Scleral Thickness in Chinese Eyes

Abstract

Purpose.: We measured scleral thickness in eyes of Chinese, and assessed interregional differences and associations with age and axial length.

Methods.: Using light microscopy, we histomorphometrically measured scleral thickness at various locations in eyeballs from Chinese patients that had been enucleated due to retinoblastoma, uveal melanoma, or absolute painful glaucoma.

Results.: The study included 281 globes from patients with a mean age of 24.8 ± 23.1 years (range, 1–83 years) and mean axial length of 24.3 ± 3.9 mm (range, 17.0–35.7 mm). In multivariate analysis in children aged ≤2 years, thicker posterior scleral thickness was marginally significantly associated with older age (P = 0.07; standardized correlation coefficient β, 0.21; correlation coefficient B, 62.2; 95% confidence interval [CI], −4.3,128.8) after adjusting for shorter axial length (P = 0.01). In participants aged ≥5 years, larger posterior scleral thickness was significantly associated only with shorter axial length (P < 0.001; β, −0.49; B, −24.1; 95% CI, −30.6,−17.6), but not with age (P = 0.93). The ratio of posterior scleral thickness to scleral thickness in the pars plana region decreased significantly with increasing axial length. The correlation coefficients were higher for the ratio of scleral thickness posterior pole/pars plana region than for posterior pole/ora serrata, posterior pole/equator, or posterior pole/midpoint posterior pole to equator. Scleral thickness measurements were not significantly (all P > 0.10) associated with adult glaucoma.

Conclusions.: Scleral thickness increased up to an age of 2 years, while afterwards scleral thickness was independent of age and decreased with longer axial length. Differences in the associations between the regional scleral thickness ratios and axial length suggested a scleral thinning taking place in axially elongated eyes predominantly in the posterior globe segment.

The sclera forms the shell of the eyeball and is of utmost importance to stabilize size and shape of the eye globe. It also is highly important for the optical power of the eye, since axial length as one of the major determinants of the optical system of the eye is strongly influenced by the sclera. An axial extension of the eye globe directly leads to axial myopia, which has become one of the most important causes for visual impairment, visual field defects, and blindness worldwide.^1–3 Its importance will further increase due to the myopic shift taking place in particular in Asian countries at the Pacific rim.^4,5 Investigations by Heine in 1899⁶ and other more recent studies^7–10 have revealed that axial elongation is associated with a thinning of the sclera in the region at or behind the equator. In these studies, this scleral thinning was more marked the closer to the posterior pole. These studies, except for a recent one,¹⁰ did not address whether axial elongation was associated only with a thinning of scleral tissue in a process of remodeling of available scleral tissue, or whether axial elongation also was associated with an increase in scleral tissue volume, suggesting an additional new formation of scleral tissue. Further, none of these investigations was performed on eyes of Chinese. We, therefore, performed this study to measure scleral thickness in eyes of Chinese patients with a normal axial length and in eyes of Chinese patients with axial elongation. It may be of particular interest, since the marked increase in the prevalence of axial myopia has occurred predominantly in countries located at the Western Pacific rim, including China, without corresponding histologic studies performed on Chinese eyes yet. The answer to the questions addressed in our study may be of interest for the discussion whether Chinese eyes with axial elongation compared to Caucasian eyes with axial elongation show similar histomorphometric findings with respect to scleral thickness, and the answers may be of interest for the discussion on the mechanism of axial elongation or myopization of the eye globe. 

Methods

The Medical Ethics Committee of the Beijing Tongren Hospital approved the study protocol, according to the Declaration of Helsinki. In agreement with the approval by the ethics committee, informed consent was not obtained, since the globes had been enucleated up to 50 years before the study was initiated. 

Eye globes of Chinese patients that had been enucleated due to retinoblastoma, malignant uveal melanomas or absolute painful glaucoma were included into the histomorphometric study. Eyes with congenital glaucoma were excluded. In a similar manner, eyes of children with an age of less than 18 years generally were excluded if the clinical data showed an elevated IOP of more than 21 mm Hg or if the clinical or histologic examinations suggested a potentially elevated IOP, due to changes, such as iris neovascularization or anterior chamber occlusion. We collected data on sex, age, clinical diagnosis, and treatment. Within the group of patients with retinoblastoma, all tumors were at stage D or E based on the International Intraocular Retinoblastoma Classification.¹¹ Some patients with retinoblastoma had undergone several cycles of chemotherapy. In the melanoma group, all patients were without previous treatment for the tumor. Eyes with uveal melanomas invading the parapapillary region or the trabecular meshwork were excluded. In the glaucomatous group, vision was completely lost. 

The study population was divided into nonaxially elongated globes with an axial length of ≤26.0 mm, and into axially elongated globes with an axial length of more than 26.0 mm. The cut-off value of 26.0 in histological studies (roughly correlating with an axial length of 26.5 mm as measured intravitally) has been suggested to mark the transition from moderate to high myopia.^12,13 High myopia is associated with an increased prevalence and incidence of secondary changes, such as myopic maculopathy and glaucomatous optic neuropathy.^14,15 

After enucleation, the eyeballs were immediately fixed in a solution of 4% formaldehyde for at least 24 hours. The axial length and the horizontal and vertical diameters of the globe were measured with a caliper. The globes then were processed for histological sectioning. In the eyes with a tumor, the meridian of the cutting section depended on the location of the tumor. Eyes without a tumor were sectioned in a horizontal plane, running through the pupil, optic nerve head, and macula. All specimens were prepared in routine manner for microscopic examination, including dehydration in alcohol, imbedding in paraffin, and sectioning with a thickness of 3 to 5 μm, and staining by the periodic acid-Schiff (PAS) method or with hematoxylin-eosin. For all eyes, one section running through the central part of the optic nerve head was selected for further evaluation. 

The histological slides were examined by light microscopy and histomorphometrically assessed applying a digitized image analysis system (Moticam 2006, 2.0M pixel USB2.0, and Motic Digital Medical Image Analysis System; Motic China Group, Co. Ltd., Xiamen, China). On both sides of the histological section of the globe, we measured the scleral thickness at the corneal limbus, mid of pars plana, ora serrata, equator, midpoint between equator and posterior pole, posterior pole, outside of the optic nerve head at the merging point of the optic nerve sheaths with the posterior sclera, and at the border optic disc in the region of the peripapillary scleral flange (Figs. 1, 2).¹ For all locations at which scleral thickness was measured on both sides of the globe (corneal limbus, pars plana, ora serrata, equator, midpoint between equator, and posterior pole), the mean of both measurements was taken for further statistical analysis. In a similar manner, the mean of the scleral thickness measurements on both sides of the optic nerve head (outside of the merging point of dura mater with sclera) was taken for the statistical analysis. 

Figure 1

 

Locations of scleral thickness measurements at the limbus, pars plana region, ora serrata, equator, midpoint between the equator and the posterior pole, and the posterior pole.

Figure 2

 

Locations of scleral thickness measurements at just outside of the merging point of dura mater (black star) and posterior sclera (black double arrows) and at the peripapillary scleral flange (yellow star).

Statistical analysis was performed using a commercially available statistical analysis program (SPSS, version 22.0; IBM-SPSS, Chicago, IL, USA). In a first step, the distribution of the values was tested using the Kolmogorov-Smirnov test. We found that the scleral thickness parameters did not show a Gaussian distribution. In a second step, we calculated the mean values ± SDs, as well as the medians and the ranges of the measurements. In a third step of the statistical analysis, we compared the scleral thickness measurements between various regions of the eye, using the nonparametric Wilcoxon test for paired samples. In a fourth step of the statistical analysis, we searched for associations between the scleral measurements and age, sex, and presence of glaucoma, first in a univariate analysis (calculating Pearson's correlation coefficient), and eventually in a multivariate analysis. 

Results

The histomorphometric study included 281 globes of 281 Chinese individuals with a mean age of 24.8 ± 23.1 years (median, 22 years; range, 1–83 years). Mean axial length was 24.3 ± 3.9 mm (median, 23.8 mm; range, 17.0–35.7 mm). The total study population was divided into three subgroups: 122 eyes enucleated due to a retinoblastoma, 33 eyes with malignant uveal melanomas, and 126 eyes enucleated because of painful absolute glaucoma. There were 76 globes with an axial length of more than 26 mm and defined to be axially elongated. 

In the whole study population, scleral thickness was significantly (P = 0.004) thicker at the posterior pole than in the perioptic nerve head region (just outside of the merging of the dura mater with the sclera), where it was thicker (P < 0.001) than at the midpoint between the posterior pole and at the equator or at the limbus (with no significant difference, P = 0.40, between the two latter locations), where it was thicker (P < 0.001) than in the para plana region, where it was thicker (P < 001) than at the ora serrate, where finally it was thicker (P = 0.01) than at the equator (Table 1; Fig. 3). Thickness of the scleral flange did not differ significantly (P = 0.38) from the scleral thickness at the equator (Table 1). 

Table 1

 

Scleral Thickness Measurements Obtained in Various Regions of Globes

 

Figure 3

 

Boxplots showing the distribution of scleral thickness measurements at various locations of the eye globe.

In univariate analysis in the study participants with an age of ≥5 years, scleral thickness measurements at all measurement locations of the globe decreased (all P < 0.001) with increasing axial length (Table 2). The associations between longer axial length and thinner sclera were statistically stronger (i.e., the correlation coefficients were the higher), the closer the measurements were obtained to the posterior pole (except for the thickness of the peripapillary scleral flange, Table 2; Fig. 4). The correlation coefficients increased from the limbus as measurement location (correlation coefficient r, −0.37; P < 0.001) to the posterior pole (r = −0.49, P < 0.001, Table 2). In the study participants younger than 5 years, the association between longer axial length and thinner scleral thickness measurements was statistically significant only for the locations at, or posterior to, the ora serrata (Table 3; Fig. 4). None of the scleral thickness measurements was significantly associated with sex (all P > 0.15). 

Table 2

 

Correlations (Univariate Analysis) Between Axial Length and Scleral Thickness Measurements (μm) in Study Participants With an Age ≥ 5 Years

 

Figure 4

 

Scattergram showing the correlation between axial length and scleral thickness at the posterior pole, stratified by age. Equation of the regression line (univariate analysis): scleral thickness at the posterior pole = −19.35 × axial length (mm) + 1182.

Table 3

 

Correlations (Univariate Analysis) Between Axial Length and Scleral Thickness Measurements (μm) in Study Participants With an Age < 5 Years

 

In multivariate analysis in the study participants younger than 5 years, scleral thickness at the posterior pole decreased with longer axial length (P = 0.001) and increased marginally with older age (P = 0.11, Table 4). If the analysis included only children with an age of ≤2 years, the association between thicker posterior scleral thickness and older age was marginally significant (P = 0.07; standardized correlation coefficient β, 0.21; correlation coefficient B, 62.2; 95% confidence interval [CI], −4.3, 128.8) after adjusting for shorter axial length (P = 0.01). In the study population with an age of ≥5 years, larger scleral thickness at the posterior pole was significantly associated only with shorter axial length (P < 0.001; β, −0.49; B, −24.1; 95% CI, −30.6, −17.6), while it was not significantly associated with age (P = 0.93, Table 4). 

Table 4

 

Correlations (Multivariate Analysis) of Scleral Thickness Measurement at the Posterior Pole (μm) With Axial Length and Age, Stratified by Age

 

The ratio of scleral thickness at the posterior pole to scleral thickness in the pars plana region decreased significantly with increasing axial length (Table 5; Fig. 5). If the study population was stratified by age, the correlations coefficients were higher and more significant for the older group with an age of ≥5 years (Table 5). In both age groups, the correlation coefficients were higher for the ratio of scleral thickness at the posterior pole to scleral thickness in the pars plana region than for the ratios of scleral thickness posterior pole/ora serrata, scleral thickness posterior pole/equator, and scleral thickness posterior pole/midpoint posterior pole to equator, and finally for the ratio of scleral thickness posterior pole/optic nerve head region (Table 5). The interregional differences in the associations between the regional scleral thickness ratios and axial length suggested a scleral thinning taking place in axially elongated eyes predominantly in the posterior segment of the globe. Correspondingly, when comparing the axially elongated group with the nonaxially elongated group, the differences in scleral thickness between both groups were more marked the closer to the posterior pole the measurements were obtained (Table 1). 

Table 5

 

Correlations (Multivariate Analysis) Between Axial Length and the Ratio of Scleral Thickness Measurements at the Posterior Pole and Scleral Thickness Measurements at Other Locations, Stratified by Age

 

Figure 5

 

Scattergram showing the correlation between axial length and the ratio of scleral thickness at the posterior pole to the scleral thickness in the pars plana region, stratified by age. Equation of the regression line (univariate analysis): ratio scleral thickness posterior pole/pars plana region = −0.06 × axial length (mm) + 3.2.

Neither in univariate analysis nor in multivariate analysis were the scleral thickness measurements significantly (all P > 0.10) associated with the presence of (adult) glaucoma. 

Discussion

In our histomorphometric study on Chinese eyes, scleral thickness at the posterior pole decreased with longer axial length (P = 0.001) and showed a tendency of an increase with older age (P = 0.06) in multivariate analysis in children with an age of ≤2 years. In contrast, in the older group with an age of ≥5 years, larger scleral thickness at the posterior pole was significantly associated only with shorter axial length (P < 0.001), but not with age (P = 0.93). The ratio of scleral thickness at the posterior pole to scleral thickness in the pars plana region decreased significantly with increasing axial length. None of the scleral thickness measurements was significantly associated with presence of adult glaucoma (all P > 0.10). Generally, the sclera was thickest at the posterior pole, followed by the perioptic nerve head region, the midpoint between the posterior pole and equator, the limbus, para plana, ora serrata, and finally the equator. 

The results of our study agreed with the findings obtained in a similar study on Caucasian eyes in which, as in our study, the sclera was thickest at the posterior pole, followed by the perioptic nerve head region, midpoint between the posterior pole and equator, the limbus, the ora serrata, and the finally the equator.⁹ In that study as in our study, scleral thickness at the equator was similar to the thickness of the scleral flange at the optic nerve head. The results of our study also agreed with the study by Norman et al.,⁸ who examined 11 enucleated human globes (7 normal and 4 ostensibly glaucomatous) using high-field micro magnetic resonance imaging. They found a thickness over the whole sclera of 670 ± 80 μm (range, 564–832 μm). The maximum thickness occurred at the posterior pole, with mean thickness of 996 ± 181 μm. Scleral thickness decreased to a minimum at the equator, where a mean thickness of 491 ± 91 μm was measured. The measurements obtained by Norman et al.⁸ were higher than those of our study. The discrepancy may be explained by the difference in the study material with unfixed globes in their study and fixed specimen in our investigation. Furthermore, differences in axial length between both studies also may account for the difference in the mean scleral thickness. 

The present study, the previous study by Vurgese et al.⁹ on Caucasian eyes, and numerous previous studies on human globes as well as experimental studies on animals reported on associations between scleral thickness and axial length: In Caucasians as well as in Chinese with an age of ≥5 eyes, eyes with longer axial length have a thinner sclera, with the association being more pronounced the closer one gets to the posterior pole.^6–9 Correspondingly, scleral thickness measurements at or posterior to the equator were not significantly correlated with corneal thickness measurements,¹⁰ fitting with clinical studies in which central corneal thickness was not related with axial length.¹⁶ Axial elongation of the eye globe leads geometrically to an increase in the surface area of the eye. In the case of axial myopization, it has been unclear whether the enlarged globe surface area, corresponding to an increased surface area of the sclera, could fully explain the thinning of the sclera as observed in axially elongated eyes, or whether despite its thinning, the sclera would increase in volume in the course of axial elongation of the globe.^17–20 In a previous study by McBrien et al.,²⁰ who induced myopia in young tree shrews for a period of 12 days or 3 to 20 months, a significant scleral thinning and scleral tissue loss was observed, particularly at the posterior pole of the eye. Simultaneously, the eyes enlarged and myopia developed.²⁰ After the short-term myopia treatment, the collagen fibril diameter distribution was not significantly altered, whereas after the long-term period of monocular deprivation, significant reductions in the collagen fibril diameter were found, particularly at the posterior pole. They concluded that loss of scleral tissue and subsequent scleral thinning occurred rapidly during development of axial myopia, while an increased number of small diameter collagen fibrils in the sclera of highly myopic eyes was observed only in the longer term. 

As in the previous investigation, all scleral measurements were statistically not related with the presence of glaucoma. It is in agreement with clinical studies that eyes with glaucoma acquired after the age of approximately 5 years did not differ in the size of the optic disc.²¹ The optic disc size gets larger in highly myopic eyes with an axial elongation beyond approximately 26.5 mm.^12,13 The result of our study on a lack of an association between scleral thickness and glaucoma in human globes is in contrast with findings in studies on monkeys with experimental glaucoma. In monkeys with experimental glaucoma, an acute elevation of IOP and a short-term chronic increase in IOP have been described to cause regional thinning within the posterior sclera.²² Our findings on a lack of an association between scleral thickness and glaucoma in human globes are partially in contrast with the results obtained in an investigation by Coudrillier et al.,²³ who examined scleral specimens from 22 donors with no history of glaucoma and 11 donors with a history of glaucoma. The globes were excised 3 mm posterior to the equator and affixed to an inflation chamber. They found that in the older age group of 76 to 93 years, which included the data of 7 damaged glaucoma eyes, 6 undamaged glaucoma eyes, and 12 normal eyes, the damaged glaucoma specimens had a thicker sclera than either normal or undamaged glaucoma specimens (P = 0.01 and P = 0.02, respectively). When all diagnosed glaucoma specimens were put together in one single category, there were no significant differences in thickness compared to normal specimens (P = 0.09). 

Potential limitations of our study should be mentioned. First, serial sections of the globes were not available, so that it was not possible to precisely determine whether the histological section was located in the very center of the optic disc or whether it ran slightly paracentrally. Since, however, all slides went through the center of the cornea and through a central optic nerve head region with a diameter of approximately 0.5 mm, the error in scleral thickness measurements induced by a paracentral section meridian may not have markedly influenced the results. In addition, the section meridian was independent of the axial length so that this weakness in the study material may have affected the nonhighly myopic group and the highly myopic group in a similar manner. Second, the study did not include normal human eyes but eyes that were enucleated due to retinoblastoma, a malignant choroidal melanoma, or end-stage glaucoma. Therefore, it is not clear whether the results of our study can be transferred generally onto normal human eyes. Third, the histomorphometric measurements given in this study do not present intravital measurements, since postmortem tissue swelling and tissue shrinkage by the histological preparation will have changed the dimensions. However, it was not the purpose of the present investigation to evaluate the thickness of the sclera in real dimensions, but to compare the thickness measurements between various regions of the eye and between ethnicities. Fourth, the histological sections of the melanoma group were orientated according to the main location of the tumor, while the glaucomatous globes had been opened in a horizontal direction. 

In conclusion, the sclera was thickest at the posterior pole, followed by the perioptic nerve head region, midpoint between the posterior pole and equator, the limbus, para plana, ora serrata, and finally the equator. Scleral thickness increased up to an age of 2 years, while afterwards scleral thickness was independent of age and decreased with longer axial length. The latter associations were statistically stronger for measurements obtained closer to the posterior pole. Differences in the associations between the regional scleral thickness ratios and axial length suggested a scleral thinning taking place in axially elongated eyes predominantly in the posterior globe segment. 

Acknowledgments

The authors alone are responsible for the content and writing of the paper. 

Disclosure: L. Shen, None; Q.S. You, None; X. Xu, None; F. Gao, None; Z. Zhang, None; B. Li, None; J.B. Jonas, None