An opaque bubble layer (OBL) is the accumulation of water vapor and carbon dioxide in the corneal tissue. OBL formation is related to photodisruption by a femtosecond laser. In recent years, many studies have explored possible risk factors for OBL formation during femtosecond laser–assisted stromal in situ keratomileusis (FS-LASIK) procedures. Thicker corneas, using a hard-docking technique, more corneal astigmatism, and a steeper corneal curvature, have been associated with an increased incidence of OBL in FS-LASIK.1–3
Some have also focused on OBL formation during small incision lenticule extraction (SMILE) procedures. Although the occurrence of OBL has not been shown to have an effect on long-term clinical outcomes, it may interfere with lenticule separation and extraction.4 Li et al5 suggested that a thicker corneal cap could reduce the risk of OBL. Furthermore, Liu et al6 reported that SMILE procedures with a 150-µm cap thickness resulted in less OBL formation than those with a 110-µm cap thickness. However, to date, there has been no objective, quantitative study evaluating the impact of cap thickness on OBL.
We analyzed the influence of cap thickness on changes in corneal curvature and biomechanical parameters after SMILE in an earlier study.7 In the current study, we aimed to investigate the effect of corneal cap thickness on the formation of OBL in SMILE cases.
Patients and Methods
The current prospective, contralateral eye study was performed from January 2018 to June 2019 at Shenyang Aier Eye Hospital. In total, 50 patients (100 eyes) who underwent SMILE for the correction of myopia for both eyes were recruited. Each patient had surgery performed with a cap thickness of 110 µm in one eye and 140 µm in the other eye. Selection of cap thickness for each eye was randomized via coin toss. To ensure that a similar lenticule thickness was acquired in both eyes, the difference in spherical equivalent between the two eyes was set to less than 0.50 diopters (D). The current study adhered to the tenets of the Declaration of Helsinki and was approved by the ethics committee of Shenyang Aier Eye Hospital. All patients provided signed informed consent before surgery.
All surgeries were performed by the same surgeon (XF) using a 500-Hz VisuMax femtosecond laser system (Carl Zeiss Meditec). Patients were asked to fixate on a green target light and, when a centration with 80% contact was achieved, suction was applied. Two corneal cap thickness settings were used: one eye was randomized to receive a 110-µm cap thickness and the other eye a 140-µm cap thickness. All other surgical parameters were the same between patients and are described in detail in Table 1. After laser treatment, the lenticule was separated and extracted via a side-cut incision. The side-cut angle was 90° in the 12-o'clock position with a width of 2 mm. Following surgery, patients received conventional anti-inflammatory agents for 1 month.
OBl Measurement and Classification
Each operation was automatically video-recorded with the VisuMax femtosecond laser storage system. Because there is no standardized method for the quantitative analysis of OBL, we referenced methods from a previous study.1 OBL measurements were taken following two steps.
Step 1 was the image extraction step. During this step, videos were imported into Photoshop software (Adobe Systems) for screen capture. The scanning of the posterior (lower) surface of the lenticule was completed immediately, and images were exported in JPEG format and defined as phase I OBL. After the anterior (upper) surface of the lenticule was completed (before side-cut scanning started), images were again exported and defined as phase II OBL.8
The scanning area for photographs was selected using the elliptical marque tool of the Photoshop software and saved in the TIF format. Single images of each phase of OBL were selected at the same size.
Step 2 was the image analysis step. ImageJ software (National Institutes of Health) was used and images were converted into an 8-bit format. Then, OBL areas were selected and highlighted in red by setting “Threshold” values to 65 to 90. The ImageJ software operation used involved the following steps: (1) “Image”?“Type” ?“8 bit”; (2) “Image”?“Adjust”?“Threshold.” The percentage of OBL in total scanning area (%area) and mean gray values were obtained.
ImageJ software was used to analyze the OBL area. This software can notably mark some scattered OBL, which is not easily distinguishable to the naked eye. So, two primary categories of phase II OBL were determined. According to the distribution characteristics of OBL areas, we proposed the following classification for phase II OBL: OBL with a concentrated distribution was grouped into one category and defined as focal OBL (Figure AA, available in the online version of this article), whereas OBL with a scattered distribution was defined as scattered OBL (Figure AB).
(A) Example of a focal opaque bubble layer (Area%: 14.51%, mean gray value: 72.23) and (B) scattered opaque bubble layer (Area%: 16.80%, mean gray value: 69.13).
Statistical software (SPSS 23.0; IBM Corporation) was used to analyze all data. Continuous variables were described as means ± standard deviations. The Shapiro–Wilk test was used to test data normality. Normally distributed data were analyzed using the two-tailed paired t test and non-normally distributed data using the Wilcoxon signed-rank test. A chi-squared test was used to analyze differences between the two categories of phase II OBL and the two different corneal cap thickness groups. P values less than .05 were considered statistically significant.
All SMILE procedures were performed successfully. The mean age of the 50 patients (21 men, 29 women) was 25.0 ± 5.1 years (range: 18 to 34 years). The mean pre-operative spherical equivalent of tested eyes was −4.90 ± 0.96 D (range: −2.75 to −7.63 D) and the mean central corneal thickness was 551.02 ± 25.39 µm (range: 506.0 to 609.00 µm). Other baseline characteristics of the 110 and 140-µm cap thickness groups are described in Table A (available in the online version of this article). The refractive outcomes were shown in a previously published study conducted by our research group.7
Comparison of Baseline Characteristics Between the Eyes in the Two Study Groups (N = 50)
The study used two parameters for the quantitative analysis of OBL: percent area, which is the percentage of OBL area found in the laser scanning area, and mean gray value. The phase I OBL area (%) was higher in the 110-µm group than in the 140-µm group (respectively: 1.32% ± 1.20%, range: 0.21% to 7.67%; 0.94% ± 0.59%, range: 0.22% to 2.90%, <I>P = .002). The phase II OBL percent area in the 110-µm group was significantly higher than in the 140-µm group (respectively: 11.70% ± 7.35%, range: 0.79% to 35.71%; 6.64% ± 4.68%, range: 0.59 to 23.02%, P < .001). There was a significant difference (P < .001) in the mean gray value of the phase I OBL between the 110-µm group (71.84 ± 0.83, range: 70.94 to 75.87) and the 140-µm group (72.11 ± 0.80, range: 71.10 to 75.50). But there was no significant groupwise difference (P = .055) in the phase II OBL between the 110-µm group (70.17 ± 1.10, range: 68.79 to 72.90) and the 110-µm group (69.74 ± 0.80, range: 71.10 to 75.50) (Figure 1).
Comparison of opaque bubble layer (OBL) between the 110-and 140-µm groups. *P < .05.
In the current study, OBL that occurred in the cut of the corneal cap (phase II OBL) was categorized into two subgroups: focal OBL and scattered OBL. Focal OBL was observed in 64% (32 of 50) of eyes in the 110-µm group, which was significantly more than that observed in the 140-µm group (32%; 16 of 50 of eyes; chi-square = 10.256, P = .001) (Figure 2). To further determine the factors that were associated with the focal OBL and scattered OBL, we compared preoperative parameters between the two subgroups. The preoperative central corneal thickness of the focal OBL subgroup was significantly higher than that of the scattered OBL subgroup in the 110-µm group (P = .028). There were no significant differences in baseline characteristics between the two subgroups in the 140-µm group (Table 2).
Comparison of the incidence of scattered and focal opaque bubble layer (OBL) between the two study groups. *P < .05.
Comparison of Preoperative Parameters Between the Two Subgroups (N = 50)
OBL is a common complication of SMILE.9 The current prospective, contralateral eye study suggests that cap thickness is an independent risk factor for intra-operative OBL in SMILE patients. Our research also confirmed the correlation between scanning depth and OBL shown in a previous study5: deeper cutting plane was associated with smaller OBL. To the best of our knowledge, the current study is the first to classify OBL in the anterior lenticule plane using an ImageJ-based approach. We further found that eyes with a thinner corneal cap may be more likely to develop focal OBL.
Previous studies have shown that the generation of OBL may be affected by differences in flap thickness in FS-LASIK. For instance, Tang et al10 analyzed three different flap thicknesses (100, 110, and 120 µm) and found that a thin flap thickness was more likely to lead to the development of OBL. Lim et al11 found that the incidence of OBL in the 80-µm flap thickness group was approximately twice that in the 120-µm flap thickness group following VisuMax 500-kHz FS-LASIK. In a later study, Lim et al12 reported that eyes with an 80-µm flap had more frequent OBL than eyes with 90- or 100-µm flaps in the FS-LASIK procedure.
Based on our own surgical experience, differences in OBL characteristics between patients are mainly in OBL area and density. Thus, we used proportional OBL areas to describe the sizes of affected areas and mean gray values to assess for the density levels of OBL. We found that when SMILE was conducted using a 110-µm cap thickness, there was a larger OBL area in both the lower and upper lenticule planes than the 140-µm cap thickness. When the same cap thickness was used, the area of phase II OBL was significantly greater than that of phase I OBL. In the current study, patients underwent SMILE with all of the same treatment parameters besides cap thickness. Our results demonstrate that a thicker corneal cap is associated with a significantly lower OBL area.
A previous study by Son et al13 analyzed OBL development at the posterior lenticular cutting plane. In this study, the mean OBL area was 3.06% ± 4.62% of the total cornea, greater than that seen in the current study (1.32% ± 1.20% in the 110-µm group and 0.94% ± 0.59% in the 140-µm group). One possible explanation for this is the difference in surgical parameters used in each study. Son et al13 found that eyes with thicker corneas or thinner lenticules, which could render the cutting plane more shallow, may be more likely to develop OBL. Li et al5 further reported that the risk of OBL was reduced with deeper laser scanning. These studies agreed with our conclusion that increasing cap thickness deepened scanning position depth and reduced OBL occurrence. When using a thinner cap for SMILE, the cutting plane draws adjacent to the superficial anterior corneal stroma, which has more compact interwoven lamellae,14 allowing gas bubbles to become trapped and accumulate in the tissue. This may explain our observations made in the current study.
To determine the role of cap thickness in OBL occurrence, we also sought to describe the density of OBL in SMILE eyes. OBL with a grayish or more transparent appearance had a lower mean gray value in the current study. Given that our analyses of mean gray values differed little between the 110- and 140-µm groups, the thickness of the corneal cap mainly affects OBL occurrence area, but not the density. One possible reason for this is that a thinner cap, located in a highly arranged collagen structure, obstructs mixed gas diffusion and thus increases the OBL area.
Li et al5 and Ma et al8 reported a classification system for OBL in SMILE. They both classified OBL, which developed at the lenticule planes into phases I and II, respectively. Phase I OBL was subdivided into four levels according to the maximum covered distance, and phase II OBL was subdivided into central OBL and diffuse OBL. In the current study, we also defined OBL as either phase I or phase II based on different lenticule scanning phase. We found that the percent area of phase I OBL was significantly less than that of phase II OBL in both groups (approximately one-eighth). Given this, we only classified phase II OBL into different subtypes (focal OBL and scattered OBL).
The OBL classification system developed here, which was based on the different distribution characteristics of OBL areas, as identified using ImageJ software analysis, used two subgroups: focal OBL and scattered OBL. Because scattered OBL could easily be missed during surgery and may form after gas bubbles escape from a more focal OBL, we compared preoperative parameters between the two subgroups. In doing this, we found that central corneal thickness was greater in patients with focal OBL than in those with scattered OBL in the 110-µm group. Eyes with a 110-µm cap thickness had more focal OBL than those with a 140-µm cap thickness. These results indicate that central corneal thickness and cap thickness are associated with the development of obvious OBL.
A limitation of this study is that we proposed a classification of phase II OBL according to the distributed characteristics of OBL areas, but there is no objective criterion. We also did not investigate the influence of scattered OBL on surgery and surgical outcomes. Further studies evaluating differences between the two categories (focal and scattered) will be needed.
Our results reveal that increasing corneal cap thickness may reduce intraoperative OBL in SMILE used for the correction of moderate to low myopia. Our observations could help surgeons optimize surgical design and acquire better surgical results in SMILE.
- Courtin R, Saad A, Guilbert E, Grise-Dulac A, Gatinel D. Opaque bubble layer risk factors in femtosecond laser-assisted LASIK. J Refract Surg. 2015;31(9):608–612. doi:10.3928/1081597X-20150820-06 [CrossRef]
- Wei CH, Dai QY, Mei LX, Ge Y, Zhang PF, Song E. Paired eye-control study of unilateral opaque bubble layer in femtosecond laser assisted laser in situ keratomileusis. Int J Ophthalmol. 2019;12(4):654–659.
- Jung HG, Kim J, Lim TH. Possible risk factors and clinical effects of an opaque bubble layer created with femtosecond laser-assisted laser in situ keratomileusis. J Cataract Refract Surg. 2015;41(7):1393–1399. doi:10.1016/j.jcrs.2014.10.039 [CrossRef]
- Wang Y, Ma J, Zhang J, et al. Incidence and management of intraoperative complications during small-incision lenticule extraction in 3004 cases. J Cataract Refract Surg. 2017;43(6):796–802. doi:10.1016/j.jcrs.2017.03.039 [CrossRef]
- Li L, Schallhorn JM, Ma J, Zhang L, Dou R, Wang Y. Risk factors for opaque bubble layer in small incision lenticule extraction (SMILE). J Refract Surg. 2017;33(11):759–764. doi:10.3928/1081597X-20170821-02 [CrossRef]
- Liu T, Yu T, Liu L, Chen K, Bai J. Corneal cap thickness and its effect on visual acuity and corneal biomechanics in eyes undergoing small incision lenticule extraction. J Ophthalmol. 2018;2018:6040873. doi:10.1155/2018/6040873 [CrossRef]
- Wu D, Liu C, Li B, Wang D, Fang X. Influence of cap thickness on corneal curvature and corneal biomechanics after SMILE: a prospective, contralateral eye study. J Refract Surg. 2020;36(2):82–88. doi:10.3928/1081597X-20191216-01 [CrossRef]
- Ma J, Wang Y, Li L, Zhang J. Corneal thickness, residual stromal thickness, and its effect on opaque bubble layer in small-incision lenticule extraction. Int Ophthalmol. 2018;38(5):2013–2020. doi:10.1007/s10792-017-0692-2 [CrossRef]
- Krueger RR, Meister CS. A review of small incision lenticule extraction complications. Curr Opin Ophthalmol. 2018;29(4):292–298. doi:10.1097/ICU.0000000000000494 [CrossRef]
- Tang SJ, Xue JS, Zhang SJ, et al. Clinical observation and related factor analysis of opaque bubble layers produced by Visumax femtosecond laser-assisted LASIK. Chin J Ophthalmol Vis Sci. 2019;(21)1:30–34. doi:10.3760/cam,j.issn.1674-845X.2019.01.006 [CrossRef]
- Lim DH, Keum JE, Ju WK, Lee JH, Chung TY, Chung ES. Prospective contralateral eye study to compare 80- and 120-µm flap LASIK using the VisuMax femtosecond laser. J Refract Surg. 2013;29(7):462–468. doi:10.3928/1081597X-20130617-04 [CrossRef]
- Lim DH, Hyun J, Shin E, Ko BW, Chung ES, Chung TY. Incidence and risk factors of opaque bubble layer formation according to flap thickness during 500-kHz FS-LASIK. J Refract Surg. 2019;35(9):583–589. doi:10.3928/1081597X-20190814-01 [CrossRef]
- Son G, Lee J, Jang C, Choi KY, Cho BJ, Lim TH. Possible risk factors and clinical effects of opaque bubble layer in small incision lenticule extraction (SMILE). J Refract Surg. 2017;33(1):24–29. doi:10.3928/1081597X-20161006-06 [CrossRef]
- Steven JP. Depth-dependent transverse shear properties of the human corneal stroma. Invest Ophthalmol Vis Sci. 2012;53(2):873–880. doi:10.1167/iovs.11-8611 [CrossRef]
|Cap thickness||110 or 140 µm|
|Spot distance||4.5 µm|
|Optical zone diameter||6.5 mm|
|Cap diameter||7.6 mm|
|Transition zone diameter||0.1 mm|
|Scan direction||Posterior surface of the lenticule: spiral in, anterior surface of the lenticule: spiral out|
Comparison of Preoperative Parameters Between the Two Subgroups (N = 50)a
|Parameter||110-µm Cap Thickness Group||140-µm Cap Thickness Group|
|Focal OBL Subgroup||Scattered OBL Subgroup||P||Focal OBL Subgroup||Scattered OBL Subgroup||P|
|Preoperative CCT (µm)||557.2 ± 23.4||540.6 ± 27.4||.028b||557.1 ± 20.2||547.9 ± 26.9||.229|
|Preoperative sphere (D)||−4.69 ± 0.92||−4.27 ± 0.96||.144||−4.67 ± 1.06||−4.41 ± 1.02||.411|
|Preoperative cylinder (D)||−0.77 ± 0.57||−0.69 ± 0.50||.444||−0.80 ± 0.56||−0.76 ± 0.54||.812|
|Preoperative SE (D)||−5.07 ± 0.95||−4.63 ± 0.95||.188||−5.07 ± 1.03||−4.79 ± 0.94||.347|
|Preoperative Km (D)||43.60 ± 1.28||43.48 ± 1.14||.734||43.13 ± 1.27||43.58 ± 1.64||.643|
|RST (µm)||347.8 ± 23.5||337.9 ± 26.1||.174||317.1 ± 15.1||311.7 ± 25.1||.425|
Comparison of Baseline Characteristics Between the Eyes in the Two Study Groups (N = 50)a
|Parameter||110-µm Cap Thickness Group||140-µm Cap Thickness Group||t/Z||P|
|Preoperative sphere (D)||−4.54 ± 0.95||−4.50 ± 1.03||−0.802||.426|
|Preoperative cylinder (D)||−0.625 (−1.00, −0.25)||−0.625 (−1.25, −0.25)||−0.588||.577|
|Preoperative SE (D)||−4.91 ± 0.96||−4.88 ± 0.97||−0.681||.499|
|Preoperative IOP (mm Hg)||16.85 ± 1.89||16.70 ± 1.98||0.855||.397|
|Preoperative CCT (µm)||551.18 ± 25.96||550.86 ± 25.07||0.444||.659|
|Km-ant at 2-mm zone (D)||43.555 ± 1.23||43.524 ± 1.169||0.57||.571|
|Km-ant at 4-mm zone (D)||43.512 ± 1.213||43.478 ± 1.176||0.78||.440|
|Km-ant at 6-mm zone (D)||43.357 ± 1.222||43.347 ± 1.187||0.254||.800|
|RST (µm)||344.26 ± 24.68||313.40 ± 22.36||26.877||< .001b|