Evaluation of the Long-Term Corneal Biomechanics Following SMILE With Different Residual Stromal Bed Thickness in Rabbits

Abstract

Purpose: The long-term safety of small-incision lenticule extraction (SMILE) surgery for correcting high myopia with a residual stromal bed thickness (RST) less than 50% of the central corneal thickness (CCT) was investigated from a biomechanical perspective.

Methods: Twelve rabbits were enrolled in this study, in which the right eye underwent SMILE surgery (the corneal cap thickness was one third of the preoperative CCT, approximately 120 µm), and the left eye served as the control. The rabbits were evenly divided into two groups, with the RST of 50% and 30% of the preoperative CCT in each group, respectively. Corneal morphology parameters in vivo and biomechanical properties in vitro were determined at 11 months after SMILE surgery. Moreover, the distribution of displacement and Von Mises stress across the cornea were evaluated using finite element analysis.

Results: At 11 months after SMILE surgery, there was no obvious forward shift in the posterior corneal elevation and no significant increase in the mean keratometry from the posterior corneal surface. The elastic modulus of the RST 30% group increased significantly compared to the control, although no significant differences were found in the creep rate, elongation rate, and equilibrium modulus among all groups. Compared to the control, the displacement was smaller in the RST 30% group, while the maximum stress was obviously higher.

Conclusions: No iatrogenic keratectasia occurred in a rabbit model of high myopic SMILE surgery with an RST of less than 50% of preoperative CCT at 11-month follow-up.

Translational Relevance: Our findings could provide valuable insights into the safety of performing SMILE with reduced RST values for high myopia correction and guide SMILE procedures.

Introduction

Currently, laser refractive surgery of the cornea is one of the most important treatments for myopia and myopic astigmatism.¹ Small-incision lenticule extraction (SMILE) is an all-in-one femtosecond laser refractive technique.² SMILE is flapless and minimally invasive, preserving much of the intact and stronger anterior stromal tissue.³ This preservation theoretically results in less damage to the structural stability of the cornea than other refractive surgeries, such as LASIK, FS-LASIK, or PRK.⁴ 

Corneal biomechanical properties are crucial parameters for evaluating the risk of corneal ectasia after refractive surgery.^5–7 The deformation ability of the cornea is influenced by its structure and material properties. Elastic properties can reflect the resistance to deformation, whereas viscoelastic properties, especially creep, can indicate the performance of the material over a longer period. Typically, the material properties of the cornea can be obtained through in vitro tests such as tensile or inflation tests. 

Several studies have evaluated biomechanical properties of the cornea after refractive surgery in vitro.^8–10 However, a few studies have explored how corneal material properties change over time after surgery. Some studies have indicated an increase in corneal elastic modulus after LASIK, FS-LASIK, or PRK when the residual stromal bed thickness (RST) was less than 50% of the central corneal thickness (CCT) within six months.^11–13 He et al.¹⁴ reported no significant difference in elastic modulus between caps of 110 µm and 160 µm when correcting approximately −4D diopter at four months after SMILE (RST is approximately 30% and 45% of the CCT). These studies primarily established animal models for the correction of moderate myopia and had a relatively short postoperative follow-up period. In clinic, Liu et al.¹⁵ recommended maintaining a sufficient RST of 280 µm, not less than 250 µm (approximately 50% of preoperative CCT) for refractive surgery. 

In this study, SMILE rabbit animal models with RST of 30% and 50% of preoperative CCT were characterized biomechanically. Given that SMILE preserves a larger portion of the stronger anterior stroma by creating a corneal cap, the safety of deviating from clinical RST guidelines in the correction of high myopia needs further investigation. Corneal biomechanical properties and morphological parameters were determined at 11 months after SMILE. Moreover, finite element analysis was used to evaluate the displacement and Von Mises stress after SMILE surgery. The purpose of this study was to provide valuable insights into the safety of performing SMILE with reduced RST values for high myopia correction. 

Material and Methods

Experimental Animals

Twenty-four healthy Japanese large-eared rabbits, weighing between 2.5 and 3.0 kg, were enrolled in this study. The animals were sourced from the Animal Experiment Center of Shanxi Medical University. Slit-lamp microscopy examinations were conducted, which revealed that none of the animals had corneal or conjunctival infections. The rabbits were housed in an environment maintained at a constant temperature and humidity. Ethical guidelines in accordance with the Declaration of Helsinki were strictly adhered to, and the study received approval from the Ethics Committee of Taiyuan University of Technology (protocol code TYUT-202103001). 

Study Design

The rabbits were randomly allocated into two groups, with RST values of 50% (RST 50% group, n = 11) and 30% (RST 30% group, n = 13) of the CCT, respectively. The corneal cap thickness was set at one third of the preoperative CCT. The RST 50% group underwent a correction degree of approximately −3D, whereas the RST 30% group had a correction degree of approximately −9D. SMILE surgery was performed on the right eye of each rabbit, and the left eye served as the control without any treatment. Supplementary Information I provides detailed procedure of the SMILE. 

Pentacam Examination

In our experimental design, 24 rabbits were included and underwent SMILE surgery. However, only 12 rabbits survived to 11 months after surgery because of coccidiosis or sudden death of unknown cause. Corneal morphology parameters in vivo were assessed using Pentacam-HR Scheimpflug camera (Oculus, Wetzlar, Germany) before surgery, as well as at one month and 11 months after surgery. The anterior central elevation (ACE), mean keratometry from the anterior corneal surface (Km F), posterior central elevation (PCE), and mean keratometry from the posterior corneal surface (Km B) were used to evaluate postoperative corneal morphological changes. To obtain good results, the light source of the Pentacam-HR Scheimpflug camera was turned down because the iris of white rabbits lacks pigment. During the examination process, the rabbit's body was put in a bag with the head exposed outside. Then an operator held the rabbit to keep it still, and another operator lifted the upper and lower eyelids of the rabbit to expose the cornea completely. The measurement was repeated until test quality (QS) was OK or stable for at least twice. The corneal topography data from rabbit with obvious deviation from the pupil center were excluded in this study. All measurements were carried by same experienced operator. The number of examinations also depended on the rabbit's cooperation. Therefore only five rabbits in the RST 50% group and five rabbits in the RST 30% group (total 10 rabbits) were used for further morphology analysis in vivo. Corneal topography pictures are presented in Supplementary Information II. It should be stated that rabbits not cooperating in the topography examination can still be used for mechanical tests in vitro. Therefore 12 rabbits were included in the mechanical test with six per group. 

Biomechanical Tests

At the 11-month postoperative time point, the rabbits were euthanized under anesthesia by air embolism. The intact cornea was promptly separated. The corneal tissue was cut into a rectangular strip using a 3 mm wide double-edged knife along the superior-inferior directions. Biomechanical tests were conducted immediately. Because of occurrences of coccidiosis or other infectious diseases, only six rabbits in the RST 50% group and six rabbits in the RST 30% group survived to the 11 months after SMILE, and were used for further analysis. 

A uniaxial tensile test was used to determine corneal viscoelastic properties using a testing machine (Instron 5544; Instron, Norwood, MA, USA) equipped with a 5.0 N full-scale load cell (accuracy of 2%). Each specimen was subjected to 10 times of loading and unloading cycles at a rate of 2 mm/min to 0.006 N of uniaxial tension to eliminate pre-stress. The load from 0 to 0.03 N was imposed on the specimens at a rate of 0.5 mm/min to preserve the quasi-static testing conditions. When the load reached 0.03 N, it was held constantly for 60 minutes to perform the creep test. All tests were performed at room temperature using a saline solution bath to keep specimens moist. The creep rate, elongation rate, and equilibrium modulus were calculated using Equations 1, 2, and 3, respectively. Moreover, uniaxial tests were conducted to determine the stress-strain curve and stiffness coefficient curves, stiffness coefficient denoted as “k.”  

\begin{eqnarray}\gamma = {{(\Delta {{L}_2} - \Delta L{}_1)} / {(L + \Delta {{L}_1})}}\end{eqnarray}

(1)

 

\begin{eqnarray}\lambda = {{\Delta {{L}_2}} / L}\end{eqnarray}

(2)

 

\begin{eqnarray}{{E}_\infty } = {{\sigma }_0}/\varepsilon (\infty )\end{eqnarray}

(3)

where γ represents creep rate, L is the original length of the specimen, ΔL₁ denotes the elongation after stretching relative to the original length, ΔL₂ represents the elongation after creep relative to the original length, λ is elongation rate, E_∞ is equilibrium modulus, σ₀ is stabilization stress, and ε is strain. 

Corneal Geometry and Corneal Finite Element Model

Generally, mechanical behaviors of a soft material can be assessed from material properties (e.g., modulus, stiffness, strength), strain, and stress. For example, areas of stress concentration in a material are more prone to failure. Changes and distribution of strain and stress cannot easily be directly obtained experimentally; however, finite element method (FEM) has the capacity to visualize these changes within any structural component under different situations. 

In this study, the material properties of the cornea were determined by mechanical tensile test, and FEM was used to simulate corneal biomechanical behaviors. It can not only serve as a useful tool in guiding the design of surgical parameters, but also predict the therapeutic outcomes of refractive procedures. To obtain a comprehensive understanding of corneal mechanical behaviors after SMILE, a finite element model was established to analyze changes and distribution of displacement and stress across the cornea in combination with tensile test results. 

SOLIDWORKS Corporation 2023 (Dassault Systemes, Vélizy-Villacoublay, France) was used to establish three-dimensional models of intact corneas and SMILE geometric model. The surgery parameters were configured as follows: lenticule diameter (optical zone) of 6.0 mm and cap thickness of 120 µm. Additionally, two different residual stromal bed thicknesses were considered, corresponding to 50% and 30% of preoperative CCT, respectively. An additional 15 µm of the lenticule for easy extraction was also considered for the actual ablation profile in the SMILE model (Fig. 1). 

Figure 1.

 

The model features of SMILE procedure. (1 corneal cap; 2 lenticule; 3 cap opening incision.)

The Ogden model was used to describe the nonlinear hyperelastic behavior of the cornea.¹⁶ The material parameters, including the µ_i and α_i (i = 1 ... N) for the Ogden function, were evaluated by fitting experimental data from the uniaxial tensile test using ABAQUS 6.14 (Dassault Systemes) software. The material parameters obtained from the fitting of the third-order equation were used for conducting finite element simulations. The specific values of the parameters µ_i and α_i obtained through this fitting process are presented in Table 1. 

Table 1.

 

The Constitutive Parameters of the Ogden Model

 

For boundary conditions, the periphery of the cornea was fixed in all degrees of freedom. The friction coefficient between the posterior cap surface and the anterior stromal surface was assigned a value of 0.8 to simulate the clinical scenario.¹⁷ As depicted in Figure 1, the coordinate system was established with the corneal center as its origin. To better understand the postoperative biomechanical behavior of the cornea in the SMILE procedure, four paths were examined, namely, the x- and y-axis paths on the corneal anterior surface and the x- and y-axis paths on the corneal posterior surface. 

Statistical Analysis

SPSS 25.0 software (SPSS Inc, Chicago, IL, USA) was used for data analysis. The experimental data conformed to a normal distribution and were expressed as mean ± standard deviation (SD), independent samples t-test was used for comparison between groups. Experimental data not following a normal distribution were expressed as median (lower quartile, upper quartile), and two independent samples nonparametric tests were used for comparison between groups. One-way analysis of variance was used for more than two groups. The level of significance was considered when P < 0.05. 

Results

In Vivo Test

There were no significant differences among groups at baseline in intraocular pressure (IOP), CCT, corneal cap thickness, flattest radius of corneal curvature (Rf), and steepest radius of corneal curvature (Rs) (P > 0.05). The preoperative ocular parameters and surgical parameters are shown in Table 2. The statistical analysis results are shown in Supplementary Information III. Compared with the CCT at one month after surgery, the CCT of the RST 30% group and the RST 50% group increased by 31.00 µm and 27.60 µm, respectively, at 11 months after surgery (Table 3). 

Table 2.

 

The Preoperative Ocular Parameters and Surgical Parameters

 

Table 3.

 

Comparison of CCT (µm) at One Month and 11 Months After SMILE

 

As indicated in Table 4, at 11 months after surgery, Km F of both operation groups decreased significantly compared to the control (P < 0.05), although there were no significant differences among groups in ACE, PCE, and Km B between operation groups and the control (P > 0.05). In addition, significant differences were also not found in ACE, PCE, Km F, and Km B between the RST 50% and RST 30% (P > 0.05). 

Table 4.

 

Comparison of Corneal Morphological Parameters at 11 Months After SMILE

 

Stress-Strain Experiment

The average stress-strain curves under physiological strain are shown in Figure 2a. The stress-strain curves under physiological strain are shown in the Supplementary Information IV. At the same strain level, the operative groups were able to resist a significantly higher stress than the control group. The elastic modulus increased with strain in all groups. Also, the RST 30% group had a significantly higher elastic modulus than the control group (P < 0.05). The elastic modulus of the RST 50% group had values situated between those of the untreated corneas and the RST 30% group, but the difference did not reach the level of statistical significance in comparison to the RST 30% group or the control. Importantly, no significant difference was found between the RST 50% and RST 30% groups (P > 0.05) (Fig. 2b). For more detailed statistical results, please refer to Supplementary Information V. 

Figure 2.

 

The stress-strain behavior. (a) The average stress-strain curve; (b) The secant modulus in the physiological strain state; (c) The average IOP-strain curve. *P < 0.05 versus control.

The stress-strain curve or the elastic modulus reflects the mechanical properties of the material itself. To assess the ability of the cornea to resist deformation, not only the elastic modulus should be considered, but also the overall anti-deformation ability including corneal thickness (cornea stiffness), No significant difference was found among the control, RST 50%, and RST 30% groups in cornea stiffness (1.37 ± 0.47 N/m, 1.36 ± 0.56 N/m, and 1.67 ± 0.82 N/m, respectively) under 2% strain (P > 0.05). That is, under the same IOP, the greater the strain is, the easier it is to deform. Therefore the IOP-strain curve may be more intuitive. As shown in Figure 2c, under IOP at 15 mm Hg, although the strain in the control group was greater than that in the operation groups, there was no statistical difference between the control and the operation groups in their ability to resist deformation (P > 0.05). 

Viscoelasticity Experiment

The average creep curves are presented in Figure 3a. There were no statistical differences observed among all groups in terms of creep rate, elongation rate, or equilibrium modulus (p > 0.05) (Figs. 3b–d). These results suggest that the corneas exhibited similar viscoelastic behavior. 

Figure 3.

 

Creep test results in the physiological stress state. (a) The average creep curve; (b) the creep rate; (c) the elongation rate; (d) the equilibrium modulus.

Finite Element Analysis

Figure 4 shows the displacement along four paths for the intact cornea and SMILE. The intact cornea tends to exhibit a softer corneal response to the IOP (higher displacement along the same path).As shown in Figure 4, the SMILE produces a shift in the displacement versus path curve. For the displacement, along the x-axes and y-axes, the performance of the RST 30% and RST 50% groups was similar under IOP. 

Figure 4.

 

Comparison of displacement of the cornea along the different paths. The IOP was constant at 10 mm Hg. (a) Displacement on the anterior corneal surface along the x-axis. (b) Displacement on the anterior corneal surface along the y-axis. (c) Displacement on the posterior corneal surface along the x-axis. (d) Displacement on the posterior corneal surface along the y-axis.

Displacement distributions through a front view and cross-sectional view of the cornea were presented in Figure 5. In the SMILE model, the displacement distribution was asymmetric, with the maximum displacement shifting from the center of the cornea towards the incision site. 

Figure 5.

 

Displacement distributions after applying IOP of 10 mm Hg. (a) Control of the anterior corneal surface. (b) RST/ Pre CCT = 50% group of the anterior corneal surface. (c) RST/Pre CCT = 30% group of the anterior corneal surface. (d) Control on the posterior corneal surface. (e) RST/Pre CCT = 50% group of the posterior corneal surface. (f) RST/Pre CCT = 30% group of the posterior corneal surface. (g) Control of the cross-sectional view. (h) RST/Pre CCT = 50% group of the cross-sectional view. (i) RST/Pre CCT = 30% group of the cross-sectional view.

The Von Mises stress of SMILE along the x- and y-axes exhibits notable differences on the anterior corneal surface (Figs. 6a, 6b). While the stress along the x-axes tended to be bilaterally symmetric, the stress along the y-axis was not bilaterally symmetric, potentially due to the position of the corneal cap incision. The maximum Von Mises stress in SMILE was higher than in the intact cornea along the y-axis. Additionally, the Von Mises stress on the posterior corneal surface in the SMILE corneal model was evidently higher than in the intact model (Figs. 6c, 6d). The Von Mises stress increased with the refractive correction. 

Figure 6.

 

Comparison of Von Mises stress on the cornea along the different paths. The IOP was constant at 10 mm Hg. (a) Von Mises stress on the anterior corneal surface along the x-axis. (b) Von Mises stress on the anterior corneal surface along the y-axis. (c) Von Mises stress on the posterior corneal surface along the x-axis. (d) Von Mises stress on the posterior corneal surface along the y-axis.

Discussion

It is unclear whether extracting the lenticule at a deeper stroma with a less residual stromal bed is feasible for SMILE surgery without the risk of iatrogenic corneal ectasia. Our study showed that the cornea became stiffer after SMILE with a cap of one-third of preoperative CCT, and achieved similar mechanical performance to the normal cornea in the ability to resist deformation. A new mechanical equilibrium was restored and sufficient to combat increased stress across the cornea. No iatrogenic corneal ectasia occurred within 11 months follow-up after SMILE. 

Studies have shown that rabbit corneal thickness increased with age; an increase of 10 µm/month in CCT was observed from age one month to seven months, but the gain reduced to less than 1 µm/month from ages eight months to 12 months.¹⁸ Havens et al.¹⁹ reported that CCT increased from 316 µm at 10 weeks to 337 µm at 30 weeks. Moreover, in clinical studies, corneal stromal thickening persisted up to one year after refractive surgery with an increase of 18.7 µm, and a greater degree of correction resulted in greater corneal thickening,^20,21 which may lead to refractive regression. Both corneal epithelium and stroma remodeling may contribute to refractive regression after SMILE surgery, especially in the case of high refractive corrections.^22,23 In this study, CCT increased by approximatively 8 µm in the control group with natural growth. The RST 50% and 30% groups increased by approximately 27 µm and 31 µm, respectively. Our results together with previous findings suggest that the corneal thickening after SMILE may be attributed to both natural growth and the wound healing response to tissue removal. 

Changes in corneal surface curvature and height are usually used to evaluate the outcomes of corneal refractive surgery, and especially for iatrogenic corneal ectasia after surgery. The anterior surface of the cornea will be flattened to reduce the refractive power of the cornea after SMILE. In this study, at 11 months postoperatively, the ACE of RST 30% and RST 50% groups were smaller than the control, it was due to corneal stroma removed after the operation. Jia et al.²⁴ reported a similar result with this study, 1 month after SMILE, the corneal flattening in RST 30% group was greater. Our results showed that there was no significant forward shift of PCE or increase in Km B in the SMILE groups at 11 months. PCE and Km B are highly specific and sensitive indicators of early corneal ectasia.²⁵ It indicate that breaking the clinical limit of RST does not seem to increase the risk of corneal ectasia within one year. 

Zhang et al.¹³ reported that after FS-LASIK with corneal ablation exceeding 50% of preoperative CCT, corneal viscoelastic properties were closer to the untreated control at one month. In this study, corneal viscoelastic properties were assessed using creep tests, which reflect the material's ability to undergo long-term deformation under a constant stress level. Our results showed that retaining RST less than 50% of preoperative CCT had no discernible influence on corneal viscoelastic parameters (including the equilibrium modulus, creep rate, and elongation rate) at 11 months after surgery. It indicated that the cornea maintained its long-term resistance to deformation at physiological IOP when high myopia was corrected, even with an RST lower than 30% of preoperative CCT if the corneal cap thick was enough. 

SMILE has attracted attention because of the advantages of small incisions, the ability to keep the cornea intact, with more compacted anterior stroma preserved which is believed stronger than the posterior. In this study, the RST 30% group demonstrated a higher elastic modulus than the control. This result is consistent with the concept described by Reinstein et al.²⁶ that the thicker the cap, the greater the corneal biomechanics. A corneal cap thickness is commonly set to 110 to 120 µm (about one fifth of preoperative CCT) for myopic correction in the clinic SMILE surgery,¹⁵ and the stromal lenticule is usually removed from part of anterior stroma. In this study, we adopted a corneal cap thickness of one third of preoperative CCT, and the lenticule was extracted from the mid-posterior part of the corneal stroma, preserving more anterior stroma. Studies have reported that the elastic modulus of the rabbit cornea is three times greater in the anterior stroma compared to the posterior stroma.²⁷ Removal of the lenticule in deeper layers (thicker cap) may better preserve the stronger part of the cornea than in superficial layer. This is the main reason why the cornea become stiffer than the control. In a recent randomized contralateral eye study, they reported that a thicker SMILE corneal cap thickness (145 µm vs. 110 µm) resulted in better corneal biomechanical properties with no significant effects on visual acuity.²⁸ Our experiment results also confirm the theoretical model proposed by Reinstein et al.²⁶ that SMILE surgery may not follow the same criteria as LASIK, which residual stromal bed thickness must be retained more than 250 µm (about 50% CCT). 

Another reason for stiffer cornea in 30% RST group may be related to local stress increase across the cornea (as shown in Fig. 6). It's known that Young's modulus of the cornea will increase with stress.²⁹ In a recent animal study, effects of different stromal ablation amount on the corneal biomechanical properties in vivo were investigated using optical coherence elastography at one month after SMILE surgery. They found that the Young's modulus of rabbit cornea increased after SMILE with -6D group higher than −3D group.³⁰ This reported results were consistent with ours, that corneal elastic modulus increased in RST 30% group with more tissue removal after SMILE. In addition, under the higher stress, the first response of the cornea may become more compact and “strong” to combat structure failure. Our previous study showed that corneal elastic modulus increased significantly when the residual stromal bed thickness was 30% of the total corneal thickness one month after LASIK surgery.¹¹ Zhang et al.¹³ established animal model of FS-LASIK surgery, they also observed that the surgery group with more corneal tissue removed had a higher elastic modulus than the control. All of the above findings may be explained by the adaptive stress-growth effect of Davis's law, which describes the relationship between stress and growth in soft tissues. However, if a considerable amount of tissue is removed and the stress across the cornea exceeds its tolerance limit, corneal ectasia may occur. 

In the present clinic, the corneal biomechanics is usually assessed by Corvis ST. The cornea always showed a decrease in corneal biomechanical properties after refractive surgery.^31,32 A tensile test is generally finished within a few minutes (quasi-static mechanical behavior) by stretching a tissue strip to determine the elastic modulus, which reflects the mechanical properties of the material itself. Although biomechanical parameters from Corvis ST are obtained under fast air puff within ∼30 msec (dynamic mechanical behavior), and these parameters are comprehensive. Therefore the results from the Corvis ST and tensile tests can't be compared directly. In fact, to assess the ability of the cornea to resist deformation in vivo, besides corneal material mechanical properties (e.g., elastic modulus), the overall anti-deformation ability including corneal thickness (that is, cornea stiffness) should be addressed. In this study, although the elastic modulus in 30% RST group became greater than the control, there was no significant difference among three groups in cornea stiffness under 2% strain. This means that the ability of the cornea to resist deformation is similar between the control and the operation groups under physiological IOP. 

This study has several limitations, and the interpretation of the results needs to be cautious. First, the sample size of this experiment was small. Second, corneal thickness and curvature change with age, which may have an impact on the experimental results. Third, making thicker corneal caps in high myopia correction means separating the lenticule in a deeper stroma, which increases surgery difficulty, and whether it influences visual quality needs to be proven. Finally, there is a difference in corneal tissue repair ability between rabbits and humans after refractive surgery. It is worth mentioning that this experiment was carried out on healthy rabbits, not myopic ones. However, our recent work has showed that the biomechanical properties of the cornea were correlated with myopia degree.³³ The mechanical properties of the cornea are lower in highly myopic patients than in healthy individuals, and it remains unknown whether they are at risk for iatrogenic keratectasia in SMILE surgery with RST lower than 30% CCT. Caution should be taken if the proposed treatment approach is translated to the treatment of human high myopia. 

Conclusions

No iatrogenic keratectasia occurred in a rabbit model of high myopic SMILE surgery with RST of less than 50% of preoperative CCT at 11-month follow-up. 

Acknowledgments

Supported by the National Natural Science Foundation of China (Grant Nos. 12072218), Natural Science Foundation of Shanxi Province, grant number 20210302123150; and Four “Batches” Innovation Project of Invigorating Medical through Science and Technology of Shanxi Province, 2023XM017. 

Disclosure: H. Qin, None; X. Yang, None; R. He, None; Y. Song, None; J. Wei, None; X. Liu, None; C. Wang, None; C. Wu, None; J. Hou, None; Z. Gao, None; L. Chen, None; X. Li, None; W. Chen, None