Dry eye and corneal sensitivity after SMILE and FS-LASIK: a Meta-analysis

·Meta-Analysis·

Dry eye and corneal sensitivity after small incision lenticule

extraction and femtosecond laser-assisted in situ
keratomileusis: a Meta-analysis
Wen-Ting Cai1, Qing-Yu Liu1, Cheng-Da Ren1, Qing-Quan Wei2, Jun-Ling Liu1, Qian-Yi Wang1,
Ya-Ru Du3, Meng-Mei He1, Jing Yu1
1
Department of Ophthalmology, Shanghai Tenth People’s ● KEYWORDS: dry eye; corneal sensation; small incision lenticule
Hospital, Tongji University, School of Medicine, Shanghai extraction; femtosecond laser-assisted in situ keratomileusis;
200072, China Meta-analysis
2
Nanchang University, Nanchang 330031, Jiangxi Province, DOI:10.18240/ijo.2017.04.21
China
3
Nanjing Medical University, Nanjing 210029, Jiangsu Province, Cai WT, Liu QY, Ren CD, Wei QQ, Liu JL, Wang QY, Du YR, He
China MM, Yu J. Dry eye and corneal sensitivity after small incision lenticule
Correspondence to: Jing Yu. Department of Ophthalmology, extraction and femtosecond laser-assisted in situ keratomileusis: a
Shanghai Tenth People’s Hospital, Tongji University, School Meta-analysis. Int J Ophthalmol 2017;10(4):632-638
of Medicine, Shanghai 200072, China. [email protected]
Received: 2016-08-02 Accepted: 2017-02-06 INTRODUCTION

Abstract
● AIM: To assess the corneal sensitivity and the incidences
R efractory surgeries bring us many advantages as well
as some complications, such as dry eye and reduced
corneal sensation. The corneal nerves are cut during the
of dry eye after small incision lenticule extraction (SMILE) flap creation and stromal ablation, which reduce the corneal
and femtosecond laser-assisted in situ keratomileusis (FS- sensation, tear production and increase tear evaporation[1]. The
LASIK). exact pathophysiology of dry eye after refractive surgery still
● METHODS: The Meta-analysis was performed using remains uncertain. However, neurotrophic epitheliopathy due
RevMan 5.3. We searched on PubMed from inception to to the damage of sub-basal nerve plexus has been considered
March 2016. Summary weighted mean difference (WMD) one of the major contributing factors[2]. Refractive surgery
and 95% confidence intervals (CIs) were used to analyze is very popular for myopia in recent years, especially small
the datum. Random-effects or fixed-effects models were incision lenticule extraction (SMILE) and femtosecond laser-
chosen up to between-study heterogeneity. The main assisted in situ keratomileusis (FS-LASIK).
outcomes were composed of the Ocular Surface Disease The rapid development of femtosecond lasers over the past two
Index (OSDI) scores, tear film break-up time (TBUT), decades has opened up new applications in ophthalmic surgery.
Schirmer Test and corneal sensitivity. The femtosecond laser has been widely applied in LASIK,
● RESULTS: Eight eligible studies including 772 eyes (386 in which offers the creation of more predictable flaps with regard
SMILE group and 386 in FS-LASIK group) were identified. The to flap thickness, diam­eter, and hinge width[3]. Although this
parameters have no significiant difference heterogeneity technique has improved a lot nowadays, many studies have
between SMILE and FS-LASIK group preoperatively. There found that some side effects could not be avoided, such as
were significant differences between the two groups in astigmatism, aningeresting, changes of corneal hysteresis
OSDI scores at one and three months postoperatively, in (CH), corneal resistance factor (CRF), photaesthesia or total
TBUT at one and three months postoperatively, in corneal higher order aberration (tHOAs), due to the damaged corneal
sensitivity at one week, about one month and three surface[4-6].
months postoperatively. However, there was no significant SMILE has been considered as advanced technology to correct
difference observed in Schirmer Test at the follow-up myopia and myopic astigmatism. Contrary to LASIK, the flap
periods. is needless in the process of SMILE with the replacement of
● CONCLUSION: Compare to FS-LASIK, dry eye and the the shorter side cut, which is used for lenticule extraction and
corneal sensitivity recover better in the SMILE group, in about 3 to 4 mm[7]. This all-in-one surgery no longer needs
first three months after the surgery. a full flap cut. Agca et al[8] found that SMILE is better in the
632
 Int J Ophthalmol, Vol. 10, No. 4, Apr.18, 2017 www.ijo.cn
Tel:8629-82245172 8629-82210956 Email:[email protected]
nerve fiber regeneration. Some studies have reported that
patients applied with SMILE surgery have low incidence of
dry eye syndrome and central corneal sensation reduction,
although others don’t stay the same[9].
In order to evaluate the ocular surface function after refractive
surgeries, many clinical studies have been carried out.
However, SMILE, an advanced technique carried out in recent
years, whether is superior to FS-LASIK in ocular surface
function, remains unclear. And the comparison of these two
kinds of techniques in dry eye and corneal sensation is not
available. The objective of this Meta-analysis was to analyze
the incidence of dry eye and changes in corneal sensitivity
after these two kinds of refractive surgeries.
MATERIALS AND METHODS
The Meta-analysis was performed following the methods
which has been universally accepted[10].
Search Strategy We searched on PubMed with language
restriction from inception to March 2016. The key words we
used for searching were “SMILE”, “LASIK”, “femtosecond”,
“SMILE and LASIK”, “dry eye”, “corneal sensation”, “corneal Figure 1 Flow diagram of the literature search in this Meta-analysis.
sensitivity” and “ocular surface”. The potentially eligible articles
Data Extraction We extracted the following information:
were identified according to the reference lists after our filtering.
name, publication year, trial location, study design, follow-up
Inclusion and Exclusion Criteria We formulated inclusion
period, intervention, characteristics of population, preoperative
and exclusion criteria to screen the retrieved articles. The
spherical equivalent. A second reviewer double-checked all data.
four inclusion criteria were shown below. At first, either
Statistical Analysis RevMan5.3 software was used to perform
randomized or nonrandomized trials were included in our
this Meta-analysis. The follow-up intervals included pre-
analysis. Secondly, we chose the myopic patients with or
without astigmatism, and the myopic degree ranged from -2 to operation, one week, one month, three months and six months
-10 degrees. Thirdly, the studies with the comparison between post-operation. Heterogeneity was assessed by calculat­ing the
SMILE and FS-LASIK were accepted. In addition, the main I2 statistic and by performing a Chi-square test (assess­ing the
outcomes we observed included Ocular Surface Disease Index P-value). I2>50% is considered to be indicative of significant
(OSDI) scores, tear film break-up time (TBUT), Schirmer Test heterogeneity. Both random-effects and fixed-effects models
and corneal sensitivity. We extracted the datum before surgery could be applied in our analysis, which was dependent on the
and within the six month follow-up. The exclusion criteria heterogeneity between two comparable groups. We examined
included reviews, meetings, letters, studies without complete the P-value and the degree of overlap in confidence intervals
data, with inconsistent or erroneous data and some duplicate (CIs)[14]. The P-value less than 0.05 suggested the statistically
publications. significance. And we also made the forest plot to show the
Screening Process Two independent reviewers (Cai WT and comparison clearly.
Wei QQ) search on the database and filter the studies according RESULTS
to our predefined criteria. We reach the agreements after our Literature Search Figure 1 is the flowchart showing our
discussion. We exclude the studies which report the same literature retrieval process. After our detailed searching on the
clinical trials and select the one latest. database according to the key terms, totally 153 reports were
Quality Assessment We used the validated Downs and Black retrieved. We identified 67 eligible studies on basis of titles and
scale to conduct our quality assessment because the studies abstracts and got the full text of articles. We detailedly browsed
included both randomized and nonrandomized ones[11]. The the full text of articles and finally, 59 reports were excluded.
evaluation indexes are shown as followed, such as reporting, The comparisons in ten studies were not SMILE versus FS-
external validity, internal validity (bias and confounding), and LASIK so they were ruled out. Thirty studies’ outcomes
power[12]. Two authors (Ren CD and Liu QY) independently didn’t conclude dry eye or corneal sensitivity, two didn’t have
assessed the studies and reached the consensus after complete datum, one was case report, ten were review, three
negotiation. We considered the studies with the score more were not prospective studies and three were animals’ study. As
than sixteen had sufficient quality[13]. a result, eight eligible articles were included.
633
Dry eye and corneal sensitivity after SMILE and FS-LASIK: a Meta-analysis
Table 1 Characteristics of included studied
Follow-up Eyes Mean age Preoperative SE
First author Year Location Design Intervention Quality
(mo) (n) (a) (D)
[18]
Li M 2013 China NR 6 SMILE 38 28.21±7.04 -6.68±1.34 20
FS-LASIK 33 27.33±6.58 -7.96±2.61
[7]
Demirok A 2013 Turkey R 6 SMILE 28 26.2±4.4 -3.9±1.5 26
FS-LASIK 28 26.2±4.4 -4.0±1.4
Xu Y[16] 2014 China NR 6 SMILE 81 24.1±6.03 -5.7±1.71 20
FS-LASIK 97 23.96±5.14 -5.8±2.01
[15]
Denoyer A 2014 France NR 6 SMILE 30 31.1±4.7 -4.65±2.38 21
FS-LASIK 30 32.2±7.5 -4.42±1.78
Wang B[17] 2015 China NR 12 SMILE 47 25.21±6.51 -7.46±1.11 20
FS-LASIK 43 24.72±6.53 -7.44±1.13
Xia L[19] 2016 China NR 6 SMILE 69 25.15±4.42 -5.04±2.32 21
FS-LASIK 59 23.65±3.87 -5.13±1.36
[21]
Li M 2013 China NR 6 SMILE 32 27.1±4.0 -6.56±1.28 20
FS-LASIK 42 28.3±5.5 -8.46±2.15
Wei S[20] 2013 China NR 3 SMILE 61 27.44±6.52 -5.11±1.25 20
FS-LASIK 54 25.44±7.15 -5.50±1.54
SMILE: Small incision lenticule extraction; FS-LASIK: Femtosecond laser-assisted in situ keratomileusis; SE: Spherical equivalent; D: Diopter;
NR: Non-randomize; Quality (reporting, external validity, bias, confounding, power).

Characteristics of Included Studies Table 1 showed characteristics lead to the symptoms and signs of dry eye. At first, there was
of our included studies. We identified 772 eyes totally, and of no difference in OSDI score between two groups before oper-
these, SMILE was applied to 386 eyes and FS-LASIK was ation (WMD: 0.36, 95%CI: -0.98 to 1.70, P=0.60). The OSDI
applied to others. The average age ranged from 23.96 to 32.2 score in SMILE group was significantly less than FS-LASIK
years old. Seven studies in our analysis were nonrandomized group at one month (WMD: -5.48, 95%CI: -6.71 to -4.25,
observational studies and one was randomized. The follow- P<0.00001) and three months (WMD: -5.67, 95%CI: -6.77 to
up periods were more than three month and less than one year. -4.57, P<0.00001) after surgery, the score at 1wk and 6mo was
In addition, these eight studies were carried out in different not calculated (Figure 2).
countries. Six studies were applied with Chinese, one with Tear film break-up time There were six studies reporting
French and the other with Turks[7,15-21]. TBUT at different time points. No significant difference
Quality Assessment The quality assessment was showed in was observed (WMD: -0.12, 95%CI: -0.48 to 0.23, P=0.50)
Table 1. In the reporting section, all of these trials received between two groups preoperatively. The TBUT was longer in
scores of more than eight points, which meant that they all SMILE group at one month (WMD: 1.23, 95%CI: 0.05 to 2.41,
clearly described the details of the trials. For the external P<0.05) and three months (WMD: 0.66, 95%CI: 0.36 to 0.96,
validity section, all studies received full scores, which meant P<0.0001) postoperatively, while at one week and six months,
they had a good representativeness of the researchers, subjects the difference was not statistically significant (Figure 3).
and devices. The bias section could address biases in the Schirmer test There were five studies reporting the Schirmer
measurement of the intervention and the outcome. Just one Test results at different time points. Schirmer Test is one of
study is randomized controlled trial (RCT) so others were the indexs to diagnose dry eye. This index decreased after two
relatively low. The confounding section addressed bias in the kinds of refractive surgeries. And no significant difference of
selection of study subjects. We considered the quality index Schirmer Test was observed in two groups preoperation, at one
more than sixteen as adequate quality so all of our eight studies week, approximately one month, three months and six months
met sufficient quality. We used the proper statistical methods postoperatively (P>0.05) (Figure 4).
and calculated the results exactly, which reached the adequate Corneal sensitivity Totally five studies reported corneal
score in the power section. sensitivity. Both SMILE and FS-LASIK refractive surgery
Comparison of Outcome damaged the corneal nerve fiber, and corneal sensation
Ocular Surface Disease Index scores Through the Meta- was very important to evaluate the corneal function. We
analysis, we found the increase of OSDI score between SMILE could not observe the obvious difference was observed
and FS-LASIK, which means that the refractive surgery may preoperatively (P=0.20). We found the superiority of SMILE
634
 Int J Ophthalmol, Vol. 10, No. 4, Apr.18, 2017 www.ijo.cn
Tel:8629-82245172 8629-82210956 Email:[email protected]

Figure 2 Forest plot comparison of OSDI scores after treatment with SMILE and FS-LASIK A: One month postoperatively; B: Three
months postoperatively.

Figure 3 Forest plot comparison of TBUT between SMILE and FS-LASIK A: One month postoperatively; B: Three months postoperatively.

in corneal sensation at 1wk (WMD: 18.96, 95%CI: 7.00 to flap and excimer laser was used to ablat stroma[23]. The LASIK
30.91, P=0.002), 1mo (WMD: 18.48, 95%CI: 9.59 to 27.36, with the mechanical microkeratome for creating a corneal flap
P<0.0001) and 3mo (WMD: 15.39, 95%CI: 9.43 to 21.35, has been demonstrated inferior to femtosecond laser including
P<0.00001) postoperatively, while at 6mo, the two groups was FS-LASIK and SMILE. What’s more, the incidence of dry
not statistically significant (Figure 5). eye may occur less in SMILE because of the small incision
DISCUSSION on the anterior stroma, which can decrease the damage of tear
The fields of refractive surgery develop rapidly in recent years, film and the corneal nerves. FS-LASIK has been mature in
in order to meet different people’ requirements, mainly for the refractive sector. We now use femto-second lasers to make
myopia. Small incision lenticule extraction, a novel procedure, corneal flaps, which can grasp the thickness of flaps accurately
has been prevalent in clinical treatments. Reinstein et al[22] and avoid some complications[24].
retrospect the history, fundamentals and clinical outcomes of Dry eye is a syndrome of blurry vision and burning pain of the
SMILE, and find that patients who take the SMILE may have eyes, which is related with a loss, or reduction of the 3-layered
similar corrected distance visual acuity (CDVA) with FS- tear film [25]. Many patients may feel uncomfortable and
LASIK, which at least can reach the desired visual outcome dissatisfied after refractive operations due to the occurrence
for patients. Femtosecond laser was applied to create corneal of dry eye syndrome. Several studies have reported that the
635
Dry eye and corneal sensitivity after SMILE and FS-LASIK: a Meta-analysis

Figure 4 Forest plot comparison of Schirmer Test after treatment with SMILE and FS-LASIK A: One month postoperatively; B: Three
months postoperatively.

Figure 5 Forest plot comparison of corneal sensitivity after treatment with SMILE and FS-LASIK A: One week postoperatively; B: One
month postoperatively; C: Three months postoperatively.

incidences and degrees of dry are lower following SMILE differences are found in occurrence of dry eye in these two
rather than FS-LASIK[26]. One study has proved that changes in surgeries.
ocular surface are severe in FS-LASIK due to the involvement In this Meta-analysis, we conclude eight appropriate studies,
of some inflammatory factors such as IL-6 and nerve growth extract related datum, and make analysis. We choose three
factor (NGF)[27]. However, some studies find that no obvious parameters closely related with the diagnosis of dry eye,
636
 Int J Ophthalmol, Vol. 10, No. 4, Apr.18, 2017 www.ijo.cn
Tel:8629-82245172 8629-82210956 Email:[email protected]
respectively concluding OSDI scores, TBUT, Schirmer Test. ACKNOWLEDGEMENTS
Before the refractive surgery, the heterogeneity of these three Foundation: Supported by the National Natural Science
parameters in both groups has no obvious difference, which Foundation of China (No.81470648).
can approve the accuracy of the results. We also found that Conflicts of Interest: Cai WT, None; Liu QY, None; Ren
the OSDI score of SMILE patients was lower than that of FS- CD, None; Wei QQ, None; Liu JL, None; Wang QY, None;
LASIK patients at one and three months’ postoperation (both Du YR, None; He MM, None; Yu J, None.
REFERENCES
P<0.00001). What’s more, the TBUT was higher of SMILE
1 Ambrosio R Jr, Tervo T, Wilson SE. LASIK-associated dry eye
patients at 1 and 3mo postoperatively (P<0.05). However,
and neurotrophic epitheliopathy: pathophysiology and strategies for
the Schirmer Test has no significant difference between two
prevention and treatment. J Refract Surg 2008;24(4):396-407.
surgeries. According to the analysis above, we can find that
2 Lee JK, Chuck RS, Park CY. Femtosecond laser refractive surgery:
compared with FS-LASIK, SMILE may have more advantages
small-incision lenticule extraction vs. femtosecond laser-assisted LASIK.
in decreasing the occurrence of dryness, photophobia and
Curr Opin Ophthalmol 2015;26(4):260-264.
tearing. And the recovery of the tear film function occurs at
3 Kim P, Sutton GL, Rootman DS. Applications of the femtosec­ond laser
3mo, so we can consider that 3-months’ post-operation can get
in corneal refractive surgery. Curr Opin Ophthalmol 2011;22(4):238-244.
the reliable results of dry eye.
4 Li X, Wang Y, Dou R. Aberration compensation between anterior and
As the results above, the damages of corneal nerve fiber are posterior corneal surfaces after Small incision lenticule extraction and
less with SMILE patents in the first three months. Although Femtosecond laser-assisted laser in-situ keratomileusis. Ophthalmic
at six months, they don’t have the significant difference, the Physiol Opt 2015;35(5):540-551.
SMILE is still much better. And they are gradually recovery to 5 Agca A, Ozgurhan EB, Demirok A, Bozkurt E, Celik U, Ozkaya A,
the previous level. An animal study keeps the same viewpoint Cankaya I, Yilmaz OF. Comparison of corneal hysteresis and corneal
with our analytical results, which demonstrates that SMILE resistance factor after small incision lenticule extraction and femtosecond
is better in nerve fiber recovery, such as the sub-basal nerve laser-assisted LASIK: a prospective fellow eye study. Cont Lens Anterior
length and density (SLD). After 4-weeks’ observation, Eye 2014;37(2):77-80.
the stromal nerve in rabbits which underwent the SMILE 6 Chan TC, Ng AL, Cheng GP, Wang Z, Ye C, Woo VC, Tham CC, Jhanji
[28]
recovered faster than the ones in FS-LASIK group . V. Vector analysis of astigmatic correction after small-incision lenticule
This Meta-analysis reliable to assess ocular surface complicates extraction and femtosecond-assisted LASIK for low to moderate myopic
of these surgeries. Firstly, we search for the studies cautiously astigmatism. Br J Ophthalmol 2016;100(4):553-559.

and finish the searching by two authors. Secondly, we filter 7 Demirok A, Ozgurhan EB, Agca A, Kara N, Bozkurt E, Cankaya KI,

the eligible studies in our analysis, and all of them have been Yilmaz OF. Corneal sensation after corneal refractive surgery with small

satisfied with the quality assessment. Thirdly, we choose incision lenticule extraction. Optom Vis Sci 2013;90(10):1040-1047.
8 Agca A, Cankaya KI, Yilmaz I, Yildirim Y, Yasa D, Olcucu O,
the fixed or random model according to the heterogeneity.
Demircan A, Demirok A, Yilmaz OF. Fellow eye comparison of nerve
Fourthly, the studies are applied in different countries, which
fiber regeneration after SMILE and femtosecond laser-assisted LASIK: a
increase the reliability. There are also some limitations in this
confocal microscopy study. J Refract Surg 2015;31(9):594-598.
Meta-analysis. Firstly, most studies are non-randomized trials.
9 He M, Huang W, Zhong X. Central corneal sensitivity after small
So the risk of various bias will increase and reliability may
incision lenticule extraction versus femtosecond laser-assisted LASIK
decrease. Secondly, some indexes are not complete, so the
for myopia: a meta-analysis of comparative studies. BMC Ophthalmol
numbers are not enough. Thirdly, the numbers of the included
2015;15:141.
eyes are not enough. Finally, the outcomes we extract don’t
10 Egger M, Smith GD, Phillips AN. Meta-analysis: principles and
contain the visual results because of the incomplete datum.
procedures. BMJ 1997;315(7121):1533-1537.
Thus more clinical trials need to be carried out.
11 Ye J, Xu YF, He JJ, Lou LX. Association between aspirin use and age-
In conclusion, this Meta-analysis suggests that the dry eye related macular degeneration: a meta-analysis. Invest Ophthalmol Vis Sci
symptom occurred and the corneal sensitivity decreased 2014;55(4):2687-2696.
between SMILE and FS-LASIK, and latter had higher 12 Samoocha D, Bruinvels DJ, Elbers NA, Anema JR, van der Beek AJ.
occurrence rate. The incidences of dry eye recovery better Effectiveness of web-based interventions on patient empowerment: a
in the SMILE groups, mainly within the three months systematic review and meta-analysis. J Med Internet Res 2010;12(2):e23.
postoperatively. The corneal sensitivity also recoveries better 13 Wang W, Zhou M, Huang W, Zhang X. Ex-PRESS implantation versus
in SMILE patients in the first three months after surgeries. The trabeculectomy in uncontrolled glaucoma: a meta-analysis. PLoS One
comparisons at 6mo have no significant difference. 2013;8(5):e63591.

637
Dry eye and corneal sensitivity after SMILE and FS-LASIK: a Meta-analysis
14 Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring lenticule extraction (SMILE) and femtosecond laser in situ keratomileusis
inconsistency in meta-analyses. BMJ 2003;327(7414):557-560. (FS-LASIK). PLoS One 2013;8(12):e81435.
15 Denoyer A, Landman E, Trinh L, Faure JF, Auclin F, Baudouin 22 Reinstein DZ, Archer TJ, Gobbe M. Small incision lenticule extraction
C. Dry eye disease after refractive surgery: comparative outcomes (SMILE) history, fundamentals of a new refractive surgery technique and
of small incision lenticule extraction versus LASIK. Ophthalmology clinical outcomes. Eye Vis (Lond) 2014;1:3.
2015;122(4):669-676. 23 Shortt AJ, Allan BD. Photorefractive keratectomy (PRK) versus laser-
16 Xu Y, Yang Y. Dry eye after small incision lenticule extraction and assisted in-situ keratomileusis (LASIK) for myopia. Cochrane Database
LASIK for myopia. J Refract Surg 2014;30(3):186-190. Syst Rev 2006;(2):CD005135.
17 Wang B, Naidu RK, Chu R, Dai J, Qu X, Zhou H. Dry eye disease 24 Mysore N, Krueger R. Advances in refractive surgery: May 2013 to
following refractive surgery: a 12-month follow-up of SMILE versus FS- June 2014. Asia Pac J Ophthalmol (Phila) 2015;4(2):112-120.
LASIK in high myopia. J Ophthalmol 2015;2015:132417. 25 Conrady CD, Joos ZP, Pate BC. Review: the lacrimal gland and its role
18 Li M, Zhao J, Shen Y, Li T, He L, Xu H, Yu Y, Zhou X. Comparison of in dry eye. J Ophthalmol 2016;2016:7542929.
dry eye and corneal sensitivity between small incision lenticule extraction 26 Ganesh S, Gupta R. Comparison of visual and refractive outcomes
and femtosecond LASIK for myopia. PLoS One 2013;8(10):e77797. following femtosecond laser- assisted lasik with smile in patients with
19 Xia L, Zhang J, Wu J, Yu K. Comparison of corneal biological myopia or myopic astigmatism. J Refract Surg 2014;30(9):590-596.
healing after femtosecond LASIK and small incision lenticule extraction 27 Gao S, Li S, Liu L, Wang Y, Ding H, Li L, Zhong X. Early changes
procedure. Curr Eye Res 2016;41(9):1202-1208. in ocular surface and tear inflammatory mediators after small-incision
20 Wei S, Wang Y. Comparison of corneal sensitivity between FS-LASIK lenticule extraction and femtosecond laser-assisted laser in situ
and femtosecond lenticule extraction (ReLEx flex) or small-incision keratomileusis. PLoS One 2014;9(9):e107370.
lenticule extraction (ReLEx smile) for myopic eyes. Graefes Arch Clin 28 Mohamed-Noriega K, Riau AK, Lwin NC, Chaurasia SS, Tan DT,
Exp Ophthalmol 2013;251(6):1645-1654. Mehta JS. Early corneal nerve damage and recovery following small
21 Li M, Niu L, Qin B, Zhou Z, Ni K, Le Q, Xiang J, Wei A, Ma W, Zhou incision lenticule extraction (SMILE) and laser in situ keratomileusis
X. Confocal comparison of corneal reinnervation after small incision (LASIK). Invest Ophthalmol Vis Sci 2014;55(3):1823-1834.

638