It is a common phenomenon for the anterior surface areas flattened by laser ablation to be compensated by epithelial remodeling and thickening to maintain good optical quality of the cornea, after corneal reflective surgery [17, 18]. Variable patterns of epithelial thickening after surgery influence the postoperative refractive status differently [9]. When the epithelial thickening of the central regions exceeds the midperipheral regions, it will increase the optical power in the central regions, which is equivalent to adding a convex lens on the anterior corneal surface; therefore, refractive regression may occur. However, if the epithelial thickening of the midperipheral region exceeds the central regions, resulting in a negative meniscus-like epithelium, it is difficult for refractive regression to occur. Reinstein et al. found that the epithelium of the central regions thickened by 5 µm one month after LASIK for correcting myopia, refractive state drifting to myopia for 0.39 D [19]. The epithelial thickness increased by approximately 6 µm one year after surgery. Kanellopoulos showed that the epithelium thickened by 6 µm in the central regions one year after LASIK for correcting high myopia [20]. However, in the midperipheral regions, it was thickened by approximately 10 µm. Therefore, epithelial thickening does not necessarily cause refractive regression. In this study, epithelial thickening in central regions was more obvious than that in the midperipheral regions when refractive regression occurred.
The pattern of epithelial remodeling and thickening may be different between PRK and LASIK because the mechanism of corneal wound healing is different, and the probability and rate of refractive regression may also be different [21]. Earlier studies have reported that the duration and degree of epithelial thickening in PRK were different from that in LASIK [22]. However, in our previous studies, when correcting high myopia, the postoperative refractive status in FS-LASIK was more likely to drift to myopia than TPRK, which may be related to the more decreased biomechanical stability and more obvious epithelial thickening after FS-LASIK [23]. Kanellopoulos and Asimellis believed that the epithelial thickness was affected not only by the corneal curvature gradient but also by corneal biomechanics after surgery [24, 25]. The epithelial thickening is more obvious in patients with higher myopia and lower corneal thickness [22]. When FS-LASIK and PRK were combined with prophylactic cross-linking, corneal rigidity was strengthened, and the degree of epithelial thickening decreased; thus, the probability of refractive regression was reduced [26, 27].
This study prospectively investigated the postoperative refractive regression after TPRK and FS-LASIK for correcting high myopia. Refractive regression occurred approximately 1 year after surgery in most patients. The epithelial thickness was significantly higher than that before surgery when refractive regression occurred. In FS-LASIK group, the increase in epithelial thickness in the central region was more obvious than that in the midperipheral region, resulting in a greater degree of myopic regression. After 3 months of steroid treatment, the epithelial thickness was relatively lower and the refractive regression was reduced, which means that the occurrence of regression is highly correlated with epithelial thickening. This is similar to Ryu’s study [28]. However, in Ryu’s study, the preoperative epithelial thickness was lacking, and the direct relationship could not be obtained between the change in epithelial thickness and refractive status when regression occurred. Furthermore, due to the combined use of steroid and anti-glaucoma drugs, the corneal posterior surface shape might have been changed. Following this, the change of refractive state was affected. Our study also found that 3 months after drug treatment withdrawal, the refractive status remained relatively stable in most patients. However, regression in a few patients occurred again, accompanied by epithelial thickening and a decrease in visual acuity. Three patients with FS-LASIK needed enhancement surgery, and one patient undergoing TPRK needed surgical retreatment (data not reported). The reason for this difference is not clear. It was speculated that this difference is related to a higher decrease of biomechanical stability after FS-LASIK, suggesting the necessity for combining collagen cross-linking when correcting high myopia, especially in patients with high myopia and lower corneal thickness preoperatively [26, 27].
In addition to corneal epithelial hyperplasia and thickening, various studies consider that refractive regression is related to the protrusion of the posterior surface of the cornea [29–31]. Refractive regression can be treated with anti-glaucoma drugs by preventing the protrusion of the posterior surface [32, 33]. During clinical practice, topical steroidal agents and anti-glaucoma drugs were often used together to treat refractive regression, which could bring better therapeutic results because anti-glaucoma drugs could also reduce corneal epithelial thickness [34, 35]. However, in this study, to analyze the sole role of epithelial thickening during refractive regression, anti-glaucoma drugs were not used to prevent the protrusion of the posterior surface. Anti-glaucoma drugs were used in four eyes after the intraocular pressure increased by more than 5 mm Hg during steroid treatment. Meanwhile, we did not find that topical steroidal agents affected the posterior surface of the cornea. Therefore, the use of anti-glaucoma drugs in the four eyes would not affect the accuracy of the conclusions in this study.
There were limitations in our study. No control group not treated with drugs was designed. We could not confirm how long the refractive regression would last and whether it could subside by itself without intervention. In clinical practice, besides enhancement surgery, the main treatment of refractive regression was steroidal agents and anti-glaucoma drugs. For patients with refractive regression affecting visual quality, it was unreasonable not to prescribe drug treatment, which would harm the patients' benefit and cause patients' dissatisfaction. Therefore, in this study, an untreated control group with refractive regression was not designed. Compared with preoperative morphology, the corneal epithelium was indeed thickened with the central regions thickening more, which was different from that without regression in previous study. Furthermore, after steroidal treatment, the epithelium became thinner with the decreasing of curvature in corneal anterior surface, refractive error drifting to positive direction, regression subsiding and other ocular variables remained unchanged. It was enough to verify the purpose of this study, that was, the epithelial thickness played an important role in refractive regression, regardless of what factors leading to changes of epithelial thickness.