Background
Laser-assisted in situ keratomileusis (LASIK) is widely accepted as a reproducible and effective surgical procedure for correcting myopia. Surgical equipment and outcomes for LASIK have been much improved over the past few decades; however, myopic regression is still a matter of concern [
1,
2]. Previous studies have suggested a forward shift of the cornea, resulting from the compromised biomechanical rigidity, as one possible explanation for myopic regression after LASIK [
2,
3]; however, the underlying mechanism of refractive regression is as yet unclear. Studies have reported a relationship between the elevation of intraocular pressure (IOP) and corneal protrusion, and demonstrated that topical anti-glaucoma eye drops were effective in correcting approximately 0.5 diopter (D) of refractive regression. This was presumably a result of lowered IOP leading to the backward movement of the cornea and the flattening of its curvature [
4‐
7]. On the other hand, steroid eye drops have long been used to decrease myopic regression after photorefractive keratectomy (PRK), as myopic regression has been known to be related to wound healing [
8,
9].
The recent availability of corneal epithelial imaging by Fourier domain optical coherence tomography (FD-OCT) provides a practical tool for in vivo epithelial mapping, which demonstrates good repeatability in both normal and post-LASIK eyes [
10]. It allows for the non-invasive measurement of corneal epithelial thickness in the clinical routine with adequate speed and resolution. Previous studies have reported an increase in corneal epithelial thickness following laser ablative myopic surgeries, such as LASIK [
11], PRK [
12], and Small incision lenticule extraction (SMILE) [
13]. Studies have suggested that epithelial thickening was associated with myopic regression after LASIK as well as PRK, although the wound healing process would be quite different between both procedures [
14‐
16]. Changes in stromal thickness have also been suggested as being responsible for postoperative refractive regression [
17]. To the best of our knowledge, no studies have documented corneal epithelial thickness changes during medical treatment of myopic regression. Moreover, there are only a few studies demonstrating the widely accepted association between epithelial thickening and refractive myopic regression [
18‐
20]. This study, therefore, aimed to determine the relationship between refractive regression and changes in epithelial thickness using FD-OCT during medical treatment of myopic regression.
Methods
Patients
This retrospective observational study included 84 eyes of 54 subjects that had been treated with steroid and anti-glaucoma eye drops for myopic regression, occurring after LASIK, between November 2017 and May 2019. The original surgery included femtosecond-LASIK (FS-LASIK) for the correction of myopia at the B & VIIT Eye Center (Seoul, Korea) from 2004 to 2016. Untreated post-LASIK subjects whose epithelial thickness have been followed-up for 12 weeks were presented as a control group to clarify that corneal epithelial thickness change was caused by drops in treatment group. The surgical units used for each case were variable. All patients had reached at least 20/20 uncorrected visual acuity (UDVA) 1 month after surgery and maintained stable visual acuity and refraction for at least 6 months. Preoperative data and patient characteristics are summarized in Table
1. All subjects were evaluated with manifested refraction, auto refract-keratometer (ARK-1, NIDEK, Tokyo, Japan), and Pentacam tomography (OCULUS Optikgeräte, Wetzlar, Germany). Keratometric values measured using ARK and posterior curvature with Pentacm were used to estimate corneal power change before and after medication. Myopic regression was defined as a myopic shift of 0.5 D or greater in manifest refraction from the third postoperative month following refractive surgery. The inclusion criteria of this study were as follows: 1) regression of myopia with spherical equivalent refraction of − 0.5 D or over for at least 3 months, 2) a corrected distance visual acuity of 20/20 or better, and 3) 3 months of medical treatment (10 to 14 weeks) for myopic regression. The exclusion criteria were as follows: 1) subjects with suspected under-correction during the original surgery, 2) a history of ocular surgery other than refractive surgery, 3) eyes that had had contact lenses worn in the past 1 month, 4) eyes with myopia caused by lenticular change and 5) eyes with postoperative complications including an iatrogenic ectasia. All subjects were prescribed with either loteprednol etabonate (Lotemax, Bausch & Lomb, Rochester, NY, USA) or flumetholone (Santen, Tokyo, Japan) with dorzolamide/timolol (Cosopt, Merck & Co., Inc., Whitehouse Station, NJ, USA). Patients were instructed to use a topical anti-glaucoma eye drop twice a day and a steroid eye drop four times a day for the first month and twice a day for the following 2 months.
Table 1
Patient demographic characteristics
Age at operation (years) | 25.0 ± 4.8 |
Preoperative SE (diopters) | −5.2 ± 2.1 |
Preoperative CCT (μm) | 544.3 ± 25.7 |
Ablation depth (μm) | 85.6 ± 27.1 |
Treatment duration for regression (weeks) | 12.2 ± 1.2 |
Postoperative time at treatment (months) | 82.7 ± 27.9 |
To explore the influence of corneal epithelial thickness, all subjects were grouped into 3 groups based on their corneal epithelial thickness at the time of initiation of medical treatment, with group 1 consisting of CET ≤ 57, group 2, 58 ≤ CET ≤ 61, and group 3, CET ≥ 62. In addition, all subjects were categorized into three groups based on their response to medical treatment: 0, no response; 1, partial response; 2, good response. The good response was defined as successful medical treatment when patients showed the following criteria: improvement in visual acuity of two lines or more and reduction of myopia (refraction or K) with 0.5 D or more. The partial response was defined as having an uncorrected visual acuity of 20/40 or more with evidence of vision or refraction improvement that did not fulfill the criteria for good response. The no response group was defined as either lack of improvement of visual acuity/refraction or having a final visual acuity worse than 20/40. The study protocol was approved by the Institutional Review Board (CR 319130) and was conducted according to the tenets of the Declaration of Helsinki.
Measurement of corneal epithelial thickness
The corneal epithelial and total thickness data were obtained at the time of initiation and termination of medical treatment using the RTVue OCT system (Optovue Inc., Fremont, CA, USA), with a corneal adaptor module set at a wavelength of 830 nm. We scanned the cornea in eight meridians by using the ‘Pachymetry + Cpwr (corneal power)’ scan (software version A6.11.0.12) over a 6-mm diameter centered at the corneal vertex. The corneal epithelial thickness map was generated using an automatic algorithm and was divided into a total of 17 sectors: a central 2-mm diameter zone, eight paracentral zones within an annulus between the 2- and 5-mm diameter rings, and eight mid-peripheral zones within an annulus between the 5- and 6-mm diameter rings. Stromal thickness was obtained by subtracting the epithelial thickness from the corneal thickness.
Statistical analysis
Statistical analyses were performed using SPSS version 21.0 for Windows (IBM Corp, Armonk, NY, USA). Snellen visual acuity was converted to the logMAR scale for statistical analysis. Corneal epithelial thickness and other continuous variables obtained before and after medical treatment were compared using paired t-tests. Student t-tests were performed for comparison between treatment group and no treatment group. Subgroup analyses were conducted using one-way analysis of variance. Simple linear regression analyses were performed to investigate the association between corneal epithelial thickness and refraction, keratometric values, and UDVA (logMAR). A multivariable logistic regression analysis was performed to investigate factors affecting the success of medical treatment.
A receiver operating characteristic (ROC) curve analysis was conducted to evaluate the ability of corneal epithelial thickness to predict the success of treatment. In the ROC curve, the true positive rate (Sensitivity) was plotted as a function of the false positive rate (1-Specificity) for different cut-off points of corneal epithelial thickness at the time of medical treatment initiation. Statistical significance was defined as P < 0.05.
Discussion
It is widely known that the corneal epithelium, which is highly reactive to irregularities and changes in the underlying stroma, undergoes remodeling in response to corneal refractive surgery. Studies suggest that despite preservation of corneal epithelium and Bowman’s layer, epithelial remodeling occurs after LASIK in a similar manner to that observed after PRK [
11,
21]. It has been suggested that epithelial hyperplasia may contribute to myopic regression after myopic laser ablation; however, its association with refractive change is as yet unclear. Erie [
20] showed that myopic regression after PRK was significantly associated with an increase in corneal epithelial thickness, whereas Ivarsen et al. [
17] did not find any correlation between changes in corneal epithelial thickness and changes in refraction after PRK or LASIK. Studies have failed to demonstrate a correlation between the change in corneal epithelial thickness and refraction, presumably due to short-term follow-ups, complex predisposing factors, and a lack of reliable measurements across the longitudinal follow-ups. Notably, a recent study demonstrated that myopic shift observed at postoperative 1 year showed a significant correlation with increase of corneal epithelial thickness after PRK [
18].
Other than epithelial hyperplasia, forward movement of the cornea, accumulation of new stromal collagen, or lens-induced regression has been suggested to play a role in myopic regression [
1,
2], however, the mechanism for myopic regression after LASIK still remains to be elucidated. Of these suggested mechanisms, we aimed to investigate the association between corneal epithelial thickness and myopic regression. The recent introduction of corneal epithelial thickness mapping in FD-OCT prompted us to study epithelial remodeling after laser refractive surgery and we monitored changes in corneal epithelial thickness during medical treatments for clinically diagnosed myopic regression. As a result, we have demonstrated a significant association between corneal epithelial thickness and refraction.
Previous researches described a negative meniscus-like lenticular pattern, with more thickening at the paracenter than at the center after LASIK due to aspheric ablation profiles [
11,
21,
22], but myopic regressed eyes showed more epithelial thickening at the center than paracenter as shown in our data. If we compared topographic epithelial thickness pattern between no response and good response group, central hyperplasia pattern was prominent in good response group, while a negative meniscus-like lenticular pattern was observed in no response group (
supplemental figure). Of note, epithelial thickness decreased at the center by the greatest amount followed by the paracenter after medical treatment, and this topographic thickness change being accompanied by refractive changes was in agreement with Barraquer’s law of thicknesses which indicates that if there is relative tissue addition in the center of the cornea, and conversely if there is relative tissue removal from the center of the cornea there will be a hyperopic refractive shift [
23]. As a result, overall topographic epithelial thickness pattern over 6.0 mm after medical treatment was recovered to a negative meniscus-like lenticular pattern, similar to the previously published postoperative map [
11,
21,
22].
Our study raised some issues related to myopic regression and epithelial remodeling. First, this study suggested that further increase in central epithelial thickness may occur years after LASIK in some eyes. Our data showed that central epithelial thickness of regressed eyes were significantly thicker than untreated eyes where the refraction has been stable long-term (Table
2). Of note, our data demonstrated that eyes with thicker epithelium have been treated for greater myopia. Second, epithelial thickening could, at least, partly be reversed by medical treatment. It was a surprise to observe active epithelial remodeling, years after surgery that was responsive to medical treatment.
To the best of our knowledge, this study is the first to report that decrease in corneal epithelial thickness occurs during medical treatment and that there is an association with improvement in visual acuity and refraction. This study, therefore, reinforces the role of epithelial thickening in the development of myopic regression following myopic LASIK.
There are limitations in this study. First, we were unable to clarify which medication showed the observed effect due to absence of solitary anti-glaucoma drop treatment group. We hypothesized that topical steroid may have epithelium modulating effect and added anti-glaucoma drops to prevent the possible IOP rise during steroid treatment, but did not include an anti-glaucoma drop treated group in this study. Previous studies have reported the effects of steroid and anti-glaucoma eye drops in reversing myopic regression after PRK and LASIK, respectively [
5‐
9]. A previous study concluded that topical steroids did not seem to have a beneficial role in routine postoperative treatment after LASIK [
24]; however, no studies have evaluated the effect of steroids on myopic regression after LASIK. Studies have shown that steroid eye drops delay proliferation and attenuate migration of corneal epithelial cells [
25,
26]. It is conceivable that steroid retard an epithelial proliferation and migration for subjects having a greater epithelial hyperplastic response and decrease (normalize) epithelial thickness; however, the exact mechanism needs to be clarified in further studies. Topographically, a greater decrease of epithelial thickness was observed in the center than in the paracenter. This result may be supporting our hypothesis because central hyperplastic area may demand more proliferation and migration to maintain thickness. With regard to possible effects of anti-glaucoma drops on myopic regression, we could not observe the suggested finding by previous publications, the backward movement of protruded cornea by lowering the IOP [
3,
5]. In our analysis, however, flattening of the cornea has only been shown with the anterior curvature, not posterior curvature (Table
2). Furthermore, we could not observe the relationship between the magnitude of decrease of IOP and refraction change. Second, this study did not contain important data such as duration of the period over which reversal was maintained or recurrence rate after discontinuation of medication. In our experience, about 80% patients showed a stable refraction for 3 months after termination of eyedrops, but no long-term follow-up data is available. In the context of usefulness as a treatment option for myopic regression, this study lacks the data to address critical questions. A prospective study is necessary to optimize treatment strategy and to investigate how to maintain stable refraction after initial success. Third, clinical implication of our cut-off epithelial thickness may be very limited because we could not assess the real magnitude of epithelial hyperplasia due to unavailable preoperative measurement. Despite these limitations, we tested the feasibility of corneal epithelial thickness as an indicator for the initiation of medical treatment for myopic regression. Our data showed that with central corneal epithelial thickness > 60.50 μm, there was an expected improvement in visual acuity of more than two lines and 0.5 D or greater change in refraction with about 80% sensitivity and specificity. We believe that measurement of baseline epithelial thickness and longitudinal follow up could provide a better strategy for myopic regression. Based on this finding, we suggest that medical treatment may be beneficial for patients with thicker corneal epithelium and myopic regression.
Conclusion
The major clinical implication of this study was to demonstrate a significant association between change in corneal epithelial thickness and refraction. Our data suggested that decrease in corneal epithelial thickness occurred concurrently with flattening of corneal curvature, leading to an improvement in refraction and visual acuity. Furthermore, this study suggested that corneal epithelial thickness plays a role in deciding whether patients would respond to medical treatment. We believe that routine measurement and monitoring of corneal epithelial thickness could be useful in subjects seeking laser refractive surgery to understand the pathogenesis of myopic regression as well as to optimize treatment strategy.
Interestingly, this study demonstrated that corneal epithelial thickness decreased as myopic regression subsided, and the change in corneal epithelial thickness was significantly associated with refractive change.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.