Evaluation of stability and visual quality in Chinese myopic eyes 2 years following FS-LASIK combined with prophylactic corneal crosslinking

Abstract

Background

To assess corneal stability of femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking (FS-LASIK Xtra) for myopic correction and compare the visual quality to FS-LASIK at 2 years postoperative.

Methods

Data of 40 patients (80 eyes) receiving FS-LASIK Xtra were collected in the longitudinal study. They were identified as “borderline” with the risk factors including younger in age, thinner corneal thickness, irregular cornea, or weak corneal biomechanics before surgery. 20 patients (40 eyes) of them were recalled for a comprehensive visit at 2 years postoperative, along with a matched group of 27 patients (54 eyes) without risk factors and receiving FS-LASIK alone during the same period. Confounders including age, gender, and baseline refractive parameters were controlled. Refractive status, corneal morphology and biomechanics, epithelial thickness (ET), subjective and objective visual quality, and tear film stability were examined. Generalized linear mixed models were constructed for comparative and correlative analyses.

Results

Spherical equivalent (SE), thinnest corneal thickness, integrated radius, central ET changed significantly from 6 months to 2 years after FS-LASIK Xtra (all P < 0.050). There was no statistical difference in SE, biomechanics, higher-order aberrations (HOAs), contrast sensitivity, tear break-up time, and total score of Quality of Vision questionnaire between two surgical groups at 2 years postoperative (all P > 0.050). FS-LASIK Xtra group had greater postoperative objective scattering index (OSI) and corneal densitometry than FS-LASIK group (all P < 0.050). Epithelial remodeling was lower after FS-LASIK Xtra though significance was only observed in pericentral zone (P < 0.050), and its correlation with HOAs was evident in coma (P < 0.050).

Conclusions

FS-LASIK Xtra demonstrated its potential in borderline cases for effective myopic correction, along with improvements in corneal morphology and biomechanics during this 2-year visit. However, the additional crosslinking seemed to increase light scatter indicated by higher corneal densitometry and OSI, though patients reported comparable subjective visual outcomes to those of FS-LASIK alone. And postoperative epithelial remodeling still requires long-term monitoring.

Background

The efficacy and safety of femtosecond laser-assisted in situ keratomileusis (FS-LASIK) have been confirmed through years of clinical practice [1]. But the creation of corneal flap and stromal ablation disrupts the integrity of cornea, potentially predisposing patients to postoperative corneal ectasia and impacting final refractive result permanently [2]. This serious complication is particularly associated with young age, high myopia, thin predicted residual stromal thickness, abnormal corneal topography, and weakened corneal biomechanical properties [3].

Nowadays, there is a growing demand for myopic correction among young population in China due to occupational requirements or military service regulations. An accelerated corneal crosslinking (CXL) protocol has been introduced into the conventional FS-LASIK procedure to prevent corneal ectasia and known as prophylactic CXL or FS-LASIK Xtra [4]. The combined action of riboflavin and ultraviolet-A in CXL causes the photopolymerization among collagen fibers, strengthening the corneal biomechanics [5]. Favourable refractive outcomes after FS-LASIK Xtra have been shown in previous studies, which seems to expand the indication of myopic correction [6, 7]. However, there is a paucity of long-term comparisons of visual quality, corneal morphology, and biomechanics between FS-LASIK with and without prophylactic CXL. Besides, corneal epithelial remodeling is generally considered to be related to myopic regression after FS-LASIK [8], but its influence on visual quality after LASIK Xtra is not fully elucidated. This present study was therefore conducted to explore the above questions.

Methods

Preoperative and early postoperative (1, 3, and 6 months) data of patients undergoing FS-LASIK Xtra for myopia and astigmatism in Beijing Tongren Hospital from January 2020 to November 2023 were retrospectively collected in this study. The surgical inclusion criteria were: over 18 years of age at surgery, stable refraction (alteration of 1.00D or less) for at least 2 years, preoperative corrected distance visual acuity (CDVA) not less than 20/25, stopped wearing soft corneal contact lens for 2 weeks, rigid contact lenses for 1 month and orthokeratology lens for more than 3 months. The exclusion criteria were: presence or history of ocular diseases including keratoconus or suspect keratoconus, systemic diseases, history of ocular surgery, pregnancy, and lactation. Besides, these patients were identified as “borderline” before surgery considering the following risk factors comprehensively: corneal thickness less than 510 μm, residual bed thickness less than 300 μm, Pentacam Scheimpflug tomography (Oculus GmbH, Wetzlar, Germany) displayed mild inferior-superior asymmetry, yellow or red in Belin/Ambrósio Enhanced Ectasia Display, Corvis biomechanical index over 0.5 and tomographic biomechanical index over 0.29 measured by Corvis ST (Oculus GmbH, Wetzlar, Germany), younger than 30 years old, and history of eye rubbing. All the patients were carefully reassessed for refractive stability, corneal morphology, and biomechanical properties prior to surgery, and were thoroughly informed about the potential risks, including the possibility of ectasia even after prophylactic CXL. After weighing the high demand for myopic correction against the surgical risks, these patients decided to receive FS-LASIK Xtra. At the 2-year time point after FS-LASIK Xtra we contacted them by phone to complete comprehensive visual quality examinations in addition to routine refractive and corneal examinations. Meanwhile, a matched group without the risk factors and receiving FS-LASIK alone from January 2020 to November 2022 was set to compare. Potential confounders including age, gender, and baseline refractive parameters were controlled.

This study complied with the Declaration of Helsinki and was approved by the institutional review board of Beijing Tongren Hospital, Capital Medical University (TRECKY2021-216). Written informed consent was obtained from all patients.

Surgical technique

Surgeries were performed by an experienced surgeon (FZ). In FS-LASIK, femtosecond laser system (FS200, Alcon Laboratories, Inc., USA) was used to create the corneal flap. The pulse energy was 130 nJ. The intended flap thickness was 100–110 μm. The position of hinge was set at 12 o’clock. Lifting the flap, excimer laser (WaveLight EX500, Alcon Laboratories, Inc., USA) was used to perform stromal ablation, and the diameter of optical zone was 6.0–6.5 mm. Then the flap was carefully repositioned.

In FS-LASIK Xtra, after the excimer laser ablation, Vibex Xtra (Avedro Inc., Waltham, MA, USA), containing 0.22% riboflavin was placed on the exposed stromal bed for 90 s. Following the soaking, the stroma was washed with BSS and the flap was replaced. Then cornea was irradiated with ultraviolet-A light (KXL System, Avedro Inc. USA) with a wavelength of 365 nm and an irradiance of 30 mW/cm² for 90 s (total energy 2.7 J/cm²).

Measurements

Routine ophthalmologic examinations including uncorrected distance visual acuity (UDVA), CDVA, manifest refraction, noncontact tonometer, slit lamp and fundus examination were conducted. Pentacam was used to examine corneal topography and densitometry (CD) at 3 depths, consisting of anterior (front 120 μm), central, and posterior layers (back 60 μm), and 3 concentric annuluses (0–2 mm, 2–6 mm, and 6–10 mm). Biomechanics were measured by Corvis ST. Corneal epithelial thicknesses (ET) in 17 sections and 3 annuluses including central (0–2 mm), paracentral (2–5 mm), and pericentral zones (5–6 mm) were measured by RTVue-100 (Optovue, Inc., Fremont, USA).Wavefront aberrations were obtained by iDesign (Abbott Medical Optics, Inc., Santa Ana, USA) under standard scotopic conditions at the pupil size of 6 mm. Contrast sensitivity was measured by Optec 6500 (Stereo Optical Co., Chicago, USA) with mesopic lighting of 3 cd/m², at spatial frequencies of 1.5, 3, 6, 12, and 18 cycles per degree. Objective scatter index (OSI) and modulation transfer function cut off (MTF cut off) value were measured using Optical Quality Analysis System (OQAS II, Visiometrics, Spain). Results of tear break-up time and corneal fluorescein staining were recorded [9]. Patients were also asked to complete the Quality of Vision (QoV) questionnaire and the ocular surface disease index (OSDI) questionnaire to assess visual and dry eye symptoms [10, 11].

Statistical analysis

SPSS (version 29; SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Both eyes of patients were included in the statistical analysis. Descriptive statistics were presented as mean±standard deviations. The normality of quantitative variables was checked by Kolmogorov-Smirnov test. When comparing two surgical groups, independent-samples t test was used for age at surgery, Yates’ continuity correction for gender, and Mann-Whitney U test for questionnaires. Generalized linear mixed models were constructed for the other baseline variables and postoperative results. For those receiving FS-LASIK Xtra, generalized linear mixed models were also used to analyse the change of postoperative refractive outcomes, corneal morphological and biomechanical parameters, and ET with time, and to explore the correlations between epithelial thickening and visual quality. Bonferroni correction was used for multiple tests. A P value of less than 0.05 was considered statistically significant, and contrast estimate along with 95% confidence interval (CI) was also reported.

Results

Data of 40 patients (80 eyes) undergoing FS-LASIK Xtra was collected for longitudinal study. There were 27 females and 13 males, and the mean age was 24.95 ± 4.72 years (range from 18 to 33 years). The mean preoperative UDVA and CDVA were 1.24 ± 0.41 (range from 0.22 to 2.00) logMAR and − 0.04 ± 0.04 (range from − 0.08 to 0.50) logMAR. The baseline and change of their refractive outcomes over time were showed in Table 1. A myopic shift of spherical equivalent (SE) appeared at 2 years compared to 6 months postoperative (contrast estimate: -0.52; 95% CI: -0.84 to -0.20; P < 0.001).

Table 1.

Refraction, corneal morphological and biomechanical parameters, and epithelial thickness over 2 years after FS-LASIK Xtra^a

	Preoperative	Postoperative
	(80 eyes)	1 month (46 eyes)	3 months (40 eyes)	6 months (26 eyes)	2 years (40 eyes)
Refractive Outcomes
Spherical equivalent (D)	-5.782 ± 1.962	0.340 ± 0.641	0.082 ± 0.520	0.019 ± 0.462	-0.493 ± 0.542*†
Cylinder (D)	-1.16 ± 0.94	-0.06 ± 0.52	-0.01 ± 0.35	-0.04 ± 0.47	-0.26 ± 0.35
Corneal Morphological and Biomechanical Parameters
K1 (D)	43.1 ± 1.3	37.9 ± 1.6	38.9 ± 1.7	38.7 ± 2.0	38.8 ± 1.9
K2 (D)	44.9 ± 1.4	38.5 ± 1.5	39.6 ± 1.7	39.4 ± 2.2	39.9 ± 3.0
Kmax (D)	45.4 ± 3.9	43.7 ± 1.8	44.5 ± 1.8	44.7 ± 2.1	43.6 ± 4.5
TCT (µm)	501 ± 55	397 ± 24	407 ± 21	400 ± 12	417 ± 19*†
ISV	23 ± 9	46 ± 20	44 ± 17	43 ± 6	39 ± 19
IVA	0.16 ± 0.08	0.36 ± 0.18	0.38 ± 0.19	0.46 ± 0.11	0.33 ± 0.21
KI	1.05 ± 0.04	0.98 ± 0.04	0.99 ± 0.05	1.05 ± 0.05	1.00 ± 0.04
CKI	1.01 ± 0.01	1.01 ± 0.03	1.01 ± 0.03	1.00 ± 0.02	1.00 ± 0.01
IHA	8.1 ± 7.1	7.1 ± 6.4	13.5 ± 9.3	10.3 ± 5.8	11.4 ± 10.7
IHD	0.013 ± 0.008	0.030 ± 0.019	0.032 ± 0.020	0.030 ± 0.006	0.029 ± 0.020
Rmin	7.43 ± 0.26	7.74 ± 0.31	7.60 ± 0.30	7.92 ± 0.06	7.61 ± 0.32
Df	1.06 ± 1.16	5.97 ± 5.25	4.94 ± 5.35	4.46 ± 4.38	5.28 ± 5.21
Db	0.19 ± 0.93	-0.09 ± 1.02	-0.01 ± 0.85	-0.22 ± 0.99	-0.16 ± 0.81
SP-A1	94.2 ± 12.6	56.2 ± 10.1	59.6 ± 10.1	61.6 ± 9.2	70.9 ± 11.6†
IR	9.6 ± 0.9	12.1 ± 0.4	12.0 ± 0.7	12.1 ± 0.9	11.3 ± 0.7*†
ARTh	409.5 ± 85.0	97.9 ± 35.0	115.0 ± 33.9	117.4 ± 31.7	131.7 ± 28.2†
DAR	5.1 ± 4.7	5.8 ± 0.4	5.8 ± 0.3	5.6 ± 0.4	5.6 ± 0.5
Epithelial Thickness (µm)
Central zone	53.72 ± 2.77	55.68 ± 6.77	57.11 ± 4.40	56.13 ± 3.27	60.43 ± 4.42*†
Paracentral zone	53.63 ± 2.35	57.69 ± 5.41	58.88 ± 3.80	58.44 ± 2.85	60.79 ± 3.55†
Pericentral zone	53.38 ± 2.31	55.08 ± 3.86	53.96 ± 3.17	55.20 ± 2.69	54.93 ± 2.81
Standard deviation	1.48 ± 0.64	3.51 ± 1.39	3.65 ± 0.96	3.27 ± 1.00	3.30 ± 1.10
Min-Max	-6.53 ± 2.44	-16.14 ± 6.33	-15.74 ± 4.40	-13.66 ± 3.90	-14.28 ± 4.88

FS-LASIK Xtra=femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking; D=diopter; K1 = flattest keratometry; K2 = steepest keratometry; Kmax=maximum keratometry; TCT=thinnest corneal thickness; ISV=index of surface variance; IVA=index of vertical asymmetry; KI=keratoconus index; CKI=central keratoconus index; IHA=index of height asymmetry; IHD=index of height decentration; Rmin=minimun curvature radius; Df=front elevation deviation; Db=back elevation deviation; SP-A1 = stiffness parameter at the first applanation; IR=integrated radius; ARTh =Ambrósio relational thickness; DAR=deformation amplitude ratio at 2 mm; ^ageneralized linear mixed model; ^*P < 0.050 versus 6 months postoperative; ^†P < 0.050 versus 3 months postoperative

The comparative study at 2 years postoperative included 20 patients (40 eyes) receiving FS-LASIK Xtra and 27 patients (54 eyes) receiving FS-LASIK. The mean values of baseline demographic and ocular characteristics were comparable between groups (all P > 0.050, Table 2 and Fig. 1). Mean postoperative UDVA was better in FS-LASIK group (contrast estimate: 0.04; 95% CI: 0.01 to 0.08; P = 0.013). The efficacy index (postoperative UDVA/preoperative CDVA) was 0.85 ± 0.15 in FS-LASIK Xtra group and 0.97 ± 0.19 in FS-LASIK group.

Table 2.

Demographic and Ocular Characteristics of FS-LASIK Xtra Group and FS-LASIK Group^a

	FS-LASIK Xtra (40 eyes)	FS-LASIK (54 eyes)	P	Contrast estimate (95% CI)
Demographic Characteristics
Age at surgery (years)	26.37 ± 4.55 (18 to 33)	27.41 ± 4.04 (18 to 36)	0.120	-1.41 (-3.18 to 0.37)
Female/Male	16/4	23/4	0.940
Ocular Characteristics
Preoperative
UDVA (logMAR)	1.26 ± 0.44 (0.40 to 2.00)	1.40 ± 0.25 (0.70 to 2.00)	0.128	-0.15 (-0.35 to -0.04)
CDVA (logMAR)	-0.05 ± 0.04 (-0.08 to 0.05)	-0.04 ± 0.04 (-0.80 to 0.00)	0.406	-0.01 (-0.04 to 0.01)
SE (D)	-6.14 ± 1.92 (-10.13 to -2.38)	-6.78 ± 1.48 (-9.25 to -1.40)	0.182	0.64 (-0.30 to 1.58)
Cylinder (D)	-0.97 ± 0.97 (-4.00 to 0.00)	-0.92 ± 0.63 (-2.50 to 0.00)	0.854	-0.04 (-0.48 to 0.40)
2 Years Postoperative
UDVA (logMAR)	0.03 ± 0.07 (-0.08 to 0.22)	-0.02 ± 0.06 (-0.18 to 0.10)	0.013*	0.04 (0.01 to 0.08)
CDVA (logMAR)	-0.05 ± 0.04 (-0.08 to 0.00)	-0.05 ± 0.04 (-0.08 to 0.00)	0.513	-0.01 (-0.03 to 0.01)
SE (D)	-0.44 ± 0.54 (-1.50 to 1.20)	-0.32 ± 0.42 (-1.04 to 0.96)	0.319	-0.13 (-0.38 to 0.13)
Cylinder (D)	-0.28 ± 0.37 (-1.00 to 0.00)	-0.30 ± 0.37 (-1.50 to 0.00)	0.765	0.03 (-0.15 to 0.20)

FS-LASIK Xtra=femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking; UDVA=uncorrected distance visual acuity; CDVA=corrected distance visual acuity; SE=spherical equivalent; D=diopter; ^aContinuous variables were analyzed by independent-samples t test for subject-level variables, generalized linear mixed model for eye-level variables, gender was analyzed by Yates’ continuity correction; ^*P < 0.050

Fig. 1.

Refractive outcomes after FS-LASIK Xtra and FS-LASIK

Corneal morphological and biomechanical parameters

After FS-LASIK Xtra, the minimum value of the thinnest corneal thickness occurred at 1 month. Then, it gradually thickened over time, and significant differences were observed between the 2-year and 3-month as well as 6-month postoperative visits (contrast estimate: 12.19 and 13.58; 95% CI: 4.01 to 20.26, and 1.65 to 25.51; P = 0.001 and 0.019). Topometric indices remained stable (P > 0.050). Stiffness parameter at the first applanation (SP-A1) and Ambrósio relational thickness (ARTh) increased, integrated radius (IR) decreased obviously from 3 months to 2 years postoperative (contrast estimate: 9.01, 22.00 and − 0.63; 95% CI: 0.80 to 17.22, 3.61 to 40.37, and − 1.12 to -0.14; P = 0.024, 0.011 and 0.005).

Comparison of preoperative and 2 years postoperative corneal morphological and biomechanical parameters was shown in Table 3. Significant differences existed in steepest and flattest keratometry after surgery (contrast estimate: 1.86 and 2.41; 95% CI: 0.86 to 2.86, and 1.16 to 3.66; both P < 0.001).

Table 3.

Comparison of Corneal Morphological and Biomechanical Parameters between FS-LASIK Xtra Group and FS-LASIK Group^a

	Preoperative				2 Years Postoperative
	FS-LASIK Xtra (40 eyes)	FS-LASIK (54 eyes)	P	Contrast estimate (95% CI)	FS-LASIK Xtra (40 eyes)	FS-LASIK (54 eyes)	P	Contrast estimate (95% CI)
K1 (D)	43.5 ± 1.8	42.1 ± 1.2	0.005*	1.38 (0.43 to 2.34)	38.6 ± 1.9	36.7 ± 1.5	< 0.001*	1.86 (0.86 to 2.86)
K2 (D)	44.9 ± 1.5	43.6 ± 1.5	0.030*	1.29 (0.13 to 2.45)	39.7 ± 3.0	37.3 ± 1.6	< 0.001*	2.41 (1.16 to 3.66)
Kmax (D)	45.5 ± 1.6	44.1 ± 1.6	0.028*	1.38 (0.15 to 2.60)	43.4 ± 4.4	44.4 ± 2.0	0.162	-0.97 (-2.32 to 0.40)
TCT (µm)	510 ± 19	544 ± 31	< 0.001*	-34.13 (-53.62 to -14.63)	426 ± 55	426 ± 35	0.966	-0.49 (-23.29 to 22.30)
ISV	23 ± 9	19 ± 4	0.165	3.64 (-1.55 to 8.84)	40 ± 18	45 ± 10	0.146	-5.87 (-13.82 to 2.08)
IVA	0.16 ± 0.08	0.13 ± 0.04	0.216	0.03 (-0.12 to 0.08)	0.34 ± 0.21	0.37 ± 0.14	0.551	-0.03 (-0.12 to 0.07)
KI	1.05 ± 0.04	1.04 ± 0.02	0.182	0.02 (-0.01 to 0.04)	0.97 ± 0.15	0.96 ± 0.05	0.592	0.012 (-0.03 to 0.06)
CKI	1.01 ± 0.01	1.01 ± 0.01	0.836	0.00 (0.00 to 0.01)	1.00 ± 0.01	1.00 ± 0.01	0.395	0.00 (-0.01 to 0.00)
IHA	8.1 ± 7.1	7.7 ± 5.2	0.865	0.35 (-3.81 to 4.52)	11.3 ± 10.2	13.6 ± 10.5	0.388	-2.28 (7.49 to 2.94)
IHD	0.013 ± 0.008	0.012 ± 0.005	0.751	0.00 (0.00 to 0.01)	0.029 ± 0.019	0.027 ± 0.011	0.614	0.002 (-0.01 to 0.01)
Rmin	7.43 ± 0.26	7.66 ± 0.27	0.029*	-0.23 (-0.44 to -0.03)	7.64 ± 0.32	7.62 ± 0.34	0.869	0.02 (-0.18 to 0.22)
Df	1.42 ± 1.34	0.90 ± 0.92	0.229	0.52 (-0.34 to 1.39)	4.92 ± 5.08	6.02 ± 2.61	0.334	-1.09 (-3.33 to 1.14)
Db	0.54 ± 0.95	-0.12 ± 0.43	0.026*	0.66 (0.08 to 1.23)	0.02 ± 0.92	-0.33 ± 0.99	0.199	0.35 (0.19 to 0.88)
SP-A1	93.6 ± 16.2	107.6 ± 14.7	0.026*	-14.80 (-27.75 to -1.85)	64.6 ± 14.3	67.5 ± 17.2	0.532	2.84 (-11.82 to 6.15)
IR	9.3 ± 1.0	8.8 ± 0.7	0.187	0.49 (-0.25 to 1.22)	11.3 ± 0.8	11.2 ± 1.6	0.770	0.10 (-0.56 to 0.76)
ARTh	422.3 ± 62.3	491.7 ± 77.4	0.007*	-69.34 (-118.50 to -20.27)	136.8 ± 34.2	132.1 ± 41.1	0.619	4.49 (-17.87 to 26.85)
DAR	4.5 ± 0.5	4.2 ± 0.3	0.071	0.329 (-0.03 to 0.69)	5.5 ± 0.4	5.3 ± 0.5	0.202	0.19 (-0.10 to 0.48)

FS-LASIK Xtra=femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking; K1 = flattest keratometry; K2 = steepest keratometry; Kmax=maximum keratometry; TCT=thinnest corneal thickness; ISV=index of surface variance; IVA=index of vertical asymmetry; KI=keratoconus index; CKI=central keratoconus index; IHA=index of height asymmetry; IHD=index of height decentration; Rmin=minimun curvature radius; Df=front elevation deviation; Db=back elevation deviation; SP-A1 = stiffness parameter at the first applanation; IR=integrated radius; ARTh=Ambrósio relational thickness; DAR=deformation amplitude ratio at 2 mm; ^ageneralized linear mixed model; ^*P < 0.050

Corneal epithelial thickness profiles

ET in central zone at 2 years after FS-LASIK Xtra was significantly thicker than 3 and 6 months (contrast estimate: 3.95 and 2.77; 95% CI: 2.26 to 5.65, and 0.56 to 4.98; P < 0.001 and = 0.008). And ET in paracentral zone increased remarkably from 3 months to 2 years (contrast estimate: 2.42; 95% CI: 0.79 to 4.04; P = 0.001; Fig. 2).

Fig. 2.

Corneal epithelial thickness during 2 years after FS-LASIK Xtra. S=superior; I=inferior; T=temporal; N=nasal; ^*P < 0.050 versus 6 months postoperative; ^†P < 0.050 versus 3 months postoperative

No statistical difference of mean ET in central or paracentral zones was found between surgical groups (both P > 0.050). Mean ET in pericentral zone was thinner FS-LASIK Xtra group (contrast estimate: -2.51; 95% CI: -4.72 to -0.30; P = 0.027), and the change in this annulus was also less than FS-LASIK group (contrast estimate: -3.33; 95% CI: -5.13 to -1.53; P < 0.001).

Comparison of visual quality

There was no statistical difference in any higher-order aberrations (HOAs) between two surgical groups (all P > 0.050), neither of contrast sensitivity at any spatial frequency (all P > 0.050). OSI was greater and MTF cut off value was smaller in FS-LASIK Xtra group, and the former showed statistical significance (contrast estimate: 0.431; 95% CI: 0.05 to 0.82; P = 0.028). CD of 2–6 mm, central layer, and total value were greater (contrast estimate: 0.37, 0.84, and 0.84; 95% CI: 0.04 to 0.70, 0.30 to 1.31, and 0.18 to 1.50; P = 0.028, 0.003, and 0.014; Table 4). Frequency, severity, bothersome and total scores of QoV questionnaire did not differ between surgical groups (all P > 0.050), but the score of fluctuation in vision was higher in FS-LASIK Xtra group (P = 0.030, 95% CI: 0.02 to 0.03). The highest scoring visual symptoms were glare, starbursts, and fluctuation in vision after FS-LASIK Xtra, and were starbursts, glare, and halos after FS-LASIK (Fig. 3).

Table 4.

Comparison of OQAS Measurements and Corneal Densitometry between FS-LASIK Xtra Group and FS-LASIK Group^a

	FS-LASIK Xtra (40 eyes)	FS-LASIK (54 eyes)	P	Contrast estimate (95% CI)
OQAS Measurements
OSI	1.419 ± 1.191	0.984 ± 0.437	0.028*	0.431 (0.05 to 0.82)
MTF cut off	29.292 ± 9.471	33.241 ± 9.016	0.064	-4.18 (-8.60 to 0.24)
Corneal Densitometry
Diameter
0–2 mm	14.5 ± 0.7	14.2 ± 0.7	0.143	0.30 (-0.10 to 0.70)
2–6 mm	13.6 ± 0.5	13.2 ± 0.6	0.028*	0.37 (0.04 to 0.70)
6–10 mm	14.3 ± 2.4	13.4 ± 1.5	0.124	0.89 (-0.25 to 2.03)
Layer
Anterior	20.8 ± 2.4	19.7 ± 1.8	0.077	1.08 (-1.12 to 2.28)
Central	13.6 ± 1.0	12.7 ± 0.8	0.003*	0.84 (0.30 to 1.31)
Posterior	11.9 ± 0.8	11.3 ± 0.8	0.012*	0.58 (0.13 to 1.04)
Total	15.4 ± 1.3	14.6 ± 1.0	0.014*	0.84 (0.18 to 1.50)

OQAS=Optical Quality Analysis System; FS-LASIK Xtra=femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking; OSI=objective scatter index; MTF cut off=modulation transfer function cut off; ^ageneralized linear mixed model; ^*P < 0.050

Fig. 3.

Stacked histogram analysis of visual symptoms at 2 years after FS-LASIK Xtra and FS-LASIK

No discrepancy was observed in tear break-up time, corneal fluorescein staining and OSDI scores (all P > 0.050).

Univariate analysis of epithelial remodeling

The change of ET in 3 annuluses from baseline to 2 years postoperative was not correlated with postoperative SE (all P > 0.050). Correlations between central and paracentral epithelial remodeling and coma were observed but with small coefficients (β = 0.02 and 0.02; 95% CI: 0.00 to 0.03, and 0.00 to 0.04; P = 0.017 and 0.047). MTF cut off would be higher with greater increase of ET in central and paracentral zones respectively (β = 0.96 and 0.83; 95% CI: 0.44 to 1.48, and 0.03 to 1.63; P < 0.001 and = 0.042, Table 5).

Table 5.

Univariate Analysis between Epithelial Remodeling and Visual Quality at 2 Years after FS-LASIK Xtra^a

	Central		Paracentral		Pericentral		Standard deviation		Min-Max
	P	β (95% CI)	P	β (95% CI)	P	β (95% CI)	P	β (95% CI)	P	β (95% CI)
Postoperative SE	0.254	-0.03 (-0.07 to 0.02)	0.747	-0.01 (-0.06 to 0.04)	0.390	0.03 (-0.04 to 0.10)	0.237	0.08 (-0.05 to 0.21)	0.046*	-0.03 (-0.06 to 0.00)
Higher-order Aberrations
Coma	0.017*	0.02 (0.00 to 0.03)	0.047*	0.02 (0.00 to 0.04)	0.481	-0.01 (-0.05 to 0.02)	0.008*	0.13 (0.04 to 0.22)	0.010*	-0.03 (-0.06 to -0.01)
Trefoil	0.018	-0.01 (-0.02 to 0.00)	0.176	-0.01 (-0.02 to 0.00)	0.875	0.00 (-0.01 to 0.02)	0.079	0.03 (0.00 to 0.06)	0.021*	-0.01 (-0.02 to 0.00)
Spherical aberration	0.224	0.01 (-0.01 to 0.04)	0.136	0.02 (-0.01 to 0.04)	0.009*	-0.02 (-0.04 to -0.01)	0.673	0.01 (-0.04 to 0.06)	0.358	-0.01 (-0.02 to 0.01)
Secondary astigmatism	0.692	0.00 (-0.01 to 0.01)	0.957	0.00 (-0.01 to 0.01)	0.076	0.01 (0.00 to 0.02)	0.954	0.00 (-0.03 to 0.03)	0.779	0.00 (-0.01, 0.01)
Tetrafoil	0.690	0.00 (-0.01 to 0.01)	0.816	0.00 (-0.01 to 0.01)	0.541	0.00 (-0.01 to 0.01)	0.089	-0.01 (-0.03 to 0.00)	0.088	0.00 (0.00, 0.01)
OQAS Measurements
OSI	0.089	-0.08 (-0.18 to 0.01)	0.827	-0.01 (-0.11 to 0.09)	0.051	0.09 (0.00 to 0.18)	0.015*	0.27 (0.06 to 0.48)	0.022*	-0.07 (-0.13 to -0.01)
MTF cut off	< 0.001*	0.96 (0.44 to 1.48)	0.042*	0.83 (0.03 to 1.63)	0.777	-0.17 (-1.39 to 1.05)	0.858	-0.25 (-3.09 to 2.59)	0.987	0.01 (-0.78 to 0.79)
Total QoV Scores	0.280	0.00 (0.00 to 0.00)	0.052	0.00 (0.00 to 0.00)	0.426	0.00 (0.00 to 0.00)	0.813	0.00 (0.00 to 0.00)	0.514	0.00 (0.00 to 0.00)

FS-LASIK Xtra=femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking; SE=spherical equivalent; OQAS=Optical Quality Analysis System; OSI=objective scatter index; MTF cut off=modulation transfer function cut off; QoV=Quality of Vision; ^ageneralized linear mixed model; ^*P < 0.050

Discussion

In the context of the increasing demands for surgical myopic correction in China, this study comprehensively evaluated the stability of refractive status, corneal topography and biomechanics, and epithelial remodeling over a 2-year period after FS-LASIK Xtra for borderline eyes. Meanwhile, it compared subjective and objective visual quality with that of well-conditioned eyes receiving conventional FS-LASIK at 2 years postoperative. No severe adverse event was observed. Postoperative SE or cylinder did not vary between two surgical groups, consistent with the 1-year reports [6, 7, 12]. The proportion of preoperative CDVA greater than 20/20 was higher in FS-LASIK Xtra group. Conversely, postoperative UDVA was better after FS-LASIK. Some studies had reported a delayed visual rehabilitation after FS-LASIK Xtra [13]. Once the edema and haze subsided, visual acuity became comparable to or even exceeded that of non-prophylactic CXL group [14]. The discrepant outcomes might be explained by the longer follow-up. A myopic shift of SE was shown at 2 years compared to early period in this study. We also acknowledge that although two surgical groups were matched and verified to have no statistical differences in baseline characteristics, there were imbalanced distributions of refractive status and slightly larger standard deviations in FS-LASIK Xtra group. Two eyes had preoperative SE reaching − 10.00D and two eyes had astigmatism of -4.00D in FS-LASIK Xtra group. These eyes demonstrated relatively poorer UDVA after surgery, which could potentially introduce bias in the mean value. Moreover, we should also pay attention to the increased CD following CXL.

Borderline eyes after FS-LASIK Xtra appeared to achieve a similar level of corneal morphology compared to those undergoing FS-LASIK alone at this 2-year visit. The surface asymmetry index was reported to be larger after FS-LASIK concurrent with accelerated CXL but measured by different topography device in earlier postoperative stage [15]. Meanwhile, topometric indices, particularly maximum keratometry and deviation in the back elevation, remained stable in longitudinal study, which was consistent with the findings of CXL for ectasia disease [16, 17]. Parameters such as index of surface variance and index of vertical asymmetry, might have their sensitivity affected by individual variability in corneal patterns. Keratoconus index and central keratoconus index have proven to be valuable tools for identifying forme fruste keratoconus and screening before refractive surgery [18, 19]. In light of the results without statistical significance, we hypothesized that prophylactic CXL might provide benefits in reducing corneal irregularity.

Biomechanical parameters were compared between surgical groups 2 years postoperatively. Crosslinking reduced the elasticity of cornea obviously by enhancing the covalent binding of collagen [15]. SP-A1 and ARTh, which performed worse before surgery, became comparable after additional prophylactic CXL. On the other hand, corneal hysteresis and corneal resistance factor remained stable throughout the previous 6- and 12-month observations [14, 20]. Corvis output quantifies the displacement of cornea rather than reporting signal intensity of the reflected infrared light represented as corneal resistance factor [21]. Our study showed improvements in SP-A1, IR, and ARTh from early period up to 2 years after FS-LASIK Xtra, which was accompanied by a trend toward increased thinnest corneal thickness. Apart from epithelial thickening, if posterior corneal elevation could be continuously measured over a longer time scale, stromal remodeling and biomechanical changes could be better discussed. In any case, no sign of postoperative ectasia was observed with a total energy delivery of 2.7 J/cm², and patients reported good tolerance, although longer follow-up is still necessary for further investigation.

We also evaluated the corneal epithelial remodeling. Previous research described the “negative meniscus-like lenticular pattern” of the remodeling after FS-LASIK [22]. A similar remodeling pattern was observed in early period after FS-LASIK Xtra [20, 23], and its degree in high myopic patients was comparable to those receiving lower myopic corrections [24]. Our study with longer follow-up was consistent with them, showing less thickening in FS-LASIK Xtra group and significant variation in pericentral zone. This deviation might imply the long-term efficacy of prophylactic CXL in improving biomechanics and stabilizing corneal curvature, the latter was the driving force of uneven epithelial distribution for compensation [25].

Interestingly, central ET appeared to thicken significantly at 2 years compared to 6 months after FS-LASIK Xtra. The combination of CXL has been reported to alter the extent of remodeling without affecting its temporal progression over a 6-month study, and the thickening became “quiet” since the third month postoperative [23]. We interpreted this pronounced remodeling as the long-term cumulative consequence of continuous thickening after surgery, rather than an early sign of ectasia. This was supported by our observation that the thickening occurred consistently across all 17 sections with or without significance, which was distinct from the pattern with localized thinning [26]. Concurrently, corneal morphology and biomechanics were improved as discussed above.

A comprehensive evaluation of objective and subjective visual quality after FS-LASIK Xtra were conducted in this study. First, OQAS measurements especially OSI, characterizing the scatter of refractive medium [27], showed superior performance in FS-LASIK group. Variations in collagen fiber diameter and regularity, as well as elevated CD, could affect visual quality by raising forward scattering after CXL [16, 28]. And CD is frequently observed to increase at the depth where the demarcation line appears following CXL [29]. Correspondingly, in our study, CD of central layer in FS-LASIK Xtra group was higher than FS-LASIK group. This parameter was reported to remain above baseline at 12 months after CXL [30], but decrease remarkably in long-term after FS-LASIK [31]. Therefore, individuals with higher requirements for visual quality should carefully consider the option of FS-LASIK Xtra. In addition, eyes with high degree of myopia or astigmatism in FS-LASIK Xtra group exhibited poorer OSI and MTF cut off values simultaneously. Given that CD is not significantly associated with preoperative SE in previous study [31], and the degree of myopia has an influence on retinal image quality and scattering [32], we could not attribute the lower visual quality solely to CD.

Second, the combination of prophylactic CXL did not result in additional change of HOAs. HOAs tend to decline with the flattening of corneal after CXL for keratoconus [33], with spherical aberration and coma induced by decentred cones being most affected [34]. The reduction also has been observed after photorefractive keratectomy combined with CXL [35]. But corneal HOAs might change negligibly in statistics after accelerated CXL [36]. There was another study suggested that the changes of HOAs over 6 months in FS-LASIK Xtra group were similar to those in FS-LASIK group [23]. Our 2-year findings seemed to align with them. Recently, Zou et al. [37] found that corneal HOAs might also influence OSI and MTF cut off value after FS-LASIK. In conjunction with our comparative analysis, lower objective visual quality after FS-LASIK Xtra was more likely to correlate with the increase of CD as previously described.

Contrast sensitivity values of two surgical groups were not distinguished from each other. It is well established that contrast sensitivity would reduce after refractive surgery. Then, due to the potential compensatory effects of brain on HOAs, contrast sensitivity would gradually return to or precede preoperative levels, and remain stable throughout a 5-year period [38, 39]. Damage of contrast sensitivity in keratoconus is significantly correlated with the irregular corneal astigmatism, which could serve as a tool reflecting the severity indirectly [40]. We speculated that the comparable contrast sensitivity in our results might indicate the stability of cornea at 2 years postoperatively.

With respect to subjective visual quality, there was no difference in frequency, severity, bothersome, and total scores of QoV questionnaire between surgical groups. In other words, the objectively measured increase in light scatter did not translate into a worse visual experience. Patients receiving FS-LASIK mainly complained about the visual symptoms of starbursts, glare, and halos, which were correlated with HOAs caused by the slight flap deflection and spherical aberration induced by laser [41]. The disparity in score of individual visual symptom between groups was only found in visual fluctuation, which might involve the tear film stability. Some of the patients with high degree of myopia or dependence on spectacles might have a more reasonable expectation of the surgery [42], and report a higher QoV score. However, the retrospective design led to the absence of preoperative questionnaire results for comparison.

The objective evaluations of tear film stability, ocular surface damage, and subjective feelings assessed by OSDI scores showed no significant difference between two surgical groups at this 2-year visit. The transection of corneal nerve fibers during FS-LASIK affects the corneal sensitivity and the concentration of inflammatory mediators, especially in short term after surgery [43]. CXL could help improve corneal symmetry, and the good compensatory response of epithelium further smooths the surface irregularity [44]. Our study revealed that the application of prophylactic CXL would not cause additional impairment of tear film stability.

Our research group had reported that the uniform distribution of ET might be associated with myopic regression and spherical aberration in patients with high myopia 5 years after FS-LASIK [45]. In this study, there was no statistical correlation between SE and ET thickening in 3 annuluses at 2 years after FS-LASIK Xtra, though both parameters changed significantly from early postoperative period. At the same time, the weakened relationship between epithelial thickening and HOAs also suggested that the application of prophylactic CXL might reduce the amount of remodeling [23]. MTF cut off value would decrease with the inhibition of remodeling, which might be explained by the efficacy of prophylactic CXL [46]. The evidence was that, in individual patients with low level of epithelial thickening, light demarcation lines could still be observed at 2-year visit.

This study has several limitations. Surgical correction in patients with suboptimal corneal conditions necessitates rigorous and repeated evaluation of indication, and a portion of young patients failed to complete postoperative visits after enlisting the military. These reasons resulted in a relatively small sample size of this study and restricted more detailed stratified analyses of variables influencing corneal morphology and biomechanics, including age, magnitude of refractive correction, and corneal flap settings. Patients could be identified as “borderline” for multiple risk factors, and the considerable within-group variability might diminish the sensitivity of certain parameters, such as cylinder and topometric indices, and limited further refinement of surgical indications. In the future, only through rigorously and properly conducted prospective cohort studies in borderline patients (with or without the addition of prophylactic crosslinking),we could better confirm its advantages and long-term efficacy in stabilizing corneal morphology and biomechanics, and ultimately establish the necessity of this procedure.

Conclusions

In summary, the combination of prophylactic corneal crosslinking appeared to improve the corneal morphology and biomechanics in the borderline eyes, bringing them to a level comparable to that in eyes without risk factors and receiving FS-LASIK alone, under the premise of ensuring the efficacy of myopic correction over this 2-year follow-up. It should be noted that the properties of CXL seemed to increase light scatter characterized by higher corneal densitometry and OSI, although the subjective visual experience seemed not to be disturbed. And epithelial remodeling correlated to corneal stability and visual quality still requires long-term monitoring. For younger patients with non-ideal corneal conditions but having a strong desire or special occupational requirement for being glasses-free, FS-LASIK Xtra might be an option. However, its use should be based on strict assessments of indications and thorough preoperative communication with patients regarding potential visual outcomes, rather than being applied indiscriminately to all patients.

Abbreviations

FS-LASIK Xtra

Femtosecond laser-assisted in situ keratomileusis combined with prophylactic corneal crosslinking

CXL

Corneal crosslinking

CDVA

Corrected distance visual acuity

UDVA

Uncorrected distance visual acuity

CD

Corneal densitometry

ET

Epithelial thicknesses

OSI

Objective scatter index

MTF cut off

Modulation transfer function cut off

OQAS

Optical Quality Analysis System

QoV

Quality of Vision

OSDI

Ocular Surface Disease Index

CI

Confidence interval

SE

Spherical equivalent

SP-A1

Stiffness parameter at the first applanation

ARTh

Ambrósio relational thickness

IR

Integrated radius

HOAs

Higher-order aberrations

Author contributions

FZ provided supervision and final approval, made critical revision of the manuscript and fund application, designed the research, and conducted the surgeries. JL and YH conducted the research, collected data, performed statistical analysis, and wrote the manuscript. YX, ZW and YS interpreted data and made revision of the manuscript. All authors read and approved the final manuscript.

Funding

Supported by Beijing Municipal Science&Technology Commission No. Z201100005520043.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in compliance with the Declaration of Helsinki and was approved by the ethics committee of Beijing Tongren Hospital, Capital Medical University (TRECKY2021-216). Informed consent was obtained from all patients.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.