The comparison of visual outcomes, aberrations, and Bowman’s layer micro-distortions after femtosecond laser small-incision lenticule extraction (SMILE) for the correction of high and moderate myopia and myopic astigmatism

Femtosecond laser small-incision lenticule extraction (SMILE) is a newly developed refractive surgical technique, which utilizes femtosecond laser to cut no corneal flap but a stromal lenticule and extracted through a small side cut [1‐3]. Many existing studies have found that compared with laser in situ keratomileusis (LASIK), SMILE have similar surgical outcomes in terms of safety, efficacy, predictability, and corneal higher-order aberrations [4‐6].

Standard SMILE procedure is capable of correcting myopia up to − 10 diopters (D) using Visumax® femtosecond laser system. With the recent software upgrade, it also made myopic correction of over − 10 D possible. The aim of this study was to compare the visual and refractive outcomes, higher-order aberrations, and corneal morphological change of SMILE for the treatment of myopia and myopic astigmatism of > − 10 D with those ≤ − 10 D performed by the same experienced surgeon.

Sixty patients were consecutively enrolled in this prospective study. All patients underwent routine preoperative examinations and met the surgical indications for SMILE. The SMILE procedures were conducted in the refractive surgery center of the department of ophthalmology, Eye and ENT Hospital of Fudan University (Shanghai, People’s Republic of China) from October 2015 to February 2016. Inclusion criteria are: ages between 18 to 40 years old, a corrected distance visual acuity (CDVA) of 20/25 or better (≥0.1 logMAR), a stable refractive error (≤0.50 D of refractive error change in the past 2 years), and no contact lens use for 2 weeks prior preoperative examinations. Exclusion criteria include keratoconus or suspicion of keratoconus, severe dry eye, corneal scar, and a history of herpetic keratitis, cataract, glaucoma, retinal pathology, ocular surgery, or any systemic disease. Eyes with a calculated postoperative residual stromal bed thickness of less than 250 μm were also excluded.

A random eye of each patient was selected for data analyses. Patients were divided into group 1 and group 2 according to preoperative manifest refraction. In group 1 (30 eyes of 30 patients), all eyes have the sum of spherical refraction and astigmatism of − 10 D or less. In group 2 (30 eyes of 30 patients), all eyes have the sum of spherical refraction and astigmatism of over − 10 D, as well as the spherical refraction of − 7.5 D or more. Demographic, preoperative and surgical information were summarized in Table 1. This study adhered to the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of the Eye and ENT Hospital, Fudan University. All patients signed their informed consent after a detailed explanation of the risks and potential outcomes of refractive surgery and the study.

It is widely accepted that the standard small-incision lenticule extraction procedure is capable of treating eyes with spherical error of no more than − 10 D, and that SMILE has shown promising results in terms of visual outcomes, safety, efficacy and predictability [4, 10‐12]. The modified Visumax® femtosecond laser system in this study was capable of treating spherical error up to − 14 D. In our study, we presented preliminary results comparing eyes with myopia and myopic astigmatism of ≤ − 10 D and > − 10 D undergoing SMILE procedure.

In terms of safety, there was no statistical difference upon comparing the safety indices between two groups. Previous studies reported the safety indices of SMILE up to − 10 D, being 1.01 ± 0.05 [13] and 1.09 [14] at 3 and 1 month postoperatively, which was similar to our result of 1.04 ± 0.10 in group 1 and 1.10 ± 0.16 in group 2. Several studies reported that the low to moderate myopic eyes treated with SMILE had the rate of CDVA loss for one or more lines of 1.1% [1] and 1.67% [13]. In our study, no eyes lost l or more lines of CDVA. For efficacy comparison, previous results regarding SMILE up to − 10 D reported the efficacy indices of 1.04 ± 0.20 [13] and 1.09 [14] at postoperative 3 and 1 month respectively. Our results of 1.00 ± 0.20 in group 1 and 1.03 ± 0.18 in group 2 were consistent with these results, also showing no statistical difference between the two groups. For previous postoperative results of SMILE up to − 10 D, 62% had an UDVA of 20/20 or better, and 93% of 20/25 or better [15‐18]. In our study, 94% (group1) and 63% (group2) eyes had an UDVA of 20/20 or better, and 100% (group1) and 97% (group2) had an UDVA of 20/25 or better. As for predictability, studies of SMILE in treating preoperative SE of ≥6.00D myopic eyes reported a predictability of 80% [19] and 88% [20] within ±0.50D of target refraction, and of 94% [19] and 97% [20] within ±1.0D, respectively. These results were consistent with ours of 83% (group 1), 80% (group 2) within ±0.50D, and 100% (group 1), 97% (group 2) within ±1.0D (Fig. 1e). Our results showed no statistical difference of UDVA and CDVA between group 1 and group 2. All these results suggested that SMILE in correcting myopia and myopic astigmatism of > − 10 D had similar results compared with those ≤ − 10 D in terms of safety, efficacy, and predictability. For corneal wavefront aberration analyses, HOA RMS and coma increased in both groups, but spherical aberration only increased in group 2 and trefoil did not increase in either group preoperatively and 6 months postoperatively. Previous literatures showed conflicting results regarding changes in HOA before and after the surgery. Some studies indicated that HOA RMS, spherical aberrations, and coma increased postoperatively [20, 21]. Others found that only HOA RMS and coma increased, while spherical aberration remained stable [7, 9, 22]. There might exist several reasons for such discrepancies. First, different instruments for measurements and different scotopic environments could affect the results. Second, aberrational data were evaluated under different pupil sizes. Using a smaller pupil size for analysis might lose some information. But in some extreme myopic eyes, scotopic pupil size could be smaller than the intended pupil size for analysis, hence using a smaller analytical pupil size was inevitable. In this study, we found spherical aberration being higher in group 2 compared with group1 6-month postoperatively. This might be due to the fact that a smaller designated optical zone was applied for higher myopic correction. Moreover, a thicker stromal lenticule needed to be excised, which caused a larger change in anterior corneal asphericity and induced more spherical aberration. Previous PRK and LASIK studies also indicated that the optical zone had a significant impact in spherical aberrations [23‐26]. Further studies should be conducted to explain the spherical aberration increase for high level myopic correction, as well as its potential influence on visual quality such as contrast sensitivity and intraocular scattering. For coma increase, the SMILE platform currently lacks a tracking system. Minor decentration was inevitable, which might have minimal effect in postoperative visual acuity but could potentially induce coma.

In previous literature, the longest follow-up time for Bowman’s layer micro-distortions observation after SMILE surgery was 3 months, which showed that the total number of micro-distortions decreased overtime [27]. In this present study, we further extended the observation time to 6 months postoperatively, finding that micro-distortions still existed but the total number remained stable for 3 and 6 months postoperatively. This suggested that the corneal shape became stable after certain time postoperatively, and the micro-distortions could still remain. More micro-distortions were found in group 2, and the total number of micro-distortions was associated with preoperative SE and refractive lenticule thickness. These findings suggested that patients with higher myopic correction tend to have more micro-distortions. The results further confirmed certain previous findings [8, 9], which bring out the hypothesis that micro-distortions might be due to mechanical changes in cornea and greater difference in curvatures between stromal bed and corneal cap after the extraction of thicker lenticule. The total number of micro-distortions were not correlated with UDVA or CDVA. This is one of the evidence adding to the efficacy, safety and the fast visual recovery of the SMILE procedure [2, 3, 28, 29]. On the OCT images, although Bowman’s layer became rugged with micro-distortions after SMILE, the epithelium had uneven thickness but a smooth surface to ensure good refractive quality of the whole cornea (Fig. 3). Although the smooth corneal surface could explain the good visual acuity after SMILE surgery, further investigations should be conducted to reveal the potential impacts of micro-distortions in visual quality.

Our study has several limitations. We presented a comparatively small sample and the follow-up time is short. We did not analyze the vertical and horizontal coma separately when doing HOA coma analysis. Also, we did not compare the procedures with other types of surgeries such as LASIK. We are currently conducting a further study with larger sample and different types of surgeries comparisons.

In conclusion, our results demonstrated that SMILE for the treatment of myopia and myopic astigmatism of > − 10 D were good in visual outcomes in terms of safety, efficacy, predictability, and stability while having slightly under-correction compared to those ≤ − 10 D. Higher myopic treatment tends to induce more spherical aberrations. No vision-threatening complications occurred throughout the 6-month follow-up period. Micro-distortions in Bowman’s layer were associated with preoperative SE and lenticule thickness, but had no impact on visual and refractive outcomes. This procedure seems promising in terms of extending the indications of SMILE.