | Yang et al. 2016 [ |
None | OCTA | Retinal microvascular decreased in the high myopia subjects with unchanged retinal microvessel blood flow velocity. | Li et al. 2017 [ |
None | OCTA | SE and AL influence the size of the foveal avascular zone. | Tan et al. 2016 [ |
CC: choriocapillaris; PICC: peripapillary intrachoroidal cavitation.
Our findings agree with these previous descriptions about peripapillary vessel density and go further to illustrate that the vasculature at the nasal quadrant of the retina may be more vulnerable to myopia-related structural change than that at other quadrants. However, the clear mechanism of these graphical correlations in myopia disease development remained unknown. Previous researchers have stated one possible hypothesis that thinning of peripapillary RNFL may affect the vasculature network via autoregulatory mechanisms [
The current study had several limitations. First, the cross-sectional design of the study precluded a causative conclusion that AL elongation in myopia induced the thinning of RNFL and vascular density at quadrants in the peripapillary area. Second, the sample size in the present study was relatively small and from the same race within a narrow spectrum of age, mainly in twenties and thirties. Thus, the study is an exploratory and descriptive analysis, and its conclusions may not be applicable to the other races or age spectrums, for example, Caucasian, African, children, or elderly. Third, we did not adjust the confounding factors such as age, which might affect the accuracy of statistical analysis. Fourth, since the Angio Retina program used in this OCTA study divided the peripapillary RNFL and vessel density into four quadrants only (superior, inferior, nasal, and temporal), further researches with more detailed division and new generation of OCTA analyzing program are required to investigate the structural changes in the myopic eyes from the sectional perspective.
OCTA as a novel, high-speed, and noninvasive imaging technique to detect blood flow signals in the retina and the choroid has advantages over the traditional angiography techniques. As shown in Table
In conclusion, using OCT angiography, we have found that peripapillary RNFL thickness decreased significantly at the superior and inferior quadrants as the myopia increased, while there was a negative correlation between the peripapillary vessel density at the nasal quadrant and AL in the myopic eyes. Measuring changes in RNFL structure and vascular density with OCTA could be a useful method to assess and follow the severity of disease in myopic subjects.
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or nonfinancial interest in the subject matter or materials discussed in this manuscript.
The authors thank all the participants of the study. Yuanjun Li was supported by the HKUST/CSU Joint MD-PhD Program. This work was supported by the Natural Science Foundation of China (NSFC 81770930 and NSFC 81371012 to Bing Jiang), the Department of Science and Technology of Hunan (No. 2015TP2007), and the Natural Science Foundation of Hunan Province (2017jj2360 to Bing Jiang).