Tear Menisci After Laser In Situ Keratomileusis With Mechanical Microkeratome and Femtosecond Laser

Abstract

Purpose.: To investigate the effect on tear menisci after laser in situ keratomileusis (LASIK) with flap creation by either microkeratome or femtosecond laser.

Methods.: Sixty eyes of 30 myopes were analyzed. Fifteen patients underwent LASIK with Moria II microkeratome, and the other 15 patients with 60-KHz IntraLase femtosecond laser. Upper and lower tear meniscus parameters of height (UTMH, LTMH) and area (UTMA, LTMA) were measured by SD-OCT preoperatively and 1 week, 1 month, and 3 months postoperatively.

Results.: Compared with the baseline values, all tear meniscus parameters decreased significantly at each postoperative time point (all P < 0.01) in both groups. LTMH increased significantly between 1 week and 1 month and between 1 and 3 months after surgery in the microkeratome (both P < 0.01) and femtosecond laser groups (P < 0.01, P = 0.012, respectively). There were significant increases in LTMA between 1 week and 1 month after surgery in the microkeratome group (P < 0.01) and in the femtosecond laser group (P = 0.028). There were no significant differences in UTMH, UTMA, LTMH, or LTMA between two groups. The depth of ablation was negatively correlated with the LTMA at 1 week after surgery (R = −0.256, P = 0.049) for all patients.

Conclusions.: There were no significant differences in the tear meniscus parameters between the microkeratome and femtosecond laser groups. The depth of ablation was significantly correlated with the LTMA only at 1 week after surgery.

Introduction

As a safe and effective correction method for myopia, laser in situ keratomileusis (LASIK) is the most commonly performed corneal refractive surgery throughout the world. ¹ Among the complications of LASIK, dry eye is considered to be the most common complaint after surgery, and it frustrates both patients and surgeons. ² Dry eye may be associated with creation of the corneal flap and stromal ablation. ^3,4 During these procedures, the corneal afferent nerve fibers are transected, resulting in a decrease in the corneal reflexes and blinking rate and a disruption of the neurotrophic factors released from the corneal nerves. ^2,5,6 These factors are important for corneal epithelial cells. 

Mechanical microkeratomes and femtosecond lasers are used to create corneal flaps during LASIK surgery. Compared with microkeratomes, femtosecond lasers provide better predictability of flap dimensions and improve the quality of the optical surfaces. ⁷ These two factors may reduce the incidence of postoperative dry eye. ⁸ However, whether the femtosecond laser reduces the incidence or decreases the severity of dry eye symptoms compared with the microkeratome is controversial. ^6,9,10 Traditional dry eye evaluations, such as a dry eye questionnaire, Schirmer test, tear breakup time, and corneal fluorescein staining, which exhibit fairly good reliability in clinics, have been widely applied in previous studies ^6,10–14 ; however, the conflicting results of these evaluations have been yielded because many factors exist that can influence the outcomes of these evaluations. ^15–19  

The tear meniscus holds approximately 75% to 90% of the total tears ²⁰ and has been suggested as a useful measure in monitoring the tear volume. ^21,22 The importance of tear meniscus evaluation in the diagnosis of tear deficiency has been recognized, and it has been used in many dry eye studies. ^15,16,23,24 Nevertheless, assessing and evaluating the tear meniscus, especially the upper meniscus, in a noninvasive manner is difficult. ¹⁵ However, with advances in technology, such as anterior segment spectral-domain optical coherence tomography (SD-OCT), the upper and lower tear menisci can now be evaluated noninvasively and quantitatively. ^25,26 With high-resolution scans and high acquisition speeds, SD-OCT provides high sensitivity and specificity in diagnosis of dry eye. ²⁷ In the study reported here, we used a custom-built anterior-segment SD-OCT instrument that was described in previous studies ^26,28–35 to measure pre- and postoperative tear menisci in patients undergoing LASIK surgeries using either a microkeratome or a femtosecond laser to create the corneal flap. 

Methods

Patients and LASIK Procedures

This study was approved by the Research Review Board at Wenzhou Medical University. Each subject was treated in accordance with the tenets of the Declaration of Helsinki, and all the patients signed informed consent. The surgeries were performed at the Refractive Surgery Center of The Eye Hospital of Wenzhou Medical University. For inclusion into the study, patients had to be between 20 and 30 years old and have a stable refractive error during the previous 2 years (progressing by less than 0.5 diopters annually). They also had to have a spherical equivalent (SE) refraction between −1.00 diopters and −8.00 diopters and astigmatism of less than −2.00 diopters. The exclusion criteria were active ocular or systemic diseases, severe dry eye syndrome, a residual corneal stromal bed of less than 280 μm, preoperative best-corrected visual acuity less than 20/20, and postoperative enhancement. 

All patients underwent the following sequential examinations during screening: anterior segment OCT, visual acuity, manifest refraction, corneal topography, noncontact tonometry, slit-lamp biomicroscopy, funduscopy, and pachymetry. Corneal flap creation by microkeratome was performed by Shihao Chen, and flap creation by femtosecond laser was performed by Ye Yu. In the microkeratome group, a Moria II microkeratome (Moria, Antony, France) was used to create the corneal flaps, and all of the flap hinges were located at the nasal cornea. In the femtosecond laser group, a 60-KHz IntraLase FS Laser (Abbott Medical Optics, Inc., Santa Ana, CA, USA) was used to create the corneal flaps, and all of the flap hinges were located at nasal cornea, as well. A Zeiss Mel-80 scan flying-spot excimer laser (Carl Zeiss Meditec-Aesclepion, Jena, Germany) was used to perform further ablation in both groups. Patients in both groups were treated with corticosteroid eyedrops (Fluorometholone 0.1%; Allergan, Inc., Irvine, CA, USA) four times daily during the first week after surgery, then tapered over the following 4 weeks. Patients in both groups were also instructed to use artificial tears (Refresh Plus; Allergan, Inc.) four times daily for at least the first month after surgery. 

Anterior Segment OCT

A custom-built, high-speed, anterior segment SD-OCT instrument was used to measure the upper and lower tear menisci preoperatively and at 1 week, 1 month, and 3 months postoperatively. This OCT system was previously described. ^26,28–35 Briefly, the superluminescent diode-based light source (Superlum, Broadlighter, D840-HP; Superlumdiodes, Ltd., Moscow, Russia) provided a beam of 840-nm wavelength and 50-nm bandwidth with a 7.2-mm scan depth in air. The longitudinal resolution of this SD-OCT system was approximately 6 μm in tissue, the lateral resolution was approximately 15 μm, and the scan speed was 24 frames per second. A 12.5-mm vertical scanning width was used to scan across the central cornea (apex) to capture the upper and lower tear menisci simultaneously. All measurements were performed between 10:00 AM and 3:00 PM in a research laboratory where air conditioning and a humidifier were used to control the temperature (20°C–25°C) and humidity (40%–50%). All patients were asked to avoid using any eyedrops before the measurement. 

During OCT measurement, the patients were instructed to look at an external fixation target and blink normally. The upper and lower tear menisci were captured immediately after a full blink, and good images, indicated by the light reflex at the corneal apex, were selected for analysis. The size of each selected image was 2048 × 2048 pixels, corresponding to 7.72 (longitudinal) × 18.29 (lateral) mm. Measurements of lateral and vertical image distances in air were calibrated by using a Vernier caliper. The caliper was mounted in the SD-OCT beam and images of the caliper spindle surface were taken in one position. The spindle was then retracted a known distance and re-imaged. The two images were processed to obtain the number of pixels for the distance between the two positions of the spindle surface. This calibration process was done in both the vertical and horizontal directions. 

The raw images of the tear menisci were optically corrected by a refraction correction algorithm based on Snell's principle, ^33,34 assuming a tear refractive index of 1.337. ³⁶ This correction was only performed longitudinally to calculate tear meniscus area. No correction was needed at the surfaces of tear menisci and cornea because they were detected in air (Fig. 1). After refraction correction, the height and area of upper and lower tear menisci (UTMH, UTMA, LTMH, and LTMA) were measured by custom software. The detailed procedure of the image-processing software has been described, and the repeatability of these measurements was verified in our previous studies. ^28–32,35  

Figure 1

 

Measurements of upper and lower tear menisci using custom software.

Statistical Analysis

The Statistical Package for Social Sciences (SPSS version 19.0; IBM SPSS Statistics, IBM Corporation, Chicago, IL, USA) was used for the statistical analysis. Sample size and statistical power were calculated by Power Analysis & Sample Size (PASS version 11; NCSS, LLC, Kaysville, UT, USA). The results are expressed as means ± SDs. Comparisons of the baseline tear meniscus parameters between the two groups were conducted with the independent samples t-test. One-way ANOVA was performed to compare differences in tear meniscus parameters between the microkeratome and femtosecond laser groups. Repeated measures of ANOVA and least significant difference (LSD) were used for comparing upper with lower menisci and to determine if there were differences in tear menisci before and after surgery. Bonferroni correction was used in post hoc t-tests to adjust the multiple comparisons. Correlations between the upper and lower tear menisci and the depth of ablation were tested using the Pearson's correlation coefficient. Because we used data from both eyes of each individual, between-eye correlations of all tear meniscus parameters after surgery were analyzed using Pearson's correlation coefficient. The significance level was set to 0.05 (LSD) and 0.0125 (0.05/4, Bonferroni). 

Results

A total of 30 myopic patients (60 eyes) were included in this study; 15 patients (30 eyes) underwent LASIK with a femtosecond laser to create the corneal flap, and the other 15 patients (30 eyes) underwent LASIK with a microkeratome to create the corneal flap. There were no differences in the demographic or clinical data between the microkeratome and femtosecond laser groups (Table). In all 60 eyes, the depth of ablation had a significant negative correlation with the LTMA value at 1 week after the surgery (R = −0.256, P = 0.049; Fig. 2). There were no other significant correlations between the ablation depth and the tear meniscus parameters. There were also no significant postsurgical correlations between the two eyes of any of the tear meniscus parameters (all possible correlations, P > 0.05). 

Figure 2

 

Correlation between depth of ablation and lower tear meniscus area at 1 week after surgery for 60 eyes (R = −0.256, P = 0.049).

Table

 

Patient Demographic and Clinical Data

Characteristic	Group
	Microkeratome; n = 15 Patients, 30 Eyes	Femtosecond Laser; n = 15 Patients, 30 Eyes
Age, y
Mean ± SD	24.6 ± 2.9	24.0 ± 2.1
Range	20–30	21–28
Sex, n (%)
Female	8 (53.3)	7 (46.7)
Male	7 (46.7)	8 (53.3)
SE, diopter	−4.55 ± 0.83	−5.23 ± 0.96
Ablation depth, μm, ± SD	74.80 ± 12.23	81.63 ± 11.58
CCT, μm, ± SD	548.47 ± 24.39	534.87 ± 29.56

 CCT, central corneal thickness.

There were no significant differences in the baseline UTMH, UTMA, LTMH, or LTMA values between the microkeratome and femtosecond laser groups (t-test, P = 0.617, P = 0.335, P = 0.871, P = 0.546, respectively; Fig. 3). In both groups, all tear meniscus parameters decreased significantly at 1 week, 1 month, and 3 months postoperatively compared with the baseline values (all P < 0.01). After surgery, there were no significant differences in the upper tear meniscus parameters between the different time points for either group. However, there were significant increases in the LTMH values between 1 week and 1 month and between 1 month and 3 months after surgery in the microkeratome group (both P < 0.01) and in the femtosecond laser group (P < 0.01, P = 0.012, respectively). There were also significant increases in the LTMA values between 1 week and 1 month after surgery in the microkeratome group (P < 0.01) and in the femtosecond laser group (P = 0.028). There were slight increases in the LTMA values between 1 month and 3 months after surgery in the microkeratome group and the femtosecond laser group, but they were not statistically significant (P = 0.067, P = 0.221, respectively). 

Figure 3

 

Tear meniscus parameters before and after LASIK. Measurements were made before surgery and at 1 week, 1 month, and 3 months after LASIK in the microkeratome group (n = 30 eyes) and the femtosecond laser group (n = 30 eyes). (Top left) UTMH, (top right) LTMH, (bottom left) UTMA, (bottom right) LTMA. There were no significant differences in the UTMH, UTMA, LTMH, or LTMA values between the two groups (one-way ANOVA, P = 0.426, P = 0.469, P = 0.668, P = 0.514, respectively). *P < 0.05 compared with the preoperative value. °P < 0.05 compared with 7 days. ˆP < 0.05 compared with 1 month.

The variables of the lower tear menisci were all significantly higher than those of the upper tear menisci in each group (all P < 0.01). Post hoc t-tests supported these differences at each time point (all P < 0.01) except for the difference between LTMH and UTMH, which was not significant in the microkeratome group (t-test, P = 0.091), and in the femtosecond laser group (t-test, P = 0.043) at 1 week after surgery. There were no significant differences in the UTMH, UTMA, LTMH, or LTMA values between the microkeratome and femtosecond laser groups (one-way ANOVA, P = 0.426, P = 0.469, P = 0.668, P = 0.514, respectively). 

Discussion

The proposed mechanisms of LASIK-induced dry eye include destruction of the corneal nerve endings, decreased neurotrophic influences on the epithelial cells, decreased blinking rate, decreased normal and reflex stimulation of tear production, decreased tear film stability and distribution, increased evaporation of the tears, and loss of limbal goblet cells. ^2,3,6,14,37 Furthermore, injury to the corneal sensory nerves after refractive surgery could produce aberrant impulse discharges that might evoke sensations of dryness. ³⁸ Although none of the patients in either group of our study reported significant symptoms of eye dryness at any of the follow-up visits, previous studies indicated that dry eye occurred after LASIK surgery with both microkeratome and femtosecond laser. ^1,9 The incidence of corneal punctate epithelial keratopathy, as determined by slit-lamp drawing, was significantly higher in the microkeratome group than in the femtosecond group ⁶ ; however, no significant difference between the two groups in dry eye symptoms was found according to the dry eye questionnaire. ¹⁰  

The surgical technologies of both methods have improved, and the incidence of corneal punctate epithelial keratopathy has decreased in a similar fashion. In our study, no patients showed significant corneal staining or reported significant sensation of dryness after surgery. The decrease in tear volume might not have been severe enough to cause corneal staining or elicit subjective symptoms, so evaluating the tear menisci might detect tear deficiency before the appearance of significant symptoms. 

The advantages of the SD-OCT system used in our study include noninvasive measurements and the ability to measure the upper and lower menisci simultaneously. ³⁵ Invasive dry eye tests might cause false-positive or false-negative results, ^39,40 so SD-OCT measurement might more objectively assess the natural tear distribution. The same OCT system has been applied in previous studies of LASIK-associated dry eye patients, ³² mild self-reported dry eye patients, ³⁵ and contact lens wearers with dry eye symptoms. ²⁹ The tear meniscus parameters were decreased in these dry eye patients and were significantly correlated with other traditional dry eye tests. ^25,26,30 Therefore, anterior segment SD-OCT has the capacity to detect early changes in tear volume after LASIK surgery and thus can facilitate investigation of LASIK-induced dry eye throughout the follow-up period after surgery. It can also be used to compare outcomes with different corneal flap creation methods. As a quick and noninvasive test to evaluate dry eye, anterior segment SD-OCT is likely to have a higher degree of acceptance compared with traditional invasive dry eye tests among patients in our clinics. Furthermore, because tear meniscus parameters are highly correlated with the more invasive test results, measuring tear menisci by anterior segment SD-OCT has the potential to become a routine dry eye test at clinics in the future. 

In our study, the UTMH, UTMA, LTMH, and LTMA postoperative values all decreased significantly compared with the corresponding preoperative values in both groups. The LTMH and LTMA of both groups increased from 1 week to 1 month after surgery. Between 1 and 3 months, the difference was significant for LTMH but not for LTMA. We also found that in most cases the variables of the lower tear menisci were greater than those of the upper tear menisci in each group, a phenomenon that has been reported and attributed to the effect of gravity. ³⁰ An exception to that observation was that the difference between LMTH and UMTH was not significant at 1 week after surgery in both groups. This exception needs to be verified in future studies with larger sample sizes. 

The parameters of the lower meniscus may better represent the total tear volume than do those of the upper meniscus. ^25,30 Further, for the lower meniscus, the height measurement had a smaller order of magnitude than did the area measurement; therefore, changes in lower meniscus height might be a more sensitive indicator of dry eye than changes in area. Based on the changes in lower meniscus height, we found that LASIK significantly decreased the tear volume after surgery. Further, the meniscus height, and therefore tear volume, recovered continually from 1 week to 3 months after surgery in both the microkeratome and femtosecond laser groups, although the values at 3 months remained below the preoperative values. These findings are consistent with other studies and clinical observations that dry eye symptoms occur during the early stages after LASIK, peak at 1 month after surgery, and later decrease gradually during the first year after the surgery. ^1,41,42  

The depth of ablation is considered to be a risk factor of LASIK-induced dry eye due to its associated damage to the corneal sensory nerves and corneal innervation. ^1,3,42 However, one study found that there was no significant correlation between the depth of ablation and the incidence of dry eye at 1 month after LASIK surgery. ⁶ In our study, we found that the depth of ablation had a significant negative correlation with the LTMA at 1 week after surgery in all subjects (60 eyes). This finding indicated that at least within 1 week of LASIK, patients with a lower spherical equivalent diopter of myopia had a greater tear volume, although there may not be a significant difference in the self-reported symptoms or clinical signs of dry eye. At 1 month after LASIK surgery, the depth of ablation value has a smaller correlation with tear volume, and certain other factors, such as the number of goblet cells, could have more influence on the tear volume. ^2,9  

The means and SDs of the flap thickness for the femtosecond laser were lower than for the microkeratome laser; therefore, the femtosecond laser was more accurate in creating corneal flaps. Thus, theoretically, the femtosecond laser may be less invasive for the anterior corneal stroma and nerve endings. ^6,10 Salomão et al. ⁶ found that the incidence of LASIK-associated dry eye was significantly higher in the microkeratome group (46%) than in the femtosecond group (8%). However, there are several differences between their study and ours. First, they used Hansatome microkeratome (Bausch & Lomb, Rochester, NY, USA), whereas we used Moria II microkeratome. The hinges in all eyes were superior in their study, whereas in our study both groups had nasal hinges. Second, they used LASIK-induced neurotrophic epitheliopathy and the need for topical cyclosporine A treatment as indicators of dry eye, whereas we compared the values of tear menisci. Third, age is a well-documented risk factor for dry eye, and the mean age of the subjects in the study by Salomão ⁶ was 43 years (range, 20–72 years) in the femtosecond laser group and 45 years (range, 20–72 years) in the microkeratome group. In contrast, the mean age in our study was 24.0 ± 2.1 years (range, 21–28 years) and 24.6 ± 2.9 years (range, 20–30 years), respectively. In our study, the absence of significant differences between the microkeratome and femtosecond laser groups in tear menisci at 1 week, 1 month, and 3 months after LASIK surgery indicated that the femtosecond laser may not result in a lower incidence or severity of dry eye at the early stages after surgery. These objective test results are consistent with subjective questionnaire results used in another study. ¹⁰  

There are some limitations in this study. The first was the use of data from both eyes of each individual. Although using only one eye is likely to be inefficient, ⁴³ and between-eye correlations of tear meniscus parameters were not significant in this study. Because we found no significant differences in the tear meniscus parameters between the microkeratome and femtosecond laser groups, further studies with larger sample sizes and using one randomly chosen eye may be more valid. A second limitation was the calibration of the SD-OCT in air, which was necessary because calibration within the tear film is not possible. We performed refraction correction using the same refractive index of tears for all the subjects; however, instrumental error still existed in each measurement and in both groups. Although we may not have obtained the actual value of each tear volume, we were able to compare the parameters of the upper and lower menisci over time and between two groups. The refractive index of tear film could be slightly different between the two groups or between the upper and lower menisci. In fact, it might even fluctuate over time because the components in tear fluid were changing dynamically. We corrected the SD-OCT images to calculate meniscus area assuming a tear refractive index of 1.337, which might induce some error. Further studies may reveal to what extent this error influences the measurement of tear menisci. There are other factors, such as surgical suction time and IOP changes during the suction phase, that may be different for the microkeratome and femtosecond lasers during the creation of corneal flaps ^6,44 ; therefore future studies are needed to determine if these factors influence the tear menisci after surgery. Furthermore, regarding whether or not differences in tear menisci parameters exist between the microkeratome and femtosecond laser methods of flap formation, between different types of femtosecond lasers, and the possible effects of corneal flap hinge location and hinge angles, future investigations are warranted. ⁴⁵  

In summary, LASIK significantly decreased the upper and lower tear menisci after surgery, and recovery of the lower tear menisci took place over a period of 1 week to 3 months. There were no significant differences in the tear meniscus parameters with regard to flap creation by microkeratome or femtosecond laser. The depth of ablation was significantly correlated with the LTMA only at 1 week after surgery. 

Acknowledgments

We express our appreciation to Meixiao Shen, PhD, from Wenzhou Medical University for her help in SD-OCT calibration interpretation. We also express our gratitude to Britt Bromberg, PhD, of Xenofile Editing (www.xenofileediting.com) for his help in editing the manuscript. We also thank Jingwei Zheng, MD, from Clinical Research Center of Wenzhou Medical University for his help in statistical analysis. 

Supported by research grants from the National Key Technology R&D Program (2011BAI12B08, 2012BAI08B05), the Department of Education of Zhejiang Province (Y201328048), the Program for Key Science and Technology Innovation Team of Wenzhou (C20120009-03), and the Service Platform of Technological Innovation for Wenzhou Laser and Optoelectronic Industry of Zhejiang Province (2012E60011). 

Disclosure: W. Xie, None; D. Zhang, None; J. Chen, None; J. Liu, None; Y. Yu, None; L. Hu, None