Thickness profiles of the corneal epithelium along the steep and flat meridians of astigmatic corneas after orthokeratology

Orthokeratology (OK) is a contact lens technology that results in the temporary correction of myopia using specially designed reverse geometry rigid-gas-permeable (RGP) contact lenses that are worn during sleep. There has been a resurgence in clinical interest into OK over the past decade due to emerging evidence of its’ role in controlling the progression of myopia, particularly in Chinese children [1, 2]. Modern OK lenses are designed to apply positive pressure over the central cornea and negative pressure in the mid-periphery under the steeper secondary “reverse” curve of the contact lens. In doing so, it produces a flattening of the central corneal treatment zone that corrects the myopic refractive error by reducing the corneal power, and a steepening of the mid-peripheral cornea stabilizing the lens.

Although the mechanism by which OK lenses change the biomechanical and topographical properties of the cornea is still not fully understood, the reshaping of the corneal epithelium remains one of the most important theories. OK lenses appear to produce their biomechanical changes through thinning of the central corneal epithelium and thickening of the paracentral epithelium [3‐5]. In eyes with corneal toricity, the application of OK lenses may lead to differential changes in corneal epithelial thickness along the steep and flat axes, due to the different pressures exerted by the lens on the cornea in each of these meridians. Currently, there are no studies investigating the reshaping of the corneal epithelium in eyes with corneal toricity. Corneal toricity exists in a significant proportion of the general population, with studies demonstrating that over 90% of the population aged 5 to 40 years have corneal toricity of ≥0.25 diopters of cylinder (DC) in their central corneas [6, 7]. In eyes with corneal toricity, significant asymmetry of the paracentral cornea has also been found in the chords of 5 to 8 mm [8, 9]. Moreover, studies [10, 11] have found that long-term wear of OK lenses can lead to changes in corneal toricity even after discontinuation of wearing; which may partly be due to the differential remodeling of the corneal epithelium along each of the different meridians. A recent study by our group found that long-term OK lens wear resulted in a decrease in the corneal power along the flatter meridian and an increase in corneal toricity after discontinuation of OK lens wear for 1 month [12].

Therefore, investigating the effects of OK lens wear on epithelial remodeling in eyes with corneal toricity, and the differences in epithelial thickness changes along the steep and flat corneal meridians may allow us to better understand the biomechanical effects on the cornea and the mechanism by which OK lenses produce their refractive correction.

This study used a Fourier-domain optical coherence tomography (FD-OCT) system [13] to automatically generate topography maps of corneal epithelial thickness in subjects who wore OK lenses for a period of 6 months. The epithelial thickness profiles were compared between the steep and flat axes of the cornea to determine if OK lens wear induces different biomechanical and thickness changes along the different axes and regions of the cornea.

The right eyes of 33 males (57.9%) and 24 females (42.1%) were included in this study. The mean age of all subjects was 13.6 ± 6.1 years (range: 9–38 years). The mean CED of 8 mm at baseline was 21.15 ± 8.79 μm (range: 0–39 μm). After 6 months of OK-lens wear, all subjects achieved good lens centration (decentration less than 1.0 mm), as indicated by a bull’s eye pattern on corneal topography maps. Unaided visual acuity achieved 20/25 or better in all of the subject eyes. The mean spectacle plane spherical error changed from − 2.77 ± 1.34 DS at baseline (range: − 5.00 to − 1.00 DS) to 0.13 ± 0.50 DS (range: 0.00 to + 1.50 DS, t = 13.87, P < 0.001) and the mean refractive astigmatism changed from − 0.45 ± 0.32 DC (range: − 1.25 to 0.00 DC) to − 0.49 ± 0.35 DC (range: − 1.50 to 0.00 DC, t = 0.802, P = 0.426). A summary of the refractive and topographic measurements is displayed in Table 1.

The distribution of corneal toricity according to the radius at baseline and post-lens wearing is illustrated in Fig. 3. The differences in CTs between pre- and post-lens wear were significant for 0.5 mm, 2 mm, 2.5 mm, 3 mm, 3.5 mm and 4 mm in radius (Paired t test, P < 0.05).

The central corneal epithelium thinned and the mid-peripheral corneal epithelium thickened after spherical OK lens wear (Fig. 4). Along the steep axis, the epithelium thinned most at 0.5 mm inferior to the corneal apex (− 9.2 ± 4.3 μm) and thickened most at 3 mm inferior to the apex (6.4 ± 6.4 μm). Along the flat axis, the epithelium thinned most at 0.5 mm temporal to the apex (− 9.2 ± 4.7 μm) and thickened most at 3 mm temporal to the apex (4.4 ± 5.9 μm). Figure 5 illustrates the scatterplots of △ET along the steep and flat axes in different radii.

In order to compare the differences in △ET between the steep and flat axes, the mean of the △ET (△ETm) at the same radius along each meridian was calculated and is illustrated in Fig. 6. No significant differences were found between △ETSm and △ETFm at 0.5 mm and 1 mm. △ETFm (− 4.2 ± 3.4 μm) thinned more than △ETSm (− 3.4 ± 4.0 μm, P = 0.027) at 1.5 mm. At 2.0 mm, △ETSm thickened, while △ETFm thinned (△ETSm: 1.0 ± 4.9 μm, △ETFm: − 0.8 ± 3.6 μm, P < 0.001). △ETSm thickened more than △ETFm at 2.5 mm (△ETSm: 4.1 ± 5.1 μm, △ETFm: 2.8 ± 4.2 μm, P = 0.019) and 3.0 mm (△ETSm: 5.0 ± 5.0 μm, △ETFm: 3.7 ± 4.9 μm, P = 0.036).

The absolute value of differences between △ETSm and △ETFm (∣△ETSm - △ETFm∣) were significantly correlated with the baseline central CT at 2.0 mm, 2.5 mm and 3.0 mm in radius (2.0 mm: r = 0.285, P = 0.032; 2.5 mm: r = 0.422, P = 0.001; 3.0 mm: r = 0.293, P = 0.027). ∣△ETSm - △ETFm∣was significantly correlated with the baseline peripheral CT at 2.5 mm in radius (r = 0.299, P = 0.028). No significant correlation was found between ∣△ETSm - △ETFm∣ and the post-OK lens wear central CT for any radius. ∣△ETSm - △ETFm∣was significantly correlated with the post-OK lens wear peripheral CT at 2.5 mm (r = 0.383, P = 0.004).∣△ETSm - △ETFm∣ was significantly correlated with baseline spherical refractive error at 2.0 mm, 2.5 mm and 3.0 mm (2.0 mm: r = 0.342, P = 0.009; 2.5 mm: r = 0.632, P < 0.001; 3.0 mm: r = 0.426, P = 0.001). ∣△ETSm - △ETFm∣was significantly correlated with baseline cylindrical refractive error at 2.5 mm (r = 0.303, P = 0.022).

Several mechanisms have been proposed to explain how OK lenses produce their refractive and myopia control effects, with changes in the thickness and topographical profiles of the corneal epithelium being one of the key mechanisms behind its’ action. In the current study, the OK lens-induced changes in the epithelial thickness profiles were evaluated and compared between the steep and flat meridians of corneas with toricity less than 1.50 diopters. After 6 months of overnight spherical myopic OK lens wear, our results demonstrated that the central corneal epithelium thinned and the mid-peripheral corneal epithelium thickened in all four quadrants. In the para-central cornea, ETF thinned more than ETS at 1.5 mm and 2 mm; in the mid-periphery, ETS thickened more than ETF at 2.5 mm and 3 mm. The differences between △ETSm and △ETFm were significantly correlated with the baseline central CT and the baseline peripheral CT. Our results suggest that there is a correlation between the corneal toricity and the remodeling of the corneal epithelium after wearing of OK lenses.

Thinning of the central corneal epithelium has been demonstrated in both clinical and histological studies of overnight OK lens wear [16, 17]; however, the reported changes in the mid-peripheral corneal epithelium have been less consistent. In a previous study by our research group [5], we found that the paracentral corneal epithelium thickened most in the inferior quadrant, which is similar to the findings of the present study and supported by the findings of a recent study by Reinstein et al. [3] Lian et al. [4] used a custom-made OCT to measure the thickness profile changes of the corneal epithelium after OK lens wear and found that the nasal and temporal mid-peripheral (4-6 mm) epithelium thickened after OK wear; however, they found no mid-peripheral epithelial thickening along the vertical meridian. They instead found that the epithelium tended to become thinner in the superior mid-peripheral region. They contributed this thinning of the superior mid-peripheral epithelium to greater pressure of the spherical OK lens on the superior cornea exerted by pressure from the upper eyelid, as well as superior decentration of the OK lens, which may have occurred in some of their subjects.

To the best of our knowledge, there are currently no clinical studies investigating the epithelial changes induced by spherical OK lens wear along different meridians of eyes with corneal toricity. Zhou et al. [18] compared the epithelial and stromal thicknesses along the steepest and flattest meridians between subjects with keratoconus and those with healthy corneas with corneal astigmatism ≥2 D. They found that the epithelium showed mid-peripheral thinning over the cone in addition to thickening superior to the cone in cases of keratoconus. Meanwhile, the average epithelial thickness map showed only small deviations in the distribution along the steepest meridian and the flattest meridian in healthy subjects. It was proposed that the epithelium compensates for anterior surface shape irregularities by altering its thickness profiles, which may be a similar mechanism observed with the ocular surface changes induced by OK lens wear. Investigating the effects of OK lens wear on epithelial remodeling in eyes with corneal toricity may allow us to better understand its biomechanical effects on the corneal profile. This current study found that ETF thinned more than ETS in the paracentral cornea, while ETS thickened more than ETF in mid-peripheral cornea. This could be explained by the mechanism by which spherical OK lenses reshape the cornea. The degree of pressure exerted by the OK lens depends on the sagittal height of the lens and the thickness of the tear film between the posterior lens surface and the cornea. The cornea along the steeper meridian is further away from the OK lens (i.e. greater sagittal height) than the flatter meridian. As a consequence, the thinning of the paracentral cornea is not as great along the steeper meridian as compared with the flatter meridian, and the thickening of the mid-peripheral epithelium is greater along the steeper meridian. The significant correlation between ∣△ETSm - △ETFm∣and baseline corneal toricity also helps to further support this mechanism. Although no significant correlation was found between ∣△ETSm - △ETFm∣and post-wearing central CT, ∣△ETSm - △ETFm∣was significantly correlated with the post-wearing peripheral CT at 2.5 mm. Therefore, the epithelial remolding could have an impact on the peripheral corneal toricity.

In theory, if following OK lens wear the epithelium thinned more along the flat meridian of the paracentral cornea and thickened more along the steep meridian of the mid-peripheral cornea, then the difference in corneal elevation between the flat and steep meridians should be reduced and the paracentral corneal toricity should also be reduced. However, we found an increase in the corneal toricity after OK lens wear. Previous studies [19, 20] have also indicated that OK lens wear could induce an increase in astigmatism, both regular and irregular. One possible explanation for this could be that the remodeling of the epithelium is neutralized by other factors such as lens decentration, uneven force imposed by the eyelids and stromal remodeling [19‐21]. Further studies investigating the impact of corneal decentration on inducing corneal toricity are needed to clarify this issue.

Currently, most studies have only investigated the OK lens-induced epithelial changes along the horizontal meridian [20, 22]. Although Lian et al. [4] compared the thickness profile changes between the vertical and horizontal meridians, corneal toricity was not taken into account; and therefore, it is difficult to compare the findings of their study with those of the present study.

Spherical OK lens wear results in greater thinning of the corneal epithelium along the flat meridian of the paracentral cornea, and greater thickening along the steep meridian of the mid-peripheral cornea. These findings add further evidence to the understanding of the mechanisms by which OK lens wear induces changes on the cornea and provides its refractive effects. In addition, the greater thinning of the epithelium along the flat meridian may help to explain the residual corneal flattening along the flat meridian after discontinuation of OK lens wear [10, 12].