Interplay Between γ-Zone Peripapillary Atrophy and Optic Disc Parameters in Central Visual Field Impairment in Highly Myopic Eyes

Abstract

Purpose: The purpose of this study was to investigate the association between the development of glaucoma-like central visual field (VF) defects and γ-zone peripapillary atrophy (γPPA) in eyes with high myopia (HM).

Methods: This retrospective study measured γPPA length, prelaminar tissue (PLT) thickness, and lamina cribrosa (LC) defects’ width in HM eyes with central VF defects by swept-source optical coherence tomography (OCT). The central VF was assessed by using the Humphrey Field Analyzer 10-2 program, and average light sensitivity in the papillomacular bundle region (C1 region) and mean deviation (MD) values were calculated. Interactions among the central VF and structural abnormalities were analyzed using correlation tests and linear regression.

Results: Forty-five eyes from 45 patients were included. PLT thickness was positively associated with light sensitivity in the C1 region and MD values, whereas γPPA was associated only with MD values. LC defects showed no association with either the C1 region or MD values. LC defects’ width was negatively correlated with γPPA length. Stratified analysis showed that in the group with γPPA length ≥763 µm, PLT thickness was positively associated, and γPPA length was negatively associated with the C1 region and MD values. No such associations were found when the γPPA length was <763 µm. The association between LC defects with central VF remained nonsignificant.

Conclusions: Our findings identified γPPA length and PLT thickness as risk factors of glaucoma-like central VF defects in HM, particularly when γPPA length exceeds 763 µm. We suggest that HM eyes with larger γPPA and thinner PLT should be paid special attention for early detection of central VF defects.

Myopia has experienced a significant global increase in prevalence and incidence in recent years. Current estimates suggest that by 2050, approximately half of the world’s population may be affected by myopia, with approximately 9.8% experiencing high myopia (HM), defined as an axial length (AL) exceeding 26.0 mm.¹ The pathophysiology of myopia involves elongation of the eyeball; in cases of HM, excessive elongation contributes to a range of ocular complications, notably including glaucoma-like visual field (VF) defects.² In patients with HM, glaucoma-like VF defects often present with distinctive patterns compared to conventional glaucoma, exhibiting a marked tendency toward central VF impairment that contributes significantly to early and severe visual disability.³ A strong association has been observed between higher degrees of myopic refraction and central VF impairment in glaucoma.^4–6 However, the mechanisms driving the development of central VF defects in HM eyes remain insufficiently understood, thereby impeding the development of targeted treatment strategies. 

Peripapillary atrophy (PPA) is frequently observed in eyes with myopia and glaucoma, and it is commonly categorized into α-zone and β-zone based on fundus photography.⁷ The α-zone PPA is marked by irregular patterns of hyperpigmentation and hypopigmentation in the peripheral region, whereas β-zone PPA (βPPA) is characterized by large visible choroidal vessels and exposed sclera between the optic disc margin and α-zone. Recently, based on the position of the peripapillary Bruch’s membrane (BM) endpoint, βPPA has been further subdivided into new βPPA (containing BM) and γ-zone PPA (γPPA), the latter distinguished by the absence of BM.⁸ Predominantly based on optical coherence tomography (OCT) findings, the new βPPA has been shown to have a strong association with the presence and progression of glaucoma.^9,10 Conversely, γPPA has been reported to correlate with AL, with its size expanding as AL increases,^11–13 and exhibits a weaker association with glaucoma.^11,14,15 However, a consensus on the relationship between γPPA and VF defects has yet to be reached, especially as studies focusing specifically on eyes with HM remain lacking. 

The lamina cribrosa (LC) is a mesh-like structure composed of collagen bundles that create pores through which retinal ganglion cell axons and retinal blood vessels pass.^16,17 Acquired structural abnormalities in the LC are considered a primary factor in the development of typical VF defects in glaucoma. In eyes with HM, excessive axial elongation also leads to deformation and defects in the LC, which are believed to contribute to the heightened susceptibility of HM eyes to glaucoma and associated visual impairments.^18–20 Additionally, a previous study using OCT demonstrated that LC defects primarily located at the temporal periphery of the optic disc correspond to the papillomacular bundle (PMB) region in HM eyes.²¹ Prelaminar tissue (PLT), which lies anterior to the LC, consists of retinal ganglion cell axon bundles, astrocytes, capillaries, and extracellular material.^17,22 Thinning of the PLT has also been associated with VF defects in glaucoma.^23,24 In eyes with HM, it has been observed that PLT overlying LC defects become thinner, which may contribute to VF abnormalities.²⁵ 

The aim of this study is to clarify whether γPPA, LC defects, and PLT thinning contribute to the development of central VF defects in eyes with HM. Additionally, this study seeks to investigate the inter-relationships among these optic nerve-related parameters. 

Methods

Ethical Approval

This retrospective study adhered to the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of the Institute of Science Tokyo, Tokyo, Japan (Study Approval Number: M2020-161). Written informed consent was obtained from all participants. 

Participants

Patients with HM who visited the Advanced Clinical Center for Myopia between January 2015 and July 2023 were enrolled in this study. The inclusion criteria were as follows: (1) eyes with AL of ≥26.5 mm; (2) diagnosis of open-angle glaucoma (OAG) confirmed by Humphrey Field Analyzer (HFA) 30-2 examinations in accordance with the Anderson-Patella criteria; (3) completion of optic disc OCT examination; and (4) availability of HFA 10-2 examination results within 6 months of the optic disc OCT examination. 

Exclusion criteria for eyes included the following: (1) intraocular pressure (IOP) >21 millimeters of mercury (mm Hg); (2) a history of vitreoretinal surgery, glaucoma laser treatment, or glaucoma incisional or filtration surgery; (3) any other diseases affecting the VF, including macular atrophy and choroidal neovascularization; (4) poor-quality OCT images; and (5) low reliability in HFA 10-2 results, defined as fixation loss >20%, false-positive rate >15%, and false-negative rate >15%. This study collected the data on sex, age, AL (IOL Master; Carl Zeiss Co., Germany), and IOP (Goldmann applanation tonometer; Haag-Streit, Koeniz, Switzerland) from medical records and within 1 year of the date of OCT examination. If both eyes of a patient met the criteria, one eye was randomly selected for assessment. 

Swept-Source Optical Coherence Tomography Examination

Optic disc images obtained using swept-source OCT (DRI-OCT Triton; Topcon Co., Japan) were measured and analyzed. The OCT scanning protocol included a scan length of 9 mm with 12 evenly spaced radial meridian scans centered on the optic disc. Three radial scan lines were selected from the OCT images: the optic disc-fovea line and 2 adjacent radial lines located 30 degrees from the optic disc-fovea line. The tissue length of each line was measured using the IMAGEnet software (Topcon Co.), with the magnification error corrected by the built-in IMAGEnet software. Eyes were excluded from the study if more than one of the three images could not be analyzed or recognized. For OCT analysis, we assessed the LC defects, γPPA, and PLT at the temporal edge of the optic disc (Figs. 1A, 1B). Two evaluators (authors N.Z. and T.Y.) independently measured all OCT images of the included eyes, remaining blinded to other patient information. The averaged results from the two evaluators were used for the final analysis. The average values of each parameter from the three radial scan lines were calculated for subsequent statistical analyses. The width of LC defects in each scan line was measured horizontally at the widest defect location in the LC. The thickness of PLT overlying LC defects was measured perpendicularly to the estimated anterior LC surface at the point where the thinnest PLT was observed. The γPPA was defined as the region adjacent to the optic nerve head (ONH), characterized by the absence of BM and atrophy of the retinal pigment epithelium (RPE) and choriocapillaris. The length of γPPA was measured as the horizontal distance from the disc edge to the BM opening (BMO). 

Figure 1.

 

Optic nerve head (OHN) and peripapillary structural parameters and visual field parameters. (A) Fundus photograph of the optic disc. The three black arrows indicate the scan lines of swept-source OCT. (B) In the OCT image, the yellow, green, and red two-directional arrows represent the thickness of PLT, the width of LC defects, and the length of γPPA, respectively. The width of LC defects in each scan line was measured horizontally at the widest defect location in the LC. The thickness of PLT overlying LC defects was measured perpendicularly to the estimated anterior LC surface at the point where the thinnest PLT was observed. (C) The extent of the C1 region corresponding to the papillomacular bundle in the HFA 10-2 examination.

HFA 10-2 Visual Field Examination

The 10-2 program of the HFA III (HFA; Carl Zeiss Co., Germany) was used to assess the central VF. Testing points corresponding to the PMB area, identified as the C1 region and consisting of 14 testing points (Fig. 1C), were also analyzed.²⁶ The average retinal light sensitivity of the C1 region and the mean deviation (MD) values from HFA 10-2 were used as parameters for assessing central VF impairment. 

Statistical Analyses

All statistical analyses were conducted using SPSS Statistics 26.0 (IBM-SPSS, Chicago, IL, USA) and Python 3.12.3 (Python Software Foundation, Wilmington, DE, USA). Interobserver reproducibility was assessed by calculating the intraclass correlation coefficient (ICC). Demographic characteristics and ocular parameters were summarized as mean ± standard deviation. Spearman or Pearson correlation analyses were used to examine correlations between ocular variables, with the choice of method depending on whether the data followed a normal distribution. Additionally, the impact of each continuous factor on VF parameters (light sensitivity of the C1 region and MD values) was assessed using linear regression analysis. 

For stratified analysis, we used piecewise linear regression to detect change points in the relationship between γPPA and light sensitivity of the C1 region. Eyes were divided into 2 groups based on γPPA length: <763 µm (group A) and ≥763 µm (group B). Differences in continuous variables between group A and group B were analyzed using Student’s t-test or the Mann-Whitney U test. Within each group, Spearman or Pearson correlation analyses were conducted to identify potential correlations between variables, and linear regression analyses were performed to determine the effect of each continuous factor on central VF. Variables with a P value < 0.10 in univariate analysis were included in the multivariate analysis. Statistical significance was defined as a P value < 0.05. The false discovery rate was applied to control for the false positive rate. 

Results

Participants’ Characteristics and Ocular Parameters

Eighty-six eyes from 86 patients with HM who visited the Advanced Clinical Center for Myopia between January 2015 and July 2023 were enrolled in this study. A total of 41 eyes were excluded: 26 eyes had diseases affecting the VF, 8 eyes had poor image quality or unrecognizable images, and 7 eyes had elevated IOP or a history of glaucoma surgery or laser treatment. Ultimately, 60 eyes from 45 patients (30 female patients and 15 male patients) met the inclusion criteria. For patients with both eyes eligible, one eye was randomly selected. Patients’ demographic and ocular characteristics are shown in Table 1. There was excellent interobserver agreement in the measurements of the extent of the LC defects (ICC = 0.986, 95% confidence interval [CI] = 0.972 to 0.992), PLT (ICC = 0.983, 95% CI = 0.953 to 0.986), and γPPA (ICC = 0.974, 95% CI = 0.953 to 0.986). 

Table 1.

 

Patients’ Demographic and Ocular Characteristics in all Cases

 

Correlation Analysis Between Patients’ Characteristics and Optic Nerve-Related Parameters in all Eyes

The scatter plots in Figure 2 show the relationship between γPPA length and PLT thickness with central VF parameters. There was a strong positive correlation between AL and γPPA length, which was statistically significant (r = 0.475, P = 0.009; see Fig. 2C). The length of γPPA also showed a statistically significant negative correlation with the width of LC defects (r = −0.533, P < 0.001; see Fig. 2D). 

Figure 2.

 

Scatterplots show the correlation among the parameters in all cases. The γPPA was significantly negatively correlated with central VF (A, B). The AL was significantly correlated with γPPA (C) based on Pearson’s correlation analysis. The LC defects were significantly negatively correlated with PPA (D). The thickness of PLT showed a significant correlation with MD Values and light sensitivity of C1 region (E, F). The dotted lines indicate the 95% confidence intervals.

Identifying Risk Factors Associated With Central VF Values in all Eyes

Univariate and multivariate linear regression analyses were conducted to illustrate the relationships between central VF parameters and various factors, including age, IOP, AL, and optic disc-related parameters across all eyes (Table 2). The length of γPPA and the thickness of PLT were significantly associated with MD values (β = −0.407, P = 0.018 and β = 0.591, P < 0.001, respectively) in univariate analysis. When considered in multivariate analyses, γPPA length and PLT thickness (β = −0.265, P = 0.036 and β = 0.519, P < 0.001, respectively) remained significant predictors of MD values. However, when linear regression was performed with the average light sensitivity in the C1 region, only PLT thickness showed a significant association in the univariate (β = 0.703, P < 0.001) and multivariate analyses (β = 0.659, P < 0.001). 

Table 2.

 

Univariate and Multivariate Linear Regression Analysis With the Average Sensitivity of C1 Region and MD Values in all Eyes

 

Stratified Analysis Based on γPPA Length

To further understand the relationship between the C1 region and γPPA length, piecewise linear regression was used to detect the change point, which was identified at a γPPA length of 763.00 µm. Participants were divided into 2 groups based on γPPA length: <763 µm (group A) and ≥763 µm (group B). Twelve eyes were classified into group A (3 male patients and 9 female patients, mean age = 56.25 ± 12.64 years), and 33 eyes were assigned to group B (12 male patients and 21 female patients, mean age = 58.36 ± 8.98 years). Significant differences were observed between the two groups (Table 3). The AL in group A was significantly shorter than that in group B (28.99 ± 1.29 vs. 30.22 ± 1.76 mm, respectively; P = 0.033). Moreover, the width of the LC defects in group A was significantly larger compared to that in group B (206.42 ± 107.84 vs. 140.86 ± 91.94 µm, respectively; P = 0.045). However, the two groups did not show significant differences in MD values (−6.82 ± 4.08 vs. −9.58 ± 7.30 decibels (dB); P = 0.317), average light sensitivity in the C1 region (27.19 ± 6.84 vs. 24.25 ± 8.22 dB; P = 0.329), or PLT thickness (180.46 ± 61.92 vs. 148.04 ± 71.55 µm; P = 0.172). Additionally, no significant differences were found in terms of age and IOP between the two groups (all P > 0.05). 

Table 3.

 

Comparison of Patients’ Demographic and Ocular Characteristics Between Group A and Group B

 

Correlation Analysis Between Patients’ Characteristics and Optic Nerve-Related Parameters in the Two Groups

First, age, IOP, and AL in either group A or group B showed no significant correlation with optic nerve-related parameters. No significant correlation was observed among the length of γPPA, the width of LC defects, and the length of PLT in group A (all P > 0.05). In group B, statistically significant correlations were observed between the length of γPPA and LC defects (r = −0.522, P = 0.006; Fig. 3F), but not between LC defects and PLT, or γPPA and PLT (all P > 0.05). The relationships among γPPA, LC defects, and VF parameters are shown in Figure 3. 

Figure 3.

 

Scatterplots show the correlation between parameters in group A and group B. In group A, neither MD values nor C1 region showed a relationship with γPPA (A, B). The LC defects also showed a significant correlation with γPPA (C). The central VF showed a significant correlation with γPPA in group B (D, E), and the LC defects had a significant correlation with γPPA (F). The dotted lines indicate the 95% confidence intervals.

Identifying Risk Factors Associated With Visual Field Values in the Two Groups

In the univariate analysis for the average light sensitivity of the C1 region and MD values in group A, none of the variables show statistically significant associations. 

In Table 4, univariate and multivariate linear regression analyses were conducted to examine the relationship between the average MD values and various factors in group B. In the univariate analysis, significant predictors included the length of γPPA (β = −0.424, 95% CI = −0.014 to −0.002, P = 0.042) and the thickness of PLT (β = 0.603, 95% CI = 0.032 to 0.091, P < 0.001). When these two parameters were considered in multivariate analyses, the length of γPPA (β = −0.318, 95% CI = −0.011 to −0.001, P = 0.027) and the thickness of PLT (β = 0.541, 95% CI = 0.027 to 0.084, P < 0.001) remained significant predictors of the average MD values. 

Table 4.

 

Univariate and Multivariate Linear Regression Analysis With the Average Light Sensitivity of C1 Region and MD Values in Group B (γPPA ≥ 763 µm)

 

In addition, univariate and multivariate linear regression analyses were conducted to examine the relationship between the average light sensitivity of the C1 region and various factors in group B. In the univariate analysis, the significant predictors was the thickness of PLT (β = 0.744, 95% CI = 0.057 to 0.114, P < 0.001). When multivariate analyses were performed, the length of γPPA (β = −0.241, 95% CI = −0.010 to −0.0001, P = 0.047) and the thickness of PLT (β = 0.697, 95% CI = 0.051 to 0.107, P < 0.001) were identified as significant predictors of the average light sensitivity in the C1 region. 

Discussion

The present study investigated the relationship between glaucoma-like central VF defects in eyes with HM and three optic nerve-related parameters: γPPA length, PLT thickness, and LC defects’ width. We used the HFA 10-2 program for more precise detection, as central VF impairment can originate from very small areas that are challenging to assess with the HFA 30-2 VF examination.²⁷ 

Jonas et al. previously demonstrated that the size of γPPA size increases significantly with AL in a nonlinear manner, with a marked increase beginning at an AL of approximately 26.5 mm.¹¹ In the present study, we found a strong positive correlation between γPPA length and AL in all HM eyes, consistent with previous findings.^11,28,29 Our analysis demonstrated that γPPA length was significantly associated with MD values but not with sensitivity in the C1 region across all eyes. Therefore, we performed a stratified analysis for the C1 region. In the piecewise linear regression analysis based on γPPA length, we identified a significant change point in the average light sensitivity in the C1 region at a γPPA length of 763 µm. A longer γPPA was determined to be a significant risk factor for average light sensitivity in the C1 region when its length was ≥763 µm, but not in eyes with γPPA <763 µm. Using this change point, the MD value was also significantly associated with γPPA when the length was ≥763 µm, but not in eyes with γPPA length <763 µm, as well as in the C1 region. What might explain this discrepancy in results based on γPPA length? One possibility lies in the differences in AL and γPPA length. Previous studies examining the relationship between glaucomatous VF defects and γPPA reported mean AL values of approximately 26 mm at most, with mean γPPA lengths approximately 300 µm.^15,30,31 In contrast, the mean AL in our study was significantly longer at 29.89 mm, and the mean γPPA length was substantially greater at 1086.73 µm. This excessive axial elongation and γPPA enlargement may contribute to retinal nerve fiber layer (RNFL) thinning. The specific structure-function relationship between retinal light sensitivity and RNFL thickness is well-established. Previous studies have shown that the relationship between RNFL thinning and VF defect progression is not linear, with substantial RNFL thinning required before functional VF defects become detectable.^32,33 This specific relationship may also contribute to the nonlinear association between γPPA length and central VF defects. In eyes with HM, the fovea-to-optic disc distance tends to elongate, which has been shown to be associated with the development of γPPA.³⁴ When the γPPA length is reduced to less than 763 µm, central VF impairment may occur. In the present study, PLT thickness was significantly associated with central VF. In contrast, there was no significant association between PLT thickness and γPPA length; however, there was a tendency for PLT thickness to differ between the γPPA ≥763 µm group and the γPPA <763 µm group (180.46 ± 61.92 vs. 148.04 ± 71.55 µm). RFNL thinning is induced not only by longer AL but also by higher IOP and aging,^24,32 which may suggest that a complex mechanism may be involved. 

LC defects on the temporal side of the ONH are reportedly the most common type in eyes with HM.^33,35 Sawada et al. demonstrated that LC defects in glaucomatous eyes with moderate to HM significantly correlated with MD values in HFA 24-2 testing.³⁵ LC defects have also been significantly associated with PMB defects in patients with glaucoma with HM.³⁶ However, in our regression analysis of both groups based on γPPA length, we found that the thickness of the PLT was significantly associated with central VF impairment in the larger γPPA group, but not with the width of LC defects. Previously, Xie et al. also demonstrated PLT thickness, but not LC defects’ width, was correlated with VF defects detected by Goldmann perimetry, which is in good agreement with our result.²⁵ Interestingly, LC defects on the temporal side of the ONH were observed in all HM eyes in this study, suggesting that the development of LC defects may not be rare in HM eyes with central VF defects. The reasons for this discrepancy are not yet fully understood. However, the development of LC defects may have the potential to contribute to central VF impairment, whereas PLT thinning—which directly causes VF impairment—may result from not only LC defects but also other factors in eyes with HM. Further analysis is needed for a more comprehensive understanding. 

Interestingly, our study demonstrated a negative correlation between the width of LC defects and the length of γPPA in eyes with HM. This finding may be explained by the rigidity of the BM, which contains elastin and resists deformation, maintaining a stable distance between the BMO and the macula during axial elongation.³⁷ As AL elongates and the BMO remains attached to or near the ONH, substantial traction may be exerted on the ONH toward the macula via the BM. This traction may lead to the separation of the LC from the peripapillary sclera, causing LC defects in the nasal region of the ONH.^33,35 As AL elongates and the BMO separates significantly from the ONH, leading to developing γPPA, the traction on the ONH may reduce. This reduction in traction could potentially halt or slow the progression of LC defects. In this study, the width of the LC defects showed no significant correlation with central VF defects. However, the development of LC defects may indirectly affect central VF function, as the PLT is histologically located above the LC, the proximity that could contribute to functional changes. 

Some limitations existed in this study. First, the relatively small sample size, along with its cross-sectional and retrospective design, limits the ability to observe structural changes over time. Because myopia typically progresses slowly over several decades, long-term studies with larger sample size are essential for a comprehensive understanding. This study is the first to reveal a correlation between γPPA development and central visual function, potentially serving as a foundation for future research on glaucoma-like VF impairment in HM eyes. Second, excessive AL elongation in HM eyes causes structural changes in ocular tissues, such as optic disc tilt, posterior staphyloma, and peripapillary intrachoroidal cavitation, all of which contribute to VF impairment. Many of these structural changes often coexist within the same eye, complicating the analysis of VF impairment in HM. Moreover, previous studies have reported that the severity of retinal atrophy associated with HM is linked to reduced visual acuity.^38,39 However, in the present study, eyes with extensive atrophic changes were excluded; thus, the influence of retinal atrophy on visual function is considered minimal. It remains crucial to assess macular abnormalities when evaluating central VF impairment associated with γPPA development. 

In conclusion, this study demonstrated that a γPPA length exceeding 763 µm and PLT thinning are significant risk factors for glaucoma-like central VF defects in eyes with HM. Notably, this is the first study to reveal that γPPA, previously thought to be unclear in association with central VF defects, is associated with such defects in eyes with HM. In the present study, the width of LC defects showed no significant direct association with central VF impairment. However, a negative correlation between LC defect width and γPPA was observed, suggesting that the influence of LC defects on central VF impairment should not be entirely dismissed. These findings indicate that particular attention should be given to γPPA development and PLT thinning for the early detection of central VF defects in eyes with HM. The elongation of γPPA may represent a novel hallmark of central VF impairment in eyes with HM. 

Acknowledgments

The authors thank Editage (www.editage.com) for providing professional editing services for this manuscript. 

Supported by the Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (Grant Number: JP19K09987). 

Disclosure: N. Zhou, None; T. Yoshida, NIDEK (C); K. Sugisawa, None; S. Yoshimoto, None; K. Ohno-Matsui, Santen (C), CooperVision (C)