Visual Acuity in Pathological Myopia Is Correlated with the Photoreceptor Myoid and Ellipsoid Zone Thickness and Affected by Choroid Thickness

Home , Emmetropia

Retina Visual Acuity in Pathological Myopia Is Correlated With the Photoreceptor Myoid and Ellipsoid Zone Thickness and Affected by Choroid Thickness

Jie Ye, Meixiao Shen, Shenghai Huang, Yuchen Fan, Aixia Yao, Chen Pan, Xiutong Shi, Fan Lu, and Yilei Shao School of Ophthalmology and Optometry, Wenzhou Medical University, Wenzhou, Zhejiang, China

Correspondence: Yilei Shao, School PURPOSE. To quantify the thickness of the outer retinal sublayers and choroid in pathological of Ophthalmology and Optometry, myopia and examine associations between these factors and best-corrected visual acuity Wenzhou Medical University, 270 (BCVA). Xueyuan Road, Wenzhou, Zhejiang 325027, China; METHODS. The cohort was composed of 21 eyes with emmetropia and 70 eyes with high [email protected]. myopia (49 simple high myopia; 21 pathological myopia). Optical coherence tomography Fan Lu, School of Ophthalmology images were segmented to determine macular thicknesses of the choroid and the following and Optometry, Wenzhou Medical outer retinal sublayers: outer plexiform layer (OPL), Henle ﬁber layer and outer nuclear layer University, 270 Xueyuan Road, (HFL þ ONL), myoid and ellipsoid zone (MEZ), outer segment of photoreceptors (OS), and Wenzhou, Zhejiang 325027, China; interdigitation zone and RPE/Bruch complex (IZ þ RPE). Correlations between BCVA and lufan62@mail.eye.ac.cn. thickness of the outer retinal sublayers and choroid were determined. JY and MS contributed equally to the RESULTS. In pathological myopia, the choroid, HFL þ ONL, MEZ, and IZ þ RPE were thinner work presented here and should than in emmetropia and simple high myopia (P < 0.05). Simple and multiple regression therefore be regarded as equivalent authors. models showed that MEZ thickness was correlated with BCVA (both P < 0.001). The relationship between MEZ thickness and BCVA varied with choroidal thickness (P ¼ 0.006). Submitted: October 30, 2018 For a constant MEZ thickness, thinner choroids were associated with worse vision. In the ﬁnal Accepted: March 25, 2019 multiple regression predictive model, MEZ thickness, choroidal thickness, and interaction Citation: Ye J, Shen M, Huang S, et al. between MEZ and choroidal thickness (all P < 0.001) were predictors of BCVA. Visual acuity in pathological myopia is correlated with the photoreceptor CONCLUSIONS. Outer retinal alterations, especially thinning of the MEZ, occurred in myoid and ellipsoid zone thickness pathological myopia. The MEZ thickness was associated with BCVA, and this relationship and affected by choroid thickness. was affected by choroidal thickness. Invest Ophthalmol Vis Sci. Keywords: pathological myopia, visual acuity, photoreceptor dystrophy, choroid 2019;60:1714–1723. https://doi.org/ 10.1167/iovs.18-26086

rreversible visual impairment and blindness due to high lial growth factor treatment in patients with myopic neovascu- I myopia is one of the most serious worldwide vision larization.9 These studies significantly improved our problems, especially in Asia.1–3 Holden et al.4 documented understanding of pathogenic mechanisms leading to vision loss that the prevalence of myopia has increased drastically and in pathological myopia. Compared to qualitative observations, predicted that by 2050 there would be nearly 1 billion people quantitative analysis would provide much more detailed 5 with high myopia. Moreover, Liu et al. estimated that nearly information regarding small and seemingly minor changes, two of every three (65.4%) highly myopic patients develop enabling early detection of photoreceptor degeneration. pathological myopia. Further, visual impairment, with the best- Investigating the minor anatomic abnormalities of the outer corrected visual acuity (BCVA) worse than 20/60, has been 5 retinal sublayers and their associations with the BCVA may identified in 30.8% of patients with pathological myopia. provide clues to the development of vision-threatening Therefore, to develop new preventive and therapeutic strate- alterations and the pathophysiological mechanisms in patho- gies, it is essential to understand the risk factors and logical myopia. pathogenesis of visual impairment in pathological myopia. The photoreceptor layer in the outer retina provides spatial In the current study, we applied a segmentation algorithm to information during the first stage of visual processing. Much of OCT images to identify the sublayers of the outer retina and the our understanding concerning the relationship between the choroid. The goals of the current study were to investigate the disruption of photoreceptors and visual function in high changes in the thickness of the outer retinal sublayers and myopia has been obtained from qualitative observations by choroid in pathological myopia and then to develop a visual using optical coherence tomography (OCT).6–8 For example, function predictive model to determine the significant variables vision was poorer in eyes with foveal ellipsoid zone disruption that would affect BCVA. This model may deepen our and in eyes with inner and outer segment defects. The integrity understanding of the mechanism of visual impairment and of the ellipsoid zone at baseline was also one of the factors has the potential to guide us to develop new preventive and predicting better final visual acuity after anti-vascular endothe- therapeutic strategies in pathological myopia.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1715

METHODS nuclear layer (HFL þ ONL), (3) myoid and ellipsoid zone (MEZ), (4) the outer segment of photoreceptors (OS), (5) the Subjects interdigitation zone and retinal pigment epithelium/Bruch complex (IZ þ RPE), and (6) the choroidal layer (Fig. 1B). In this prospective, cross-sectional study, all subjects were The custom software for segmented image analysis was recruited from August 2017 to May 2018 at The Affiliated Eye developed based on gradient information and the shortest Hospital of Wenzhou Medical University, Wenzhou, Zhejiang, path search, as our previous studies described.11,12 For the China. This project was approved by the Ethics Committee of automated segmentation, each image was visually inspected Wenzhou Medical University and performed in accordance after the segmentation algorithm was run. The standard of the with the tenets of the Declaration of Helsinki. All participating segmentation for each layer was according to the definitions patients provided written informed consent. proposed and adopted by the International Nomenclature for The enrolled subjects were divided into the following three Optical Coherence Tomography Panel.13 The criteria used to groups: (1) emmetropia subjects with a spherical equivalent identify segmentation errors were reported in our previous (SE) of 1.0 diopter (D) to þ 0.5 D; (2) simple high myopia papers and included small peaks and curve offsets.14,15 If the subjects with SE 6.0 D or axial length (AL) ‡ 26.5 mm, automated segmentation was wrong in a few images, which without any pathological changes; and (3) pathological myopia was more likely to occur at the posterior boundary of the subjects with SE 6.0 D or AL ‡ 26.5 mm, with pathological choroid in emmetropic eyes, manual corrections were made. changes. According to the International Meta-Analysis for All analyses of the segmentation were done by one masked Pathologic Myopia (META-PM) classification system, eyes with reader.11,12 diffuse or severe atrophy were classified as pathological We used Bennett’s formula, t ¼ p 3 q 3 s (t as the real scan myopic eyes.10 Diffuse and severe atrophy includes eyes with length, p as the magnification factor determined by the OCT a yellowish-white appearance of the posterior pole, or with imaging system camera, q as the magnification factor related to well-defined, grayish-white lesions in the macular area and the the eye, s as the original measurement from the OCT image), to optic disc, even in the foveal region. The diagnosis of adjust the image magnification based on the AL. The correction emmetropia, simple high myopia, and pathological myopia factor q was determined by the formula q ¼ 0.01306 3 (AL was determined by two ophthalmologists. If the results from 1.82). When imaging the eye with an AL of 24.46 mm, the these two ophthalmologists were not consistent with each actual scanning range t would be equal to the s. When imaging other, a senior ophthalmologist gave the final judgment. These eyes with other ALs, the real scan length t was determined by three ophthalmologists all came from the High Myopia the equation t ¼ (AL 1.82) / 22.64 3 s. Department of the Affiliated Eye Hospital of Wenzhou Medical To evaluate measurement repeatability, two separate sets of University, Wenzhou, Zhejiang, China. Eyes with intraocular images collected from 14 eyes from each of the three groups pressure (IOP) more than 21 mm Hg, visual field defects, were measured and the global macular thicknesses of the outer history of intraocular surgery or related systemic diseases, or retinal sublayers and choroid were compared. The intraclass complications of high myopia such as retinoschisis and correlation (ICC) and the percentage of coefficient of choroidal neovascularization were excluded. repeatability (CoR%) were calculated. The CoR% was defined as standard deviation of the difference between two sets of Clinical Examinations scanned image measurements divided by the average of the All subjects underwent comprehensive clinical examinations, two repeated measurements. Bland-Altman plots were also including refraction, BCVA measurement as the logarithm of used to assess the agreement between the two repeated the minimum angle of resolution (logMAR), and slit-lamp measurements. biomicroscopy. AL was measured by the IOL Master (Carl Zeiss, Jena, Germany), and noncontact IOP was measured by the Full Statistical Analyses Auto Tonometer TX-F (Topcon, Tokyo, Japan). Fundus All data were calculated as means 6 standard deviations and photography was performed with a 458 digital retinal camera were analyzed with SPSS software (version 22.0; SPSS, Inc., (Canon EOS 10D SLR backing; Canon, Inc., Tokyo, Japan). Chicago, IL, USA). The SE of refractive error was calculated as the spherical dioptric power plus one-half of the cylindrical Image Acquisition Protocol and Analysis dioptric power. One-way analysis of variance (ANOVA) was All enrolled patients were imaged by an OCT system (Optovue used to compare the differences among the three groups, and RTVue XR Avanti; Optovue, Inc., Fremont, CA, USA). The scan post hoc procedures were used to compare differences between groups. The different frequencies of each sex among speed was 70,000 A-scans per second with 5-lmaxial 2 resolution. The 8-mm radial line scan protocol was applied. the three groups were tested by the v test. Simple regression Eighteen consecutive B-scan images were produced, and each models (based on the generalized estimating equations, GEE) scan was centered on the fovea. A good set of scans with a were used to analyze the associations and interactions of the signal strength index of more than 40 was selected for further imaged outer retinal sublayer thicknesses and other variables analysis. with the BCVA. Based on the GEE, results from the simple Following image collection, we divided the macula into regression models and the models assessing interactions were three subfields (Fig. 1A): (1) the central foveal region with a then used to create a final multivariate model with BCVA as the diameter of 1 mm, (2) the parafoveal region ranging from 1 to 3 outcome. A value of P < 0.05 was considered statistically mm from the central foveal region, and (3) the perifoveal significant. region ranging from 3 to 6 mm from the central foveal region. Global thickness measurements of the whole macula included all three subfields. Custom-developed software that included RESULTS correction of the image magnification based on the AL was Patient Characteristics used to quantify the thickness of the following sublayers of the outer retina and the choroid in the fundus11,12: (1) the outer Of the 91 eyes analyzed, 21 were emmetropic, and 70 had high plexiform layer (OPL), (2) the Henle fiber layer and outer myopia, including 49 with simple high myopia and 21 with

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1716

FIGURE 1. Macular regions analyzed by optical coherence tomography in the 18-radial line scanning mode. (A) The macula was divided into three subﬁelds: the central foveal region with the diameter of 1 mm, the parafoveal region ranging from 1 to 3 mm from the central foveal region, and the perifoveal region ranging from 3 to 6 mm from the central foveal region. (B) The boundaries of fundus structure were segmented by an automated algorithm, and the thickness proﬁles of the macular outer retinal sublayers and choroid were determined. Bars: 300 lm.

pathological myopia (Table 1). Pathological myopic eyes had emmetropia group were thicker than in the simple high worse BCVA, greater myopia, and longer ALs than did myopia and pathological myopia groups (P < 0.01 each). The emmetropic and simple high myopic eyes. There were no global OS layer of the emmetropia and pathological myopia significant differences among emmetropia, simple high myo- groups was thinner compared to the simple high myopia group pia, and pathological myopia patients in age, sex, or IOP (P ¼ (P ¼ 0.001 and 0.040, respectively). The global MEZ thickness 0.137, 0.329, and 0.206, respectively, Table 1). of the emmetropia group was not significantly greater than that of the simple high myopia group (P ¼ 0.640), but it was greater Thickness Differences of the Outer Retinal than that of the pathological myopia group (P ¼ 0.014). The global OPL was not significantly different among the three Sublayers and Choroid groups (ANOVA, P ¼ 0.140). Fundus photographs (Figs. 2A–C) and OCT images (Figs. 2D–F) In comparisons between the simple high myopia and with the detailed automated segmentation were acquired for pathological myopia groups (Table 3), all of the outer retinal emmetropic, simple high myopic, and pathological myopic sublayers except the OPL were significantly thinner in the eyes. For the two repeated measurements, the ICC for the pathological myopia group. For the OPL, the thickness was thickness of the outer retinal sublayers and choroid in the marginally greater in the pathological myopia group (P ¼ 0.05). three groups varied from 0.875 to 0.998, and the CoR% varied Comparisons of the outer retinal sublayers and choroid in the from 0.59 to 5.61 (Table 2). Bland-Altman analysis of the global central, parafoveal, and perifoveal regions followed similar, thicknesses for the outer retinal sublayers and choroid in all though not necessarily identical, patterns as seen for the global groups showed that most of the differences between the comparison (Supplementary Table S1). means for the two repeated measurements were within the limits of agreement (Fig. 3). Association of Outer Retinal Sublayer Thicknesses There were significant differences among the three groups and Other Variables With Best-Corrected Visual in the thicknesses of the outer retinal sublayers and the Acuity choroid along the 6-mm diameter of the macula centered on the fovea (Table 3; Supplementary Table S1; Fig. 4). The global Simple GEE-based regression models with BCVA as the choroidal thickness of the emmetropia group was larger than outcome were constructed (Table 4). Sublayer thicknesses that of the simple high myopia and pathological myopia groups were based on the global values. Older age (P ¼ 0.004), longer (P < 0.001 each). The global IZ þ RPE and HFL þ ONL of the AL (P < 0.001), thicker OPL (P ¼ 0.008), thinner HFL þ ONL (P

TABLE 1. Basic Characteristics of the Emmetropia, Simple High Myopia, and Pathological Myopia Groups

Characteristics Emmetropia Simple High Myopia Pathological Myopia P* P1 P2 P3

N 2149 21–––– Age, y 37 6 13 32 6 9376 11 0.137 0.399 1.000 0.324 Sex, F:M 15:6 31:18 17:4 0.329 0.510 0.469 0.144 SE, diopter 0.35 6 0.46 10.2 6 2.64 14.0 6 3.17 <0.001 0.009 <0.001 0.001 BCVA, logMAR 0.0 6 0.04 0.02 6 0.04 0.19 6 0.20 <0.001 0.009 <0.001 0.004 AL, mm 24.04 6 0.75 27.64 6 1.15 29.28 6 1.21 <0.001 <0.001 <0.001 <0.001 IOP, mm Hg 13.89 6 4.44 15.78 6 2.70 14.81 6 4.09 0.206 0.091 0.460 0.319 F, female; M, male. * P value among the three groups; P1, P value between emmetropia and simple high myopia; P2, P value between emmetropia and pathological myopia; P3, P value between simple high myopia and pathological myopia.

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1717

FIGURE 2. Representative fundus photographs and OCT images of emmetropia, simple high myopia, and pathological myopia eyes. Fundus photographs (A–C) and OCT B-scan images (D–F) with the detailed automated segmentation of emmetropia, simple high myopia, and pathological myopia eyes, respectively. For the emmetropic eye (A, D), the AL was 24.21 mm, the global macular MEZ thickness was 26.34 lm, and the global macular choroid thickness was 347.91 lm. For the simple high myopia eye (B, E), the AL was 27.24 mm, the global macular MEZ thickness was 25.75 lm, and the global macular choroid thickness was 100.94 lm. For the pathological myopia eye (C, F), the AL was 29.37 mm, the global macular MEZ thickness was 20.24 lm, and the global macular choroid thickness was 68.67 lm. Bars: 300 lm.

¼ 0.001), thinner MEZ (P < 0.001), thinner IZ þ RPE (P < Predictive Models of BCVA Using MEZ Thickness 0.001), and thinner choroid (P < 0.001) were significantly and Other Variables associated with worse BCVA. Sex (P ¼ 0.138) and OS thickness (P ¼ 0.813) were not correlated with BCVA. MEZ thickness had Because age, AL, and thickness of the OPL, HFL þ ONL, MEZ, IZ the greatest correlation coefficient among all of the fundus þ RPE, and choroid were significantly associated with BCVA in thickness variables. the simple regression models (Table 4), they were included in the final predictive model (Table 6). The interaction between Interactions Between MEZ Thickness and Other the MEZ and choroidal thickness was also used in the final predictive model. The MEZ thickness (P < 0.001), choroidal Variables thickness (P < 0.001), and the interaction between the MEZ Multiple regression models explored the interactions between and choroidal thickness (P < 0.001) were significant predictors the MEZ and other variables with BCVA as the outcome (Table of the BCVA. Age, AL, OPL, HFL þ ONL, and IZ þ RPE 5). The only significant interaction was between the MEZ and thicknesses were not significant predictors of BCVA (P ¼ 0.064 choroidal thickness (P ¼ 0.006). Interactions between MEZ ~ 0.963) in this final predictive model. thickness and age, AL, or IZ þ RPE thickness were not The final regression model demonstrated how the relation- significant (P ¼ 0.283, 0.904, and 0.446, respectively). ship between MEZ thickness and BCVA varied with choroidal thickness (Fig. 5). Analysis of the receiver operating charac- teristic (ROC) curve was used to calculate the choroidal TABLE 2. Repeatability of Global Thickness Measurements for the thickness cutoff value that determined visual impairment, Outer Retinal Sublayers and Choroid in the Emmetropia, Simple High Myopia, and Pathological Myopia Groups defined as BCVA (logMAR) worse than 0.10. The choroidal thickness cutoff value for visual impairment was 109.77 lm Emmetropia, Simple High Pathological with an area under the ROC curve of 0.928 (Fig. 6). The n ¼ 14 Myopia, n ¼ 14 Myopia, n ¼ 14 correlation coefficient r between the MEZ thickness and BCVA for the entire cohort was 0.439 (P < 0.001). For the group Parameters ICC CoR% ICC CoR% ICC CoR% above the cutoff value for choroidal thickness, the correlation coefficient was 0.409 ( 0.001), and for the group below OPL 0.927 5.61 0.875 2.30 0.959 2.88 r P ¼ the cutoff value, it was 0.468 (P ¼ 0.021). This provides HFL þ ONL 0.984 1.52 0.984 1.58 0.990 1.88 further evidence that the relationship between the MEZ MEZ 0.983 1.62 0.992 0.91 0.988 1.21 thickness and BCVA varied with choroidal thickness. Pearson’s OS 0.981 2.36 0.991 0.59 0.994 1.69 correlation analysis also showed that the MEZ thickness was IZ þ RPE 0.978 3.41 0.987 1.27 0.972 2.00 Choroid 0.972 3.86 0.998 1.30 0.998 2.20 significantly correlated with the choroid thickness (r ¼ 0.403, P < 0.001).

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1718

FIGURE 3. Bland-Altman analysis for two repeated measurements of the global thickness of the outer retinal sublayers and choroid in all groups (including emmetropia, simple high myopia, and pathological myopia). (A) Outer plexiform layer; (B) Henle ﬁber layer and outer nuclear layer; (C) myoid and ellipsoid zone; (D) outer segment of photoreceptors; (E) interdigitation zone and retinal pigment epithelium/Bruch complex; (F) choroidal layer. Values on the horizontal and vertical axes correspond to the mean and to the difference of the two repeated measurements, respectively. Solid lines and dashed lines indicate mean differences and limits of agreement. Error bars indicate 95% conﬁdence interval of the mean difference and limits of agreement.

DISCUSSION of the intraretinal layer and choroid thicknesses in high myopia (Table 7), and our ﬁndings were consistent with those reports. We used OCT to evaluate thickness changes in the outer retinal sublayers and in the choroid of eyes with simple high myopia However, few of the previously published papers differentiated and pathological myopia. Previous papers reported alterations between simple high myopia and pathological myopia. To our

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1719

FIGURE 4. Box and whisker plots of global thickness comparisons of the outer retinal sublayers and choroid among the three groups. (A) Outer plexiform layer; (B) Henle ﬁber layer and outer nuclear layer; (C) myoid and ellipsoid zone; (D) outer segment of photoreceptors; (E) interdigitation zone and retinal pigment epithelium/Bruch complex; (F) choroidal layer. The median is represented by the middle line within each box, and the second and third quartiles are represented by the lower and upper segments of the box, respectively. The whiskers of the plot represent the minimum (bottom whisker) and the maximum (top whisker).

knowledge, this is the first report of changes in the thickness of visual acuities for a given MEZ thickness. Therefore, preventing outer retinal sublayers, including the photoreceptor layer, in further degeneration of the MEZ should be an important pathological myopia. The MEZ thickness was significantly clinical goal for pathological myopia eyes. correlated with BCVA. Further, the relationship between MEZ Previous studies only evaluated the disruption of photore- thickness and BCVA was influenced by choroidal thickness, ceptors in high myopia qualitatively to determine the influence such that patients with relatively thin choroids had poorer on visual acuity.6–8 Disruption of the foveal ellipsoid zone and

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1720

TABLE 3. Global Macular Thicknesses (lm) of the Outer Retinal Sublayers and Comparisons Among the Emmetropia, Simple High Myopia, and Pathological Myopia Groups

Regions Emmetropia Simple High Myopia Pathological Myopia P* P1 P2 P3

Global OPL 17.79 6 2.77 17.61 6 1.93 18.73 6 2.01 0.140 0.742 0.165 0.050 HFL þ ONL 62.66 6 5.10 57.88 6 5.99 54.28 6 6.38 <0.001 0.003 <0.001 0.022 MEZ 23.02 6 2.12 22.76 6 2.29 21.34 6 1.88 0.022 0.640 0.014 0.014 OS 27.72 6 3.41 30.27 6 2.76 28.69 6 2.65 0.003 0.001 0.279 0.040 IZ þ RPE 27.17 6 3.93 23.27 6 3.30 20.49 6 1.17 <0.001 0.001 <0.001 <0.001 Choroid 243.40 6 51.63 158.02 6 43.67 84.30 6 22.55 <0.001 <0.001 <0.001 <0.001

* P value among the three groups; P1, P value between emmetropia and simple high myopia; P2, P value between emmetropia and pathological myopia; P3, P value between simple high myopia and pathological myopia.

defects in the inner and outer segments were found in myopic changes will be further studied by collecting larger sample eyes with poor vision.6,7 In the current study, we quantified sizes to validate this hypothesis. the thickness of photoreceptor sublayers in high myopia with The MEZ is a specialized region in the photoreceptors that and without macular pathological change. The high repeat- contains the mitochondria, Golgi, and endoplasmic reticulum, ability of the outer retinal sublayer and choroidal thickness so it is important with regard to the production of ATP, G- measurements shows that the OCT images can offer sufficient proteins, photopsin, and other chemicals16,17 that are impor- precision. The quantitative methods used in this study enabled tant to maintain photosensitivity. In simple high myopia detection of small changes in the outer retinal sublayers. Thus, without pathological change, the MEZ layer was slightly the measurement of changes in photoreceptor sublayer thinner than in emmetropic eyes, but the difference was not thickness might be a valuable tool to provide insight into the significant. This might explain why serious visual impairment origin of visual loss in pathological myopia. always occurred in pathological myopia but not in simple high Thinning of the MEZ layer in pathological myopia was myopia. Therefore, along with BCVA, MEZ thickness can be a measured with high repeatability; that is, the ICC ranged from suitable biomarker in the clinic for pathological myopia. 0.983 to 0.992 and the CoR% ranged from 0.91 to 1.62. Further longitudinal studies are required to determine if early Importantly, MEZ thickness was the most relevant fundus detection of MEZ thickness alteration would be a significant microstructural factor related to visual impairment. The predictor of visual acuity during the progression of patholog- difference in AL between the emmetropia and simple high ical myopia. myopia groups was 3.60 mm (P < 0.001, Table 1), while the In this study, MEZ thicknesses, together with other difference in the global MEZ thickness for these two groups, important clinical variables such as age and AL and OCT 0.26 lm, was not significant (P ¼ 0.640, Table 3). Similarly, the anatomic parameters such as the thickness of the choroid and difference in AL between the simple high myopia and the IZ þ RPE, were investigated to determine if they were pathological myopia groups, 1.64 mm, was significant (P < correlated in any way with visual acuity in high myopia. The 0.001, Table 1). However, in contrast to difference in MEZ association between MEZ thickness and visual acuity was thickness between the emmetropic and simple high myopic independent of age, AL, and RPE thickness. The statistical eyes, the difference in MEZ thickness between the simple high interaction between MEZ and choroidal thickness was myopic and pathological myopic eyes, 1.42 lm, was significant significant, indicating that choroidal perfusion affects the (P < 0.014, Table 3). These results suggest that the significant relationship between MEZ and visual function in high myopia. thinning of MEZ occurred during the stage of pathological For a given MEZ thickness, worse visual acuity occurred in eyes retinopathy and is unlikely to be due simply to retinal with a thinner choroid. It is well known that choroidal thinning is significantly associated with visual impairment in high stretching by axial elongation. However, there might be a 18–20 critical AL beyond which the retina and choroid are myopia and other ocular diseases. Our current study ‘‘overstretched,’’ triggering the pathological myopia and confirmed that MEZ thinning was correlated with visual acuity resulting in the changes in the MEZ thickness. In our future as well. Further analysis showed that MEZ thinning was study, whether there was a critical AL to trigger pathological correlated with the choroidal thinning, indicating that thinning of the MEZ might be due to the lack of oxygen and nutrition from the choroid, which is indispensable for MEZ metabo- TABLE 4. Simple Regression Models Based on Best-Corrected Visual lism.21,22 Additional studies are required to research the Acuity Outcome cellular mechanism of the interaction between MEZ and 95% CI choroidal perfusion in pathological myopia. Standard Parameters Coefficient Error Lower Upper P Value TABLE 5. Multiple Regression Models for Interactions Between the Age, y 0.003 0.001 0.001 0.006 0.004 MEZ and Other Variables Based on Best-Corrected Visual Acuity Outcome Sex 0.044 0.029 0.101 0.014 0.138 AL, mm 0.030 0.006 0.019 0.041 <0.001 Standard OPL, lm 0.016 0.006 0.004 0.028 0.008 Coefficient, Error, P 95% CI HFL þ ONL, lm 0.007 0.002 0.011 0.003 0.001 Parameters 3105 3105 Value (3105) MEZ, lm 0.026 0.006 0.037 0.015 <0.001 OS, lm 0.001 0.004 0.010 0.008 0.813 MEZ 3 age 5.91 5.51 0.283 (4.89, 16.71) IZ þ RPE, lm 0.013 0.003 0.019 0.006 <0.001 MEZ 3 AL 2.41 19.96 0.904 (36.71, 41.53) Choroid, lm 0.001 0.0002 0.001 0.001 <0.001 MEZ 3 (IZ þ RPE) 12.80 16.81 0.446 (45.74, 20.14) MEZ 3 choroid 3.13 1.13 0.006 (5.35, 0.91) CI, confidence interval.

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1721

TABLE 6. Final Multiple Regression Models Based on Best-Corrected Visual Acuity Outcome

Standard Coefﬁcient, Error, 95% CI P Parameters 3104 3104 (3103) Value

Age 20.63 11.14 (0.12, 4.25) 0.064 AL 135.73 75.80 (1.28, 28.43) 0.073 OPL 29.96 53.16 (7.42, 13.42) 0.573 HFL þ ONL 22.61 18.70 (5.93, 1.40) 0.227 MEZ 540.30 124.45 (78.42, 29.64) <0.001 IZ þ RPE 2.00 42.48 (8.13, 8.53) 0.963 Choroid 68.75 18.68 (10.54, 3.21) <0.001 MEZ 3 choroid 2.84 0.81 (0.13, 0.44) <0.001

Although the associations between age and AL with BCVA were independent of the MEZ, they were excluded in the final predictive model. Nevertheless, age and AL might still be important factors associated with visual impairment in pathological myopia. Higher prevalence and severity of pathological myopia occur in patients with older age and longer ALs.1,5 With aging and axial elongation, choroidal FIGURE 6. ROC curve analysis of the global thickness comparisons of thinning occurs.23–26 Therefore, we hypothesize that aging the outer retinal sublayers and choroid for visual impairment. and axial elongation might lead to decreasing choroidal perfusion, thus resulting in visual impairment. These factors will be investigated in the future. ceptors, in the future, the use of advanced adaptive optics We acknowledge four limitations in the current study. First, imaging might also be helpful in showing photoreceptor the eyes with pathological myopia in category 4 of the META- alterations, that is, the loss of cone/rod cells, and deepen our PM classification system were not included because this would understanding of photoreceptor status in pathological myopia. have required the inclusion of older patients with longer ALs.19 The images from older patients with longer ALs were always of Fourth, this was a cross-sectional study. Longitudinal studies low quality and could not be analyzed correctly. Moreover, that would provide more reliable data regarding changes in the group constituted a small sample size and included more fundus as eyes transition from simple high myopia to females in each of the three myopia groups, while our study pathological myopia, and they would help to discern the required larger sample sizes and well-matched sex composition effect of changes in the MEZ on visual function. These too will to attain reliable statistical results. Second, the definition of be the subject of further studies. emmetropia in the current study (1.00 to þ0.50 D) was In summary, we demonstrated that changes in the thickness different from the World Health Organization definition (0.50 of the outer retinal sublayers and the choroid were correlated to þ0.50 D) for the high prevalence of myopia in China.27 with visual impairment in pathological myopia. Thinning of the However, this subtle discrepancy in the definition of emme- MEZ was significantly associated with worsening BCVA, and tropia is unlikely to have influenced our conclusions because this correlation varied with choroidal thickness. MEZ thickness we were more concerned with the clinical and structural is an important retinal anatomic factor with high functional variables of the pathological myopia group. Third, while we impact, and clinicians should closely examine and monitor were focused on OCT-documented changes in the photore- alterations of the MEZ in pathological myopia. The interrela-

FIGURE 5. Correlation between MEZ thickness and BCVA. (A) Scatterplot showing MEZ thickness versus BCVA in all eyes. (B) Scatterplot of subjects above the visual impairment cutoff value of choroidal thickness (109.77 lm). (C) Scatterplots of subjects below the visual impairment cutoff value of choroidal thickness. The dashed lines are the 95% conﬁdence intervals for the solid trend lines. The correlation between MEZ thickness and BCVA in all eyes was signiﬁcant (r ¼0.439, P < 0.001), but the correlation was different between the subjects above the visual impairment cutoff value of choroidal thickness (r ¼0.409, P < 0.001) and the subjects below the visual impairment cutoff value of choroidal thickness (r ¼0.468, P < 0.001).

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1722

TABLE 7. Summary of Previous Studies on Retinal and Choroidal Microstructure in Myopia Patients

Microstructure Study Alteration Related to Myopia

Current study Global thickness of HFL þ ONL and IZ þ RPE: pathological myopia < simple high myopia < emmetropia Global thickness of MEZ: pathological myopia < emmetropia/simple high myopia Global thickness of OS: emmetropia/pathological myopia < simple high myopia Global choroidal thickness: pathological myopia < simple high myopia < emmetropia Intraretinal layers of Abdolrahimzadeh et al.28 Thickness of ONL and OPL: high myopia < emmetropia the outer retina Liu et al.11 Thickness of OPL: high myopia > emmetropia Thickness of HFLþ ONL: high myopia < emmetropia Flores-Moreno et al.29 Thickness of photoreceptors and RPE aggregate negatively correlated with axial length Choroid Gupta et al.,30 Significantly thinner choroid in highly myopic eyes Abdolrahimzadeh et al.28 Wong et al.19 Thickness of subfoveal choroid: pathological myopia < simple high myopia

tionship between the changes in MEZ, choroid, and visual 11. Liu X, Shen M, Yuan Y, et al. Macular thickness profiles of functions over the natural course of pathological myopia intraretinal layers in myopia evaluated by ultrahigh-resolution warrants further study. optical coherence tomography. Am J Ophthalmol. 2015;160: 53–61.e2. 12. Chen Q, Tan F, Wu Y, et al. Characteristics of retinal structural Acknowledgments and microvascular alterations in early type 2 diabetic patients. Supported by research grants from the National Key Research and Invest Ophthalmol Vis Sci. 2018;59:2110–2118. Development Program of China (Grant 2016YFC0102500, Grant 13. Staurenghi G, Sadda S, Chakravarthy U, Spaide RF. Proposed 2016YFE0107000), the National Nature Science Foundation of lexicon for anatomic landmarks in normal posterior segment China (Grant 81570880), Natural Science Foundation of Zhejiang spectral-domain optical coherence tomography: the IN*OCT Province (Grant LQ16H120007), and the Zhejiang Provincial Key consensus. Ophthalmology. 2014;121:1572–1578. Research and Development Program (Grant 2015C03029). 14. Debuc DC, Salinas HM, Ranganathan S, et al. Improving image Disclosure: J. Ye, None; M. Shen, None; S. Huang, None; Y. Fan, segmentation performance and quantitative analysis via a None; A. Yao, None; C. Pan, None; X. Shi, None; F. Lu, None; Y. computer-aided grading methodology for optical coherence Shao, None tomography retinal image analysis. J Biomed Opt. 2010;15: 046015. References 15. Liu X, Shen M, Huang S, Leng L, Zhu D, Lu F. Repeatability and reproducibility of eight macular intra-retinal layer thicknesses 1. Ohno-Matsui K, Lai TY, Lai CC, Cheung CM. Updates of determined by an automated segmentation algorithm using pathologic myopia. Prog Retin Eye Res. 2016;52:156–187. two SD-OCT instruments. PLoS One. 2014;9:e87996. 2.WongTY,FerreiraA,HughesR,CarterG,MitchellP. 16. Gunkel M, Schoneberg J, Alkhaldi W, et al. Higher-order Epidemiology and disease burden of pathologic myopia and architecture of rhodopsin in intact photoreceptors and its myopic choroidal neovascularization: an evidence-based implication for phototransduction kinetics. Structure. 2015; 23:628–638. systematic review. Am J Ophthalmol. 2014;157:9–25.e12. 17. Smith SB, O’Brien PJ. Acylation and glycosylation of rhodop- 3. Hayashi K, Ohno-Matsui K, Shimada N, et al. Long-term sin in the rd mouse. Exp Eye Res. 1991;52:599–606. pattern of progression of myopic maculopathy: a natural history study. Ophthalmology. 2010;117:1595–1611. 18. Wang NK, Lai CC, Chu HY, et al. Classification of early dry- type myopic maculopathy with macular choroidal thickness. 4. Holden BA, Fricke TR, Wilson DA, et al. Global prevalence of Am J Ophthalmol. 2012;153:669–677. myopia and high myopia and temporal trends from 2000 through 2050. Ophthalmology. 2016;123:1036–1042. 19. Wong CW, Phua V, Lee SY, Wong TY, Cheung CM. Is choroidal or scleral thickness related to myopic macular degeneration? 5. Liu HH, Xu L, Wang YX, Wang S, You QS, Jonas JB. Prevalence Invest Ophthalmol Vis Sci. 2017;58:907–913. and progression of myopic retinopathy in Chinese adults: the 20. Shao L, Xu L, Wei WB, et al. Visual acuity and subfoveal Beijing Eye Study. Ophthalmology. 2010;117:1763–1768. choroidal thickness: the Beijing Eye Study. Am J Ophthalmol. 6. Cheng C, Teo K, Tan CS, et al. Myopic retinoschisis in Asians: 2014;158:702–709, e1. structural features and determinants of visual acuity and 21. Linsenmeier RA, Braun RD. Oxygen distribution and con- prognostic factors for progression. Retina. 2016;36:717–726. sumption in the cat retina during normoxia and hypoxemia. J 7. Sayanagi K, Ikuno Y, Soga K, Tano Y. Photoreceptor inner and Gen Physiol. 1992;99:177–197. outer segment defects in myopic foveoschisis. Am J Oph- 22. Jonas JB, Ohno-Matsui K, Spaide RF, Holbach L, Panda-Jonas S. thalmol. 2008;145:902–908. Macular Bruch’s membrane defects and axial length: associ- 8. You QS, Peng XY, Xu L, Chen CX, Wang YX, Jonas JB. Myopic ation with gamma zone and delta zone in peripapillary region. maculopathy imaged by optical coherence tomography: the Invest Ophthalmol Vis Sci. 2013;54:1295–1302. Beijing Eye Study. Ophthalmology. 2014;121:220–224. 23. Maruko I, Iida T, Sugano Y, Oyamada H, Akiba M, Sekiryu T. 9. Milani P, Pellegrini M, Massacesi A, et al. Is ellipsoid zone Morphologic analysis in pathologic myopia using high- integrity essential for visual recovery in myopic neovascular- penetration optical coherence tomography. Invest Ophthal- ization after anti-VEGF therapy? Graefes Arch Clin Exp mol Vis Sci. 2012;53:3834–3838. Ophthalmol. 2017;255:1713–1720. 24. Fujiwara T, Imamura Y, Margolis R, Slakter JS, Spaide RF. 10. Ohno-Matsui K, Kawasaki R, Jonas JB, et al. International Enhanced depth imaging optical coherence tomography of photographic classification and grading system for myopic the choroid in highly myopic eyes. Am J Ophthalmol. 2009; maculopathy. Am J Ophthalmol. 2015;159:877–883.e7. 148:445–450.

Downloaded from iovs.arvojournals.org on 09/24/2021 Thinning of MEZ and Choroid in Pathological Myopia IOVS j April 2019 j Vol. 60 j No. 5 j 1723

25. Ho M, Liu DT, Chan VC, Lam DS. Choroidal thickness 28. Abdolrahimzadeh S, Parisi F, Plateroti AM, et al. Visual acuity, measurement in myopic eyes by enhanced depth optical and macular and peripapillary thickness in high myopia. Curr coherence tomography. Ophthalmology. 2013;120:1909– Eye Res. 2017;42:1468–1473. 1914. 26. Abdolrahimzadeh S, Parisi F, Scavella V, Recupero SM. Optical 29. Flores-Moreno I, Ruiz-Medrano J, Duker JS, Ruiz-Moreno JM. coherence tomography evidence on the correlation of The relationship between retinal and choroidal thickness and choroidal thickness and age with vascularized retinal layers visual acuity in highly myopic eyes. Br J Ophthalmol. 2013; in normal eyes. Retina. 2016;36:2329–2338. 97:1010–1013. 27. Morgan IG, Rose KA, Ellwein LB. Is emmetropia the natural endpoint for human refractive development? An analysis of 30. Gupta P, Saw SM, Cheung CY, et al. Choroidal thickness and population-based data from the refractive error study in high myopia: a case-control study of young Chinese men in children (RESC). Acta Ophthalmol. 2010;88:877–884. Singapore. Acta Ophthalmol. 2015;93:e585–e592.

Downloaded from iovs.arvojournals.org on 09/24/2021