The techniques of refractive surgery are evolving with rapid, ongoing advances in research. It has been established that a treatment zone centered at the visual axis, instead of the pupillary axis, is the key to optimizing the visual outcome of a refractive surgery [12, 13]. At present, in corneal refractive surgery, the cutting of the corneal center is divided into four distinct methods, namely, the center of the pupil (PC) [14], CSCLR [15], corneal vertex normal (CVN) [16, 17], and between the pupillary and visual axis [18]. However, there is controversy regarding optimal centration in corneal refractive procedures. In other words, in the virtual image of the light source, also known as the first Purkinje-Sanson image, the corneal light reflex is formed by the reflection of light from the anterior corneal surface. A number of researchers have postulated that the CSCLR from the cornea lies closer to the corneal intercept of the visual axis than the PC and have thus recommended using the corneal CSCLR as the center of the refractive surgery [19]. However, the CSCLR was used as the surgical center for all patients in the present retrospective study, with adjustment of the angle kappa, and all patients were satisfied with the results. In addition, SMILE showed good safety, efficacy, and predictability.

HOAs are optical imperfections of the eye that alter the quality of the retinal image despite optimal correction of spherical defocus and astigmatism. Since HOAs reduce the retinal image quality and produce variations in optical vergence across the entrance pupil of the eye, they may provide optical signals that contribute to the regulation and modulation of eye growth and refractive error development. The third-order coma and trefoil and fourth-order spherical aberration constitute the major components of HOAs, with the coma and trefoil aberrations considered variations of magnification with respect to the aperture. If the paraxial magnification is equal to the real ray marginal magnification, then an optical system would be free of coma or trefoil. Spherical aberration can be considered a variation of the focal length with the aperture. These aberrations can cause difficulty in seeing at night, glare, halos, blurring, starburst patterns, or double vision (diplopia). In this report, all HOA parameters were compared before and after SMILE by Pentacam, and all exhibited significant differences except trefoil (although with longer follow-up times, the statistical difference gradually decreased). This result indicated that trefoil could be stabilized over time.

Several prior studies [20, 21] have indicated that the angle kappa decreases with age, although the changes in its direction or value are not significant. Hashemi [22] reported that the angle kappa decreased by about 0.015 per year increase in age. In this report, we divided the angle kappa into three distinct groups, namely, r < 0.1 mm, 0.1 ≤ r < 0.2 mm, and r ≥ 0.2 mm, finding that there was no significant difference between the groups with age. The reason can be attributed to the fact that the inclusion criteria included age (mean 23.377 ± 6.816); therefore, a valid comparison of age could not be made. Secondly, there were significant differences in SA and OT before and 12 months after SMILE, which implies that the early differences in HOAs after SMILE surgery were not significant in patients with different angle kappa, while some may change after 12 months, especially SA in the group with the larger angle kappa. This may have been caused primarily by the aspherical reconstruction of the corneal surface after partial corneal central cutting. Thirdly, there were no significant differences between the groups in terms of DC, DS, and SE, a finding that was surprisingly quite different from those of previous reports. For instance, Basmak [23] found that the angle kappa decreased significantly with increasing DS. Ding [24] reported that a large angle kappa was observed mostly in patients with medium and low myopia, while a small angle kappa was mostly associated with high myopia, as patients with high myopia often need to move close to objects to see them clearly. Clinically, these changes show a significant shortening of the near point distance, and the eye three-linkage effect of the near-reflection system could be stimulated (accommodation, convergence, and pupil shrink) to maintain monocular function in both eyes. In addition, the binocular convergence function must be strengthened, i.e., the inward rotation of the eyeball, so that high myopia can exhibit a small angle kappa. The reasons for the different conclusions in this study can be possibly attributed to insufficient sample size, together with the predominance of patients with middle and low myopia, with patients with high myopia accounting for only 21%, which may have caused statistical bias and led to this peculiar trend.

To better understand whether the orientation of the angle kappa affected the postoperative HOAs, the angle kappa values were divided into four distinct groups according to quadrant (Table 4). The result showed that there were no statistically significant differences between the quadrants. However, a few changes were observed which may have been due to wound healing and the postoperative reestablishment of the aspheric vertices on the corneal surface, resulting in alteration of the angle kappa [24]. At the same time, we drew a scatter plot of the angle kappa distribution (Fig. 1), which showed that the angle kappa was mainly concentrated in the first and second quadrants, as the fovea was located on the pupil axis and slightly temporal to the posterior pole, resulting in the normal angle kappa being slightly positive.

We performed correlation analyses on all preoperative and postoperative parameters. The results showed that higher degrees of myopia were associated with greater changes in the total HOA, SA, and coma. We also found that patients with large angle kappa values showed greater increases in SA after the operation; this may have been due to postoperative aspherical changes in the corneal surface, forming a new off-axis center and altering the postoperative angle kappa. The results showed significant differences in SA and VC between the preoperative and early postoperative periods, which can be explained by the technique itself [25]. The patients’ side cuts were designed in the vertical direction, and healing of the surgical incision and the resultant scar formation could have led to an increase in the VC. Thus, changes in asphericity and the gradient refractive index of the corneal surface after SMILE can potentially cause an increase in SA.

We performed a tracking trend analysis of HOAs in the different groups of patients. The results showed the greatest changes in all HOAs during the first month, after which the trend gradually slowed; however, no significant differences were observed between the three groups. Although increases in HOAs were seen in all patients, the patients nevertheless showed excellent visual quality, with good postoperative refractive outcomes obtained in all groups, which may be because the angle kappa is generally small in myopic eyes and had been adjusted. Thus, the postoperative decentration in the eyes of the large angle kappa group was not sufficient to affect the visual and refractive low-order aberration outcomes. HOAs can only affect patients in certain situations, such as having larger pupils at night, potentially leading to the experience of glare, diplopia, and reduced contrast sensitivity. In 2003 [26], a case report of a patient with a large bilateral angle kappa provided the first direct comparison of centering on the entrance pupil versus centering on the CSCLR. The patient had undergone LASIK centered on the entrance pupil in the right eye and over the CSCLR in the left eye. The left eye demonstrated significantly better visual acuity, smaller refractive error, and a smaller amount of postoperative decentration. Kermani [18] did report a significant decrease in the total HOA in the CSCLR group, and this effect was not noted in the pupil-centered group. Moreover, in the same year, Okamoto [27] compared myopic LASIK centered on the CSCLR with centration on the entrance pupil in 556 eyes with unknown angle kappa values. LASIK centered on the CSCLR was found to be significantly safer, more effective, and had lower induction of coma, as well as total HOA, in comparison to LASIK treatments centered on the pupil. However, in 2011, Soler [28] published the only randomized double-masked comparison of the pupil-centered vs. corneal reflex-centered hyperopic LASIK, and the findings of this study concluded that there were no statistical differences. In 2013, Reinstein [19] compared different angle kappa values using the CSCLR treatment, finding no significant differences in safety, accuracy, induced astigmatism, contrast sensitivity, or night vision disturbances after the surgery. These findings are consistent with our results. However, these authors did not conduct further comparative analysis of HOAs, and the results of our report thus provide the clinical basis for our next prospective study. This is primarily because the size of the angle kappa showed no significant effect on the result following the CSCLR treatment, and considering the accuracy of the small kappa angle correction, later prospective studies may consider including large angle kappa for further investigation.

The small angle kappa in SMILE surgery may not need to be adjusted, but there are very few studies describing the correction of the angle kappa after SMILE; nevertheless, large angle kappa must be adjusted to obtain better visual quality. This report indicated that patients with large angle kappa were able to increase VC and SA due to the aspheric and gradient refractive indices of the corneal surface, as well as healing of the early incision and scar formation. Although all patients had increased HOAs after SMILE, the procedure still showed excellent safety, efficacy, and predictability. In this study, the rate of patient follow-up declined between 6 months and 1 year after surgery, which may have had a certain influence on the results. Nevertheless, based on the surgical treatment and observation standard [29, 30], the patients were essentially recovered and stable within 3 months after surgery and drug treatment had ceased. In addition, to avoid missing patients with poor postoperative results, we conducted long-term telephone and network follow-ups, in which none of the patients reported visual-related discomfort or abnormalities. Therefore, the results of this study have some reference value.

This is the first study comparing the effects of the different quadrants of angle kappa on HOAs. Although the different quadrants showed some differences in HOAs, these were non-significant. It is possible that due to the corneal surface reestablishment and change of diopter of myopia after SMILE surgery, the three-linkage effect of the near-reflection system was reduced, and thus, the angle kappa was reconstructed as well as altered; however, this trend was not apparent. The current study was limited by its design and small sample size. We also did not evaluate subjective visual quality and symptoms, such as contrast sensitivity, halo, and glare, in our patients. Thus, further studies with larger sample sizes and diopter differences are still needed, which can also evaluate the potential effect of angle kappa adjustment on light scattering, visual quality, and contrast sensitivity.