Preliminary Clinical Study of a LenSx Femtosecond Laser-Assisted Limbal Relaxing Incision for the Correction of High Myopia with Low to Moderate Astigmatism in Posterior Implantable Collamer Lens Implantation

Purpose. To evaluate the safety, efficacy, and predictability of implantable collamer lens (ICL) implantation combined with a LenSx femtosecond laser-assisted limbal relaxing incision (LRI) for the correction of corneal astigmatism. Methods. This prospective study enrolled 64 eyes (54 patients) with high myopia with low to moderate regular corneal astigmatism. They were divided into an ICL group with ICL implantation (18 patients, 20 eyes), a TICL group with toric ICL implantation (17 patients, 23 eyes), and a LenSx + ICL group with a LenSx femtosecond laser-assisted LRI and an ICL implantation (19 patients, 21 eyes). Visual acuity, astigmatism correction ability, and visual quality were measured before and 1, 3, and 6 months after surgery. Results. The postoperative visual acuity of the 3 groups was higher than the preoperative visual acuity (
 
 P
 <
 0.01
 
 ), and the improvements in the LenSx + ICL group and the TICL group were greater than those in the ICL group (
 
 P
 <
 0.01
 
 ). The LenSx + ICL and TICL groups had less residual astigmatism and a higher astigmatism correction index (CI) than the ICL group (
 
 P
 <
 0.01
 
 ). There was no significant difference among the three groups in total high-order aberrations (HOAs) before and after surgery (
 
 P
 >
 0.05
 
 ). Conclusion. LenSx femtosecond laser-assisted LRI can effectively correct low to moderate corneal astigmatism during ICL implantation surgery. It can achieve similar clinical effects in the short term compared with TICL implantation.


Abstract Background
To evaluate the safety, e cacy, and predictability of the LenSx femtosecond laser-assisted limbal relaxing incision (LRI) for correction of high myopia with low to moderate corneal astigmatism (-0.75 to -1.50D) in implantable collamer lens implantation (ICL).

Methods
This prospective clinical control study included 56 eyes from 46 patients. They were divided into an ICL group with ICL implantation (14 patients,16 eyes); a TICL group with toric ICL implantation (17 patients, 23 eyes); and a LenSx + ICL group with a LenSx femtosecond laser-assisted LRI and an ICL implantation (15 patients, 17 eyes). The visual acuity, astigmatism correction ability and visual quality were measured before and 1, 3, and 6 months after surgery.

Results
The postoperative visual acuity of the 3 groups were higher than the preoperative visual acuity (P < 0.01), and the improvements in the LenSx + ICL group and the TICL group were greater than those in the ICL group (P < 0.01). The LenSx + ICL and TICL groups had less residual astigmatism and higher astigmatism correction index (CI) than the ICL group (P < 0.01). There was no signi cant difference in the three groups of total high order aberrations (HOAs) before and after surgery (P > 0.05).

Conclusion
LenSx femtosecond laser-assisted LRI combined with ICL implantation can effectively correct low to moderate corneal astigmatism in high myopia. Compared with TICL implantation, it can achieve similar clinical effects in the short term.

Background
In recent years, implantable collamer lens (ICL) implantation has become the mainstream of intraocular refractive surgery due to its effectiveness, safety and predictability. Ophthalmologists often meet patients with high myopia combined with astigmatism of -1.50D or less in clinical work. For such patients, there are usually two solutions. One way is to correct the astigmatism with a TICL implantation. However, the lens may rotate in the ciliary sulcus which will cause the decrease of astigmatism correction or even create a new astigmatism [1]. Besides this, the patients have to wait a relatively long time for surgery. The other way is a simple ICL implantation without astigmatism correcting, but long-term follow-up has shown that patients with uncorrected astigmatism often fail to achieve the best corrected visual acuity, accompanied by a decline in visual quality, especially in contrast sensitivity and night vision. Some scholars have attempted to make limbal relaxing incisions (LRIs) on the cornea to correct the astigmatism and achieved certain results [2,3]. However, it is di cult to accurately control the length, position and depth of the manual incision, and inappropriate location may cause di culty for the surgeon [4].
In 2009, femtosecond laser-assisted cataract surgery (FLACS) was invented. In addition to controlling the femtosecond laser for corneal incision, lens capsule release, and lens nucleus fragmentation, some scholars have tried to resolve corneal astigmatism with the help of femtosecond laser-assisted LRIs and achieved ideal results [5][6][7][8]. So we think about whether we can apply the FLACS technique to ICL implantation surgery, and use the femtosecond laser to make LRIs to correct the low to moderate corneal astigmatism before ICL implantation. This can avoid the shortcomings of the manual incision, and prevent the problem in TICL implantation. We innovatively combined femtosecond laser-assisted corneal LRIs with traditional ICL implantation to correct low to moderate astigmatism and to verify the safety, e cacy, and predictability of this method.

Methods
From January 2019 to May 2019, 46 patients amomg 18-44 years old with high myopia among − 6.00D to -14.75D and low to moderate astigmatism among − 0.75D to -1.50D were recruited from the Chongqing Aier-Mega Eye Hospital (Chongqing, China). According to the principles of random selection and voluntary participation, the patients were divided into 3 groups, including 16 eyes of 14 patients in the ICL group, 23 eyes of 17 patients in the TICL group, and 17 eyes of 15 patients in the LenSx + ICL group.
There were no statistically signi cant differences among the 3 groups before operation ( Table 1). The inclusion criteria for the three groups were as follows: stable refractive error (≤ 0.50D change of refractive error in the past 2 years), anterior chamber depth (ACD) of 2.80 mm or more, and an endothelial cell density greater than 2000 cells/mm 2 . Note:There was no statistical difference among the three groups before operation.
The uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), residual astigmatism, and correction index (CI) were collected before and 1, 3, and 6 months after surgery. The iTrace analyzer measured and recorded the total high-order aberrations (HOAs) and modulation transfer function (MTF) value to observe the astigmatism correction and visual quality of the patients after surgery. Vector analysis was performed, utilizing the Alpins vector method with ASSORT software (ASSORT Pty; Cheltenham, Victoria, Australia) to show the magnitude and directionality of the residual astigmatism. According to the conventional ICL implantation method, all patients dropped Levo oxacin before operation (4 times each day, 3 days in total). For the TICL group and the LenSx + ICL group, the astigmatism axis was marked on the eye 1 hour before the surgery while the patient was in a sitting position at the slit lamp. In the ICL group, after surface anesthesia, the surgeon made a 3 mm main incision at the scleral limbus, then slowly implanted the lens into the anterior chamber and used a special hook to adjust the lens into the ciliary sulcus under the iris to ensure a centered position of the ICL. The postoperative eye was treated with a special ointment and covered well. In the TICL group, the surgeon simply rotated the lens according to the astigmatism axis provided by the STAAR company after implantation. In the LenSx + ICL group, astigmatism axis locations was determined by using an online software (Abbott Medical Optics, USA; available at http://www.lricalculator.com). Individual surgeon's surgically induced astigmatism was considered as 0.5D (Fig. 1). Then the surgeon marked the astigmatism axis on the patient's eye. The computer automatically calculated the position, depth, and arc length of the limbal relaxing incision according to the vector calculator. The depth of the limbal relaxing incision (the setting depth is 85% of the corneal thickness) was measured under the guidance of an anterior optical coherence tomography segment. The laser was started, and the vacuum was released after the laser operation was over. The vacuum ring was then removed, the patient was moved to a conventional operating table, and the remaining steps were completed according to the conventional ICL surgery procedure.
All the operations were performed by an experienced surgeon without any complications.

Statistical Analysis
All statistical analyses were performed using SPSS (version 19; SPSS Inc). Figures were drawn using GraphPad Prism (version 8; GraphPad Software). Data are presented as mean ± standard deviation.
Student's t-test was used for normally distributed data, and the Mann-Whitney test was used for nonnormally distributed data. One-way ANOVA was used to analyze the differences among the 3 groups of different operative methods, and LSD was used to perform additional multiple comparison tests. P < 0.05 was considered statistically signi cant.

Surgical Complications
The postoperative follow-up found that in the LenSx + ICL group, there was only a short arc-length relaxing incision left under the cornea after the surgery (Fig. 2). There was no bleeding or in ltration of the incision, the tightness was good, and no complications (such as infection or poor healing of the incision) occurred.

Visual Acuity
The UDVA and CDVA before and after surgery in each group were showed in the Table 2. Figure 3 shows that the UDVA in the 3 groups 1, 3, and 6 months after surgery were higher than the CDVA before operation, and the differences were all statistically signi cant (P < 0.01). The improvements between postoperative UDVA and preoperative CDVA were compared between each group (Fig. 4). The improvements in the LenSx + ICL group and the TICL group were higher than those in the ICL group at 1, 3, and 6 months after surgery (P < 0.05). Further comparison was made between the improvement of the LenSx + ICL group and the TICL group. At 1, 3, and 6 months after surgery, the mean improvements in the TICL group were higher than those in the LenSx + ICL group, but there was no statistically signi cant difference between the two groups (P > 0.05).
We further compared the CDVA before and after surgery in the 3 groups (Fig. 5). At 1, 3, and 6 months after operation, the CDVA were improved compared with those before surgery, and the differences were statistically signi cant (P < 0.05).

Residual Astigmatism and CI
Residual astigmatism was shown using the astigmatism double-angle plot of the Alpins vector analysis method ( Fig. 6-7). Compared with the ICL group, the LenSx + ICL group and the TICL group had smaller residual astigmatism at 1, 3, and 6 months after operation, and the differences were statistically signi cant (P < 0.01). Further comparison was made between the LenSx + ICL group and the TICL group, and the residual astigmatism of the LenSx + ICL group was slightly lower than that of the TICL group at 1 month after surgery, whereas that of the TICL group was lower than that of the LenSx + ICL group at 3 and 6 months after surgery; however, these differences were not statistically signi cant (P > 0.05). Table 3 and Fig. 8 showed the CI (CI = |SIA|/|TIA|) values for the 3 groups at 1, 3, and 6 months after operation.
Among these, the LenSx + ICL group and the TICL group had higher CI than the ICL group, and the differences were statistically signi cant (P < 0.01). However, there was no statistically signi cant difference between the LenSx + ICL group and the TICL group after surgery (P > 0.05). None of the total HOAs in the 3 groups before or 1, 3, or 6 months after surgery showed statistically signi cant differences (P > 0.05) ( Table 4 and Fig. 9). Note:Pre-=preoperative,Post-=postoperative,Means ± SD Table 5 and Fig. 10 show that there were no statistically signi cant differences in the MTF values among the 3 groups before the operation at pupil diameters of 3 mm or 5 mm (P = 0.7812, P = 0.8544). The comparison among the 3 groups after surgery showed that the MTF values of the TICL group were always slightly higher than those of the LenSx + ICL group, but the difference was not statistically signi cant (P > 0.05). The MTF values in the TICL group were always higher than those of the ICL group, and the differences were statistically signi cant (P < 0.01). Except for the condition of a 3 mm-diameter pupil at 6 months after surgery, the MTF values in the LenSx + ICL group were always better than those in the ICL group, whether at 1, 3, or 6 months after the operation (P < 0.05).

Discussion
Astigmatism is one of the common types of refractive error in the eye. Previous studies have reported that nearly 80% of individuals with refractive errors have astigmatism of varying degrees, and about 25% of individuals have astigmatism greater than − 1.00D [9]. Astigmatism over − 0.75D can cause diplopia, blurred vision, and photophobia [10]. STARR company launched TICL implantation for astigmatism correction, however, the lens may rotate in the ciliary sulcus after implantation, which may lead to a decrease in astigmatism correction or create a new astigmatism. Patel C [11] showed that an axial deviation of 10 degrees in TICL can only correct 2/3 of the estimated astigmatism, and an deviation of 30 degrees has no correction effect at all. If the deviation exceeds 30 degrees, there will be symptoms such as diplopia, dazzle, and decreased visual acuity, which have a signi cant impact on patients [12][13][14]. Some scholars have tried to make LRIs to solve the corneal low to moderate astigmatism in ICL implantation [15]. However, the traditional LRIs are performed manually, it is di cult to achieve accurate length, position, and depth of the incision, which often makes the astigmatism correction effect unpredictable [4]. In cataract surgeries, some scholars have tried to solve corneal astigmatism with the help of femtosecond laser-assisted LRIs and achieved ideal results. This study compares the traditional ICL implantation, TICL implantation, and femtosecond laser-assisted LRIs with ICL implantation to assess the safety, e cacy, and predictability of this new approach to treating astigmatism, and we hope to propose a new treatment for the correction of low to moderate astigmatism in ICL surgery.
This study found that the UDVA of the 3 groups were improved compared with the preoperative CDVA. It showed that all 3 studied methods can improve the visual acuity in a certain way. Further comparison of the improvements between postoperative UDVA and preoperative CDVA showed that compared with the ICL group, both the LenSx + ICL group and the TICL group had better visual improvement, and the LenSx + ICL group achieved similar corrected visual acuity to that of the TICL group.
Compared with the ICL group, the residual astigmatism was smaller in the LenSx + ICL group and the TICL group after surgery, which means that the LenSx + ICL group and the TICL group experienced better astigmatism correction than the ICL group. Although the residual astigmatism between the LenSx + ICL group and the TICL group displayed some differences, these were not statistically signi cant. To make further comparisons about the astigmatism correction capabilities of the 3 methods, CI values were introduced. CI = 1 means that the expected astigmatism was completely corrected, CI < 1 indicates under correction, and CI > 1 indicates over correction. These CI values were less than 1 in all 3 groups, indicating that all 3 groups were under corrected after surgery. Among these, the LenSx + ICL group and the TICL group had higher CI values than the ICL group, and just like the residual astigmatism, there were no statistically signi cant differences between the LenSx + ICL group and the TICL group. We found that the CI values in the LenSx + ICL group became smaller, just as the residual astigmatism values became higher. It is considered that during long-term corneal incision healing in the LenSx + ICL group, the diopter of the astigmatism is increased, which indicates astigmatism regression and thus may result in a slight decrease in visual acuity and visual quality. This is consistent with the research by Tetikoğlu [16], which found that the healing process of surgical incision directly affects the corneal astigmatism and thus affects the change in corneal aberration.This also suggests us that we can further explore the problem of astigmatism regression caused by corneal incision healing, and adjust the nomogram so as to obtain a more stable and longer-term astigmatism correction effect.In recent years, refractive surgeons have aimed not only to improve the visual acuity of the patient but also to improve the patients' visual quality through a series of personalized operations. Myopia and astigmatism are low-order aberrations (LOAs) and cannot truly re ect the visual quality of patients. Therefore, we collected and studied the HOAs of 3 methods to analyze the visual quality of postoperative patients. We used the iTrace visual quality analyzer to analyze the patients' visual quality before and after surgery. The iTrace analyzer uses optical path tracing technology, and the visual quality of the patient can be simulated by computer software. The iTrace analyzer can convert the point spread function to a Fourier transformation to obtain the MTF curve. At present, many scholars [17] use iTrace analyzer to compare the visual quality of patients between femtosecond laser-assisted cataract surgery and traditional phacoemulsi cation surgery. In this study, the researchers selected 2 major indicators of visual quality: total HOAs and MTF value. Lower HOAs and higher MTF values indicated better visual quality. This study collected patients' total HOAs and MTF values with a pupil diameter of 3 mm and 5 mm before and after surgery. The total HOAs in the 3 groups before and 1, 3, and 6 months after surgery displayed no statistically signi cant differences, indicating that no new HOAs were introduced, regardless of the method used. This is consistent with some earlier research, including that of Zhou Xingtao et al., [18] who used the OQAS objective visual quality analysis system to compare the visu al quality of wavefront aberration-guided LASIK surgery and ICL implantation and concluded that the ICL implantation introduced no new aberrations and achieved better visual quality than the LASIK operation. In another long-term comparative observation of 3 years, Zhou Xingtao et al. [19] also found that compared with traditional refractive surgery, ICL implantation offers more stable visual acuity and smaller aberrations. This study found that the MTF values of the TICL group were always slightly higher than those of the LenSx + ICL group; however, the differences were not statistically signi cant. In comparison to the ICL group, the LenSx + ICL and TICL groups had higher MTF values, except for the comparison between the LenSx + ICL and the ICL group under the 3 mmdiameter pupil in 6 months after the surgery. The results show that, compared with the ICL implantation, the visual acuity obtained in the LenSx + ICL group was obviously superior, and the visual quality of the patients was better. In addition, the LenSx + ICL group achieved a similar effect compared with that of TICL implantation.
Postoperative follow-up found that in the LenSx + ICL group, only a short arc-length incision was left under the cornea after the surgery. There was no bleeding or in ltration of the incision, and the postoperative tightness was good. There were no complications, such as infection and poor healing of the incision. Masket et al. 7 also showed that femtosecond laser-assisted corneal incisions are much neater and tighter than manual corneal incisions. Studies by Chung SH et al. [20] also showed that femtosecond laser-assisted cataract phacoemulsi cation surgery yields good watertightness and neatness in the corneal incision. Traditional manual corneal incisions are relatively poorly sealed, and this inevitably results in edema of the surrounding tissues during the post-operative period, causing an increase in high-order aberrations after surgery. In addition, the postoperative CDVA in all 3 groups were better than the preoperative CDVA, and the differences were statistically signi cant. This shows that whether the traditional ICL implantation or the femtosecond assisted ICL implantation was used, there was no loss in CDVA, which re ects the safety of the surgery and suggests that for some patients with amblyopia, this surgical method can correct visual acuity to a degree that is unobtainable with ordinary frame glasses, thus improving patient quality of life.

Conclusion
In summary, this study provides a new method to resolve high myopia combined with low to moderate astigmatism to a certain extent, and it preliminarily veri es the safety, e cacy, and predictability of this method of astigmatism correction. Femtosecond laser-assisted LRI combined with ICL implantation may be an alternative because of time constraints in the TICL lens. In addition, it is probable that the femtosecond parameters (such as incision depth and arc length) can be further optimized to obtain stable, long-term astigmatism correction. However, due to ethical considerations, this study did not set up a manual LRI combined with an ICL implantation group, and the number of patients included were limited. We plan to compare the manual LRI and the femtosecond LRI in the next step. Therefore, a larger samples and a longer follow-up time are needed to further verify the surgical effects.  The improvement between postoperative LogMAR UDVA and preoperative LogMAR CDVA in three groups.

Abbreviations
(The results are expressed as the means ± SD, * means P<0.05, ** means P<0.01) Figure 5 The LogMAR CDVA before and after surgery at 1, 3 and 6 months in three groups. (The results are expressed as the means ± SD, ** means P<0.01, *** means P<0.001) Figure 7 The comparison of postoperative residual astigmatism in three groups at 1, 3 and 6 months. (The results are expressed as the means ± SD, *** means P<0.001)