Abstract
Purpose.:
To measure corneal and limbal epithelial thickness (ET) in normal subjects and to evaluate its variation with age by using anterior segment optical coherence tomography (AS-OCT).
Methods.:
A total of 180 normal subjects (180 healthy eyes) were enrolled and divided into four groups according to age: A (0–20 years), B (21–40 years), C (41–60 years), and D (>60 years). Cornea and limbus were imaged with OCT. Corneal ET (CET) was obtained automatically by the built-in analysis software of the OCT system. Limbal ET (LET) in four quadrants was manually measured from OCT images.
Results.:
Corneal ET of a central 2-mm diameter zone in groups A, B, C, and D were 53.4 ± 2.8 μm, 53.4 ± 2.7 μm, 53.2 ± 3.0 μm, and 52.9 ± 3.3 μm, respectively, N showing no significant change with aging. In the paracentral zone extending to 6-mm diameter, correlation analysis suggested that CET was inversely associated with age (P < 0.05). Limbal ET in the nasal and the temporal quadrants were similar and decreased with aging, the averages were 58.3 ± 8.1 μm, 54.1 ± 6.1 μm, 51.2 ± 6.1 μm, 51.6 ± 5.2 μm for groups A, B, C and D, respectively; while age seemed to have no effect on LET of the superior and the inferior quadrant.
Conclusions.:
The paracentral corneal epithelium, as well as the nasal and temporal limbal epithelium, became thinner with aging, while the central CET seemed to remain constant. Measurement with AS-OCT of the corneal and limbal ET could aid in clinical assessment and planning treatments of the cornea.
We used the Fourier-domain OCT system (RTVue-100; Optovue Inc., Fremont, CA, USA) with a cornea anterior module long adapter lens (1.96-mm scan depth and 6-mm scan width) in this study. The device worked at 830-nm wavelength and had a scan speed of 26,000 axial scans per second. A Pachymetry+Cpwr scan pattern was used to map the cornea over a 6-mm diameter area, the settings were 8 meridional B-scans, consisting of 1024 A-scans each 5-μm axial resolution. Following correct fixation and centering on the pupil, each scan was acquired within 5 seconds. Images of limbal area in each quadrant (superior, inferior, nasal, and temporal) were obtained using the cross-line scan mode. The subject was asked to fixate at a peripheral target to maintain the perpendicularity of the OCT beam at the surface of the targeted tissue, which was essential for obtaining accurate thickness values.
All tests were performed by one trained operator. To avoid potential artifacts or interferences from other ocular examinations, OCT scans were performed first, on one eye (randomly selected) of each subject for three times.
Statistical analyses were performed with statistical software (SPSS for Windows, version 19.0; SPSS, Inc., Chicago, IL, USA). Basic descriptive statistics were calculated on all data gathered, and values were reported as means ± SD. Difference in each parameter across age groups was tested with one-way ANOVA (or nonparametric Kruskal-Wallis test where appropriate). Comparisons between groups were conducted with Student's t-test (or Mann-Whitney U test where appropriate). Spearman's correlation analysis was used to determine the significance level of the association of age with CET and LET. Two-tailed P values of <0.05 was considered significant.
The average central CET (53.2 ± 2.9 μm) measured by this OCT system (Optovue, Inc.) was within the range (48.0–59.9 μm) reported in peer-reviewed literature.
3–6,12–14,21–29 Our study demonstrated the epithelial layer of the cornea had a nonuniform thickness profile, as suggested in previous studies.
5,10,29 In an aging eye, the topographic variability of CET would increase, which was considered as a natural tendency to counterbalance the underlying stromal irregularities; however, age effect on thickness of the epithelial layer was not fully elucidated.
11–15,21 Present findings showed that the differences of CET in 2- to 6-mm diameter area between under and over 40 years were ranged between 1.19 to 2.65 μm, which were slightly larger than the repeatability measurement fluctuations reported by Ma et al.
30 With such data validity, our study demonstrated that epithelial thickness in paracentral cornea decreased with aging. However, these differences, though significant, were below the resolution of the AS-OCT device used, perhaps future technologies like swept-source imaging will help us to confirm our initial findings and elucidate the anatomy better.
Corneal epithelium is renewed and maintained by limbal epithelial stem cells (LESCs) during normal homeostasis or after wounding.
31,32 In the course of proliferation and differentiation, these cells migrate centripetally while detaching from the basement membrane outward to the surface to form 5 to 7 layers of corneal epithelium, and finally shed by desquamation.
33 Investigations into the normal corneoscleral transitional zone suggested that the diameter and density of cells in both the limbus-cornea and limbus-palisade changed with advancing age, whereby the mean diameter increased with a corresponding decrease in cell density.
9,34 Zheng and Xu
35 utilized in vivo confocal microscopy to observe the microstructure of limbus and found that the presence of Vogt palisades, which was believed to harbor LESCs, declined with aging; whereas the limbal basal epithelial cells showed age-related size enlargement and quantity reduction. Such results implied that as the age increased, LESCs kept a dynamic evolvement with attenuating proliferative potential, leading to degenerated capacity in maintaining the integrity and stability of the corneal epithelium. This may be an explanation of our findings that the old had thinner corneal epithelium compared with the young.
Interestingly, age seemed to have no effect on central CET, which concurred with previous findings.
11–15,21 Majo et al.
36 proposed that limbus only contributed cells to the corneal epithelium in response to wound healing, whereas during normal homeostasis the renewal was carried out by stem cells scattered throughout the corneal epithelium itself. In other studies, Chang et al.
37,38 found that the capacity for epithelial cell proliferation and migration appeared to be as active in the central cornea as in the limbus; epithelial recovery in response to wound healing remained equal even after ablation of the limbus, cell density of the central corneal epithelium could increase faster than at the periphery. Laboratory evidence suggested that epithelial cells separated from human central cornea were capable of forming clonogenic spheres which were positive for p63 immunolabeling, indicating that central corneal epithelium contained cells with stem cell- or progenitor cell-like properties. These novel findings could be an explanation why the central corneal epithelium was less susceptible to age influences, also supplemented the conventional view that maintenance and repair of the corneal epithelium were only carried out by cells proliferating and moving centripetally from the limbus.
In terms of limbal epithelium evaluation, we observed that mean LET in the nasal and temporal quadrants was thinner than in the superior and inferior quadrants; the former exhibited significant age-related decrease while the latter seemed to remain constant with aging. The regenerative capacity of LESCs might account for such variations.
It was well recognized that the biological functions and activities of LESCs were regulated by exquisite programs. Various extrinsic (e.g., ultraviolet radiation) or intrinsic (e.g., cytokines) signals could cause failure to maintain a normal microenvironment in the limbal region, thus giving rise to the development of ocular disorders.
41–43 Histological and slit-lamp examinations have shown the presence of melanocytes, among the basal cells of the rete pegs in Vogt palisades, correlated with the pigment distribution in the limbal area.
9,44,45 This melanin pigment is supposed to protect LESCs from the deleterious effect of ambient ultraviolet radiation.
46 According to Zheng and Xu's
35 study, aging also gave rise to a reduction in melanocytes quantity, which could impair LESCs repository due to the decreasing amount of pigment. In an aging eye, LESCs in nasal and temporal limbus were more likely to be the victims of solar damage than those in the superior and inferior limbus, because the former had higher risk of exposure to ultraviolet radiation without eyelids covering up. In addition, an intrinsic weakness in LESC reserves was implied by less prominent Vogt palisades in the nasal and temporal limbus,
47,48 suggesting that these regions might be less likely to undergo effective self-repair and self-renewal in normal homeostasis or after injury.
However, age-related changes of LET presented in this study were not observed previously. In fact, very little was known about epithelial thickness profiles at limbal region. With different OCT systems, epithelial thickness in the nasal and temporal limbus was measured to be 60.1 to 79.6 μm, usually thinner than that in the superior and inferior limbus which was reported to be 68.0 to 106.8 μm.
21,22,27,39 Our findings showed lower LET in each quadrant compared with these values. Such discrepancy could be related to intrinsic differences of instruments. For instance, configurations, properties, and software algorithms of different OCT devices were not calibrated identically.
40 On the other hand, both in vivo and histological studies have suggested that the epithelial layer gradually became thinner from central cornea to limbus,
21,27,39,49 while our results failed to demonstrate such variations. This could be attributed to different methods in image processing of the raw OCT scans. Specifically, the pachymetry map-based CET was accurately acquired by the internal software with an automated algorithm, while LET was measured by manual segmentation using an external software. Results yielded from different measurement protocols were incomparable, therefore the changing patterns of epithelial thickness from central cornea to limbus could not be inferred from the present findings.
The current study suffered from several limitations. First, given the fact that the OCT system was unable to discriminate the precorneal tear film, of which the thickness was reported 4.79 ± 0.88 μm in a recent study,
50 a possible uncertainty was raised: whether the age-related changes we have observed were indeed attributed to the corneal epithelium thinning or faked by the tear film incorporated in the measurement. It was proposed that steady tear film thinning, mainly due to evaporation, eventually caused tear film breakup at different locations. Since the tear breakup time decreased with aging,
51,52 we postulate that the tear film thinning rate increases with aging. This could lead to a reduction of tear film thickness (TFT) which might skew the results in the present study. However, OCT images were acquired within 5 seconds since the eyes opened, not long enough to observe the breakup of tear film (dry eyes were excluded in subject selection); the decline of TFT with aging was still lack of evidence. In addition, observations showed that the inferior tear film was thinner than the superior in normal eyes.
53,54 Such a spatial distribution pattern of TFT was quite different from that of CET. We hope future imaging technologies with higher resolution and a function of automatically excluding tear film in the measurement, will help us evaluate CET more accurately and confirm its association with aging. Second, although data from the central 6-mm cornea is essential for vision and adequate for keratoconus screening, it is insufficient for design and fitting of contact lenses, or for monitoring disorders involving the peripheral cornea and planning clinical procedures for them. OCT scans covering the entire cornea are preferable for obtaining more information. The last limitation lies in the subjective interpretation of images when determining the limbus. In some cases, blurred visualization of the scleral spur made it difficult to precisely define the posterior margin of limbus and to guarantee the accuracy of LET measurements. Since LET, for the first time, was defined using the proposed averaging approach, these measurement data should be considered preliminary and subject to further refinement to verify reliability and validity. After all, so few was reported about LET and its association with the properties of the limbal epithelial cells. Future studies combined with other clinical or laboratory investigations such as confocal microscopy and immunostaining of the targeted tissues or cells, will be needed to help understand the physiology or pathogenesis of human limbus in health and diseases.
To date, reports of CET measurement in normal population were not adequate. Results of our study have supplemented some normative information about human corneal epithelium. In conclusion, age seemed to have no effect on epithelial thickness of the central 2-mm cornea, or of the superior and inferior limbus; while in the majority of paracentral cornea up to 6 mm in diameter, as well as in the nasal and temporal limbus, the epithelial layer became thinner with increasing age. Studies of diverse populations and larger sample sizes may further validate these initial findings. Human corneal and limbal epithelial thickness profiles in normal corneas, measured by AS-OCT, could serve as a suitable basis for investigations into ocular surface pathology.
Supported by grants from the Key Clinic Medicine Research Program, the Ministry of Health, China (201302015); the National Science and Technology Research Program, the Ministry of Science and Technology, China (2012BAI08B01); the National Natural Science Foundation of China (81170817, 81200658); the Scientific Research Program, Science and Technology Commission of Shanghai Municipality, Shanghai (13441900900, 13430720400, 134119a8800, 13430710500). The authors alone are responsible for the content and writing of the paper.
Disclosure: Y. Yang, None; J. Hong, None; S.X. Deng, None; J. Xu, None