Pii: s0886-3350(03)00575-3

Effect of amniotic membrane after laser-assisted
subepithelial keratectomy on epithelial healing
Clinical and refractive outcomes
Hyung Keun Lee, MD, Jin Kook Kim, MD, Sung Soo Kim, MD, Eung Kweon Kim, MD,Kwang One Kim, MD, In Sik Lee, MD, Gong Je Seong, MD Purpose: To evaluate the effect of an amniotic membrane (AM) on reepithelialization
time, corneal haze, and postoperative visual and refractive outcomes after laser-
assisted subepithelial keratectomy (LASEK) for myopia and myopic astigmatism.
Setting: Department of Ophthalmology, Yonsei University College of Medicine,
and Balgeunsesang Ophthalmology Clinic, Seoul, Korea.
Methods: One hundred fifty-two eyes of 84 patients with myopia or myopic astig-
matism were prospectively evaluated for 6 months after LASEK. An AM was
placed as a strip on the inferior limbus in 94 eyes of 54 patients after LASEK;
58 eyes of 30 patients served as the control group. Postoperative epithelial heal-
ing time, uncorrected visual acuity (UCVA), best corrected visual acuity, remaining
refractive error, and corneal haze were examined.
Results: The reepithelialization time was shorter in the AM group (2.40 days Ϯ
0.94 [SD]) than in the control group (3.90 Ϯ 0.97 days) (PϽ.001). At 6 months, 86
eyes (91.5%) in the AM group had a UCVA of 20/25 or better and 90 eyes
(95.7%) had a UCVA of 20/40 or better; 48 eyes (82.8%) and 53 eyes (91.4%) in
the control group had a UCVA of 20/25 or better and 20/40 or better, respec-
tively. The mean spherical equivalent in the AM group was Ϫ0.48 Ϯ 0.54 diopter
(D) and in the control group, Ϫ0.94 Ϯ 0.60 D (PϽ.001). The corneal haze was sig-
nificantly less in the AM group than in the control group (PϽ.001).
Conclusion: Amniotic membrane use after LASEK induced rapid epithelial heal-
ing with more favorable visual and refractive outcomes and lower corneal haze
scores than conventional LASEK.
J Cataract Refract Surg 2004; 30:334–340  2004 ASCRS and ESCRS Laser-assisted subepithelial keratectomy (LASEK) ocular pain after surgery or stromal opacity similar to was introduced and popularized with the advan- tages of decreased pain, tearing, irritating symptoms, Amniotic membrane (AM), the innermost mem- and corneal opacity compared to photorefractive kera- brane lining the placenta facing the fetus, is known to tectomy (PRK). It also enables the correction of refrac- induce rapid corneal epithelial healing and is used to tive errors, especially in eyes with thin corneas.1,2 There reconstruct the ocular surface in cases of partial limbal- is, however, some controversy about the possibility of cell deficiency and persistent epithelial defects.4–7 Recent delayed epithelial healing resulting in complaints of studies show that AM transplantation results in lessstromal infiltration of inflammatory cells and a reduced Accepted for publication June 17, 2003. loss of keratocytes in rabbit corneas.8,9 These results Reprint requests to Eung Kweon Kim, MD, Institute of Vision Research, suggest that AM application to the cornea after LASEK Department of Ophthalmology, College of Medicine, Yonsei University, would affect the epithelial healing pattern and stro- 134 Shinchon-dong, Sudaemoon-gu, Seoul, Korea. E-mail: eungkkim@yumc.yonsei.ac.kr. This study prospectively evaluated the effect of AM placement after LASEK on epithelial healing time, post-operative visual outcomes including refractive status andvisual acuity, and corneal opacities.
Patients and Methods
One hundred fifty-two consecutive eyes of 84 patients were enrolled between September 2000 and December 2001.
The preoperative ophthalmic examination of all patients in-cluded slitlamp biomicroscopy, intraocular pressure, fundusexamination, pupil diameter measurements, Schirmer test,manifest refraction, corneal keratometry, corneal topography,corneal pachymetry, and visual field examination. No patienthad a history of refractive procedures or cataract surgery, Figure 1.
(Lee) One day after LASEK with AM strip transplantation.
keratoconus, diabetes, glaucoma, connective tissue disorders, Arrowheads mark the upper margin of the AM.
After the preoperative examinations, the patients were proparacaine hydrochloride 0.5% (Alcaine௡) instilled. An told about the AM and its intended use after LASEK. They alcohol solution cone (J2905, Janach) with an 8.5 mm diame- were also told that this might be the first human trial after ter was placed on the eye. Twenty percent of the alcohol photoablation, and written informed consent was obtained solution was instilled inside the cone, left for about 20 sec- once the patients agreed to participate. The patients who did onds, and then carefully washed off with a balanced salt not agree to the use of the AM on the LASEK-treated eye solution so the epithelium around the flap was not disturbed.
were considered the control group after they provided in- The epithelial flap was gently lifted with an epithelial microhoe (J2915A, Janach). If the edges were difficult to lift,the alcohol application was repeated for another 10 seconds.
Preparation of Preserved Human Amniotic Membrane The epithelial flap was then peeled back as a sheet toward the The AM was prepared following the methods of Lee 12 o’clock position using a spatula (J2910A, Janach). To avoid and Tseng.10 Human placenta was obtained during an elective tears in the epithelial flap, the basal lamina was carefully Cesarean section in a seronegative (human immunodeficiency separated from Bowman’s layer instead of separating the virus, human hepatitis type B and C, and syphilis) woman.
epithelium from the basal lamina. If the epithelial flap was Informed consent was obtained from the AM donor after the not applied to the denuded stromal as an intact sheet, the purpose of the AM was explained. Under a laminar flow hood, the placenta was cleaned of blood clots with sterile phosphate- Excimer laser treatment was performed in the usual man- buffered saline solution containing penicillin 50 ␮g/␮L, ner using the nomogram for PRK with the EC-5000 laser streptomycin 50 ␮g/␮L, neomycin 100 ␮g/␮L, and ampho- system (Nidek). The flap was washed with a balanced salt tericin B 2.5 ␮g/␮L. The amnion was separated from the solution and then repositioned carefully with a spatula. In chorion by blunt dissection and flattened on a nitrocellulose the control group, the treatment was finished with the appli- paper, with the epithelium-basement-membrane surface fac- cation of a therapeutic soft contact lens on the operated eye.
ing away from the paper. The paper with adherent AM was In the AM group, the AM was washed thoroughly and then cut into 1.5 cm blocks and stored at Ϫ80ЊC until immersed in a balanced salt solution containing gentamicin transplantation in a sterile vial containing Dulbecco’s modi- 8 ␮g/␮L and cefaxolin 4 ␮g/␮L for 30 minutes before sur- fied Eagle medium (GIBCO Life Technologies, Inc.) and gery. The membrane was then cut into 1.5 cm ϫ 0.3 cm glycerol (GIBCO Life Technologies, Inc.) at the ratio of 1:1 rectangular pieces. With the mesenchymal side facing the (vol/vol). Six months after delivery, the AM donors were cornea, the membrane was attached to the stromal bed after retested by the previous serologic examinations for the win- excimer laser ablation. The slender AM strip was placed on dow period of transmittable diseases. Only the AM that the inferior limbus so it would not touch the ablated corneal passed prenatal and postnatal serologic tests as seronegative bed and secured with 2 interrupted 10-0 nylon sutures placed on the limbal conjunctiva. A therapeutic soft contact lenswas also placed on the eyes with the AM strip (Figure 1).
Laser-Assisted Subepithelial Keratectomy Procedure One drop of ofloxacin 0.3% (Tarivid௡) and diclofenac The LASEK procedure was the same in the control and 0.1% (Optanac௡) was given to the patients in both groups AM groups. A speculum was applied to the patient’s eye and immediately after LASEK. All patients were checked daily J CATARACT REFRACT SURG—VOL 30, FEBRUARY 2004 illumination; 1, for an opacity of minimal density seen withdifficulty under direct and diffuse illumination; 2, for an easily visible opacity; 3, for a dense opacity that significantly Characteristics
P Value
decreased visualization of intraocular structures such as the iris and retina; and 4, for an opaque cornea.
Statistical analysis was performed using Statistical Analy- sis System (version 6.12, SAS Institute Inc.). A P value lessthan 0.05 was considered statistically significant.
AM ϭ amniotic membrane; NS ϭ not significant The AM group and the control group consisted of 94 eyes of 54 patients and 58 eyes of 30 patients,respectively. Preoperative data are shown in Table 1.
There were no statistically significant differences be-tween the 2 groups in the preoperative independentvariables.
Eyes with an AM had a significantly faster rate of epithelial healing than the control eyes (PϽ.001).
Round or oval epithelial defects under the epithelialflap and therapeutic soft lens were observed on the dayof treatment; the defects decreased concentrically ateach follow-up. The mean epithelial healing time was2.40 days Ϯ 0.94 (SD) in the AM group and 3.90 Ϯ0.97 days in the control group. Three days after surgery,82 eyes (87.2%) in the AM group and 18 eyes (31.8%)in the control group were completely reepithelialized.
Ten eyes (10.6%) in the AM group showed complete Figure 2.
(Lee) Completion of epithelial healing in the control and reepithelialization within 36 hours of surgery, and 2 eyes (3.45%) in the control group showed completereepithelialization after 5 days (Figure 2).
until the epithelial defect completely healed. They were in- At 1 week, 61 eyes (64.9%) in the AM group and structed to apply 1 drop of diclofenac and ofloxacin every 31 eyes (53.5%) in the control group had a UCVA 2 hours and artificial tears (Hyalein 0.1%) every hour until of 20/25 or better. At 6 months, 48 eyes (82.8%) Complete epithelialization was determined by daily slit- in the control group had a UCVA of 20/25 and 53 lamp observation. Once the epithelium was healed, the thera- eyes (91.4%), of 20/40 or better; in the AM group, peutic contact lenses and AM, if used, were removed from 86 eyes (91.5%) had a UCVA of 20/25 or better and the cornea. Then, ofloxacin 0.3% and fluorometholone 0.1% 90 eyes (95.7%), of 20/40 or better (Table 2).
(Fluorometholone௡) were administered 4 times daily for There were no statistically significant differences in 1 week and 2 times daily for 1 month.
BCVA between the 2 groups during the follow-up.
Uncorrected visual acuity (UCVA), best corrected visual acuity (BCVA), manifest refraction, tonometry, and slitlamp Most patients in both groups showed no changes in biomicroscopy were performed at each follow-up examina- BCVA. However, 7 eyes (7.5%) in the AM group and tion. Subepithelial corneal haze levels were checked with a 5 eyes (8.6%) in the control group lost 1 or 2 Snellen slitlamp 1, 3, and 6 months after surgery. Two of the authors, lines of BCVA at the final examination.
who did not know whether the eye examined was part of Statistically significant differences in the mean refrac- the AM group or the control group, observed and graded tive error were found between 1 month and 6 months the corneal opacities from 0 to 4, according to the methodof Hanna et al.11 A grade of 0 was given for “totally clear”; (Table 3). At the final examination, 70 eyes (74.5%) 0.5, for a faint corneal opacity seen only by oblique indirect in the AM group and 40 eyes (69.0%) in the control J CATARACT REFRACT SURG—VOL 30, FEBRUARY 2004 Visual acuity results in the control and AM groups over time after LASEK.
Month(s) After LASEK
AM ϭ amniotic membrane; LASEK ϭ laser-assisted subepithelial keratectomy*pϽ.01, chi-square test group had a mean spherical equivalent (SE) within in the control group had a corneal haze score greater Ϯ0.50 diopter of the attempted myopic correction, than 2 (PϽ.01). At 6 months, the AM group had a more with a significant between-group difference (PϽ.01).
favorable corneal haze score. Forty-five eyes (47.8%) in The cylinder magnitude between the groups was also the AM group and 21 eyes (35.2%) in the control group significantly different from 1 month to 6 months had a haze score below grade 1. Three eyes (5.2%) in the control group and 4 eyes (4.3%) in the AM group Corneal haze was examined and graded under the had grade 3 corneal opacity 6 months after surgery slitlamp according to the previously described grading system. At 1 month, the corneal haze score was less The correlations between duration of the epithelial than grade 1 in all eyes. But the AM group had less defect, corneal opacity, SE, and cylinder magnitude in haze than the control group (PϽ.01). At 3 months, all eyes are shown in Table 4. The strongest correlation 8 eyes (8.5%) in the AM group and 20 eyes (34.7%) was between corneal opacity and SE at 6 months (corre-lation coefficient ϭ Ϫ0.711, P ϭ .000). The correlation Refractive error in the control and AM groups after coefficient showed a positive correlation between the duration of the epithelial defect and the stromal opacity Mean D Ϯ SD
(r ϭ 0.653, P ϭ .000) and negative correlations between Control Group
P Value*
the stromal opacity and the SE (r ϭ Ϫ0.607, P ϭ Discussion
From our results, we conclude that the epithelializa- tion and the wound-healing process after LASEK was influenced by the AM strip. Both visual and refractive outcomes were acceptable in all patients. However, the AM group had a shorter epithelial healing time and more favorable visual and refractive outcomes. At the beginning of the study, we used the AM to cover the entire LASEK-treated surface directly. However, inmany eyes, the regenerated corneal epithelium was de- tached during removal of the AM so a longer period was needed for reepithelialization. We therefore modi- AM ϭ amniotic membrane; Cyl ϭ cylinder; LASEK ϭ laser-assisted fied the so-called overlay technique to use the AM as subepithelial keratectomy; SE ϭ spherical equivalent*pϽ.05, Student t test J CATARACT REFRACT SURG—VOL 30, FEBRUARY 2004 Corneal haze score in the control and AM groups after LASEK.
Month(s) After LASEK
AM ϭ amniotic membrane; LASEK ϭ laser-assisted subepithelial keratectomy*pϽ.01, chi-square test The exact mechanism of how AM use after LASEK first is that the AM strip may act as a mechanical barrier is effective in corneal epithelial healing was not demon- against white-blood-cell migration that emerges from strated in this study. The AM was not directly in contact the limbal blood vessels or tear fluid in the conjunctival with the wound bed, so the direct physical and mechani- sac. After excimer laser refractive surgery, the concentra- cal roles of the AM would hardly affect the wound tion of inflammatory cells and cytokines such as trans- healing after LASEK. A bandage soft contact lens forming growth factor-␤ is increased immediately.12,13 (BSCL) was used in both groups until the epithelial The greatest amount of tear flow occurs in the meniscus healing was complete. Both groups were studied at the adjacent to the lower lid.14 The AM, which was placed same time by the same surgeon and observers. We on the inferior limbus, might act as a mechanical barrier therefore think the BSCL had the same effect on both so it would decrease inflammatory cell infiltration in groups and did not act as a bias in 1 group.
the wound bed and reduce epithelial cell destruction We suggest 2 possible mechanisms to explain how by the white blood cells and the inflammatory factors the AM strip facilitates corneal epithelial healing. The The second possibility is that the various cytokines Correlation coefficients between duration of epithelial de- and epithelial-growth-associated factors within the AM fect and corneal opacity, refractive error, and cylinder magnitude at would help epithelial regeneration and wound healing after LASEK. The AM has been shown to express epi- dermal growth factor, hepatocyte growth factor, and keratinocyte growth factor15 and suppress the proinflam- matory cytokines during wound healing, just as interleu- kin does.16 Although the exact role of these factors in AM associated with facilitating cornea wound healing is not known, the factors may play some role in corneal epithelial healing. However, further studies of how the AM promotes epithelial healing are needed.
The UCVA was better in the AM eyes than in the control group from 1 week to 6 months after surgery.
We believe the faster epithelial healing that resulted in a smoother refractive surface in the AM group provided better UCVA a week after LASEK. At 6 months, cornealopacities, which might cause irregular astigmatism and CM ϭ cylinder magnitude; DED ϭ duration of epithelial defect; SD ϭstromal opacity; SE ϭ spherical equivalent; Sig ϭ significance increase postoperative refractive errors, were fewer in J CATARACT REFRACT SURG—VOL 30, FEBRUARY 2004 2. Dastjerdi MH, Soong HK. LASEK (laser subepithelial the AM group. We think these could be related to the keratomileusis). Curr Opin Ophthalmol 2002; 13:261– We did not find statistically significant differences 3. Litwak S, Zadok D, Garcia-de Quevedo V, et al. Laser- in the BCVA between the 2 groups until 6 months assisted subepithelial keratectomy versus photorefractive postoperatively. After PRK or LASEK, the corneal haze keratectomy for the correction of myopia; a prospective is associated with BCVA loss.17 Up to 6 months postop- comparative study. J Cataract Refract Surg 2002; 28:1330–1333 eratively, most patients showed no changes in BCVA 4. Azuara-Blanco A, Pillai CT, Dua HS. Amniotic mem- and corneal haze below grade 2. We do not think this brane transplantation for ocular surface reconstruction.
mild to moderate grade of corneal haze would affect 5. Tseng SCG, Prabhasawat P, Barton K, et al. Amniotic Although the refractive outcomes in the 2 groups membrane transplantation with or without limbal allo- were relatively comparable, there was a statistically sig- grafts for corneal surface reconstruction in patients withlimbal stem cell deficiency. Arch Ophthalmol 1998; nificant difference in the remaining refractive errors.
Besides the spherical component of the postoperative 6. Tsubota K, Satake Y, Ohyama M, et al. Surgical recon- refractive error, the magnitude of astigmatism was statis- struction of the ocular surface in advanced ocular cicatri- tically significantly different between the groups. More- cial pemphigoid and Stevens-Johnson syndrome. Am J over, we found correlations among corneal opacity, 7. Chen H-J, Pires RTF, Tseng SCG. Amniotic membrane epithelial healing time, and remaining refractive errors.
transplantation for severe neurotrophic corneal ulcers.
From these results, we think it is possible that the AM affects postoperative wound healing after LASEK.
8. Wang MX, Gray TB, Park WC, et al. Reduction in The excessive synthesis of collagen and glycosami- corneal haze and apoptosis by amniotic membrane ma- noglycans by keratocytes in the ablation zone can lead to trix in excimer laser photoablation in rabbits. J Cataract a thickening that yields myopic regression, astigmatism, 9. Park WC, Tseng SCG. Modulation of acute inflamma- irregular topography, and optical aberration.18 Tabin tion and keratocyte death by suturing, blood, and amni- et al.19 report that surgically induced refractive errors otic membrane in PRK. Invest Ophthalmol Vis Sci including astigmatism may be the result of irregular epithelial thickening or epithelial hyperplasia. There- 10. Lee S-H, Tseng SCG. Amniotic membrane transplanta- fore, the postoperative wound healing after excimer laser tion for persistent epithelial defects with ulceration. AmJ Ophthalmol 1997; 123:303–312 surgery could significantly affect the postoperative visual 11. Hanna KD, Pouliquen YM, Waring GO III, et al. Cor- and refractive outcomes. It is possible that corneal haze neal wound healing in monkeys after repeated excimer and surgically induced refractive errors could be regu- laser photorefractive keratectomy. Arch Ophthalmol lated and moreover reduced with antiinflammatory 12. Tuominen ISJ, Tervo TMT, Teppo A-M, et al. Human In conclusion, a strip of AM fixed on the limbus tear fluid PDGF-BB, TNF-␣ and TGF-␤1 vs cornealhaze and regeneration of corneal epithelium and subbasal after LASEK may reduce the duration of epithelial de- nerve plexus after PRK. Exp Eye Res 2001; 72:631–641 fects and decrease corneal haze, which improves visual 13. Ramirez-Florez S, Maurice DM. Inflammatory cells, re- outcomes in LASEK eyes. Further studies are needed fractive regression, and haze after excimer laser PRK.
to investigate the subcellular mechanisms of the AM in facilitating epithelial healing. We also believe that a ¯ gu¯t MS, Bavbek T, Kazokoglu H. Assessment of tear longer follow-up of the patients is needed.
drainage by fluorescein dye disappearance test after ex-perimental canalicular obstruction. Acta Ophthalmol1993; 71:69–72 References
15. Koizumi N, Inatomi T, Sotozono C, et al. Growth factor 1. Lee JB, Seong GJ, Lee JH, et al. Comparison of laser mRNA and protein in preserved human amniotic mem- epithelial keratomileusis and photorefractive keratec- brane. Curr Eye Research 2000; 20:173–177 tomy for low to moderate myopia. J Cataract Refract 16. Solomon A, Rosenblatt M, Monroy D, et al. Suppression of interleukin 1␣ and interleukin 1␤ in human limbal J CATARACT REFRACT SURG—VOL 30, FEBRUARY 2004 epithelial cells cultured on the amniotic membrane stro- pic astigmatism; the Melbourne Excimer Laser Group.
mal matrix. Br J Ophthalmol 2001; 85:444–449 J Cataract Refract Surg 1996; 22:924–930 17. Van Gelder RN, Steger-May K, Yang SH, et al. Compar- ison of photorefractive keratectomy, astigmatic PRK, From the Institute of Vision Research, Department of Ophthalmology, laser in situ keratomileusis, and astigmatic LASIK in the College of Medicine (H.K. Lee, S.S. Kim, E.K. Kim, Seong) and treatment of myopia. J Cataract Refract Surg 2002; 28: BK21 Project for Medical Science (E.K. Kim); Yonsei University; and Balgeunsesang Ophthalmology Clinic (J.K. Kim, I.S. Lee), Seoul, Korea. 18. Shah S, Chatterjee A, Smith RJ. Predictability and out- Supported by a grant (02-PJ1-PG1-CH02-0003) from the Korea comes of photoastigmatic keratectomy using the Nidek Health 21 R&D Project, Ministry of Health and Welfare, Republic EC-5000 excimer laser. J Cataract Refract Surg 2002; 19. Tabin GC, Alpins N, Aldred GF, et al. Astigmatic change None of authors has a financial or proprietary interest in any prod- 1 year after excimer laser treatment of myopia and myo- J CATARACT REFRACT SURG—VOL 30, FEBRUARY 2004

Source: http://xn--vb0bx00a3oduny1olld.kr/BnVDataUpload/thesis/2004%20Feb%20%EB%9D%BC%EC%84%B9%20%EC%88%98%EC%88%A0%20%EC%8B%9C%20%EC%96%91%EB%A7%89%EC%9D%98%20%EC%82%AC%EC%9A%A9%EC%9D%B4%20%EA%B0%81%EB%A7%89%20%EC%83%81%ED%94%BC%20%EC%84%B8%ED%8F%AC%EC%9D%98%20%EC%9E%AC%EC%83%9D%EA%B3%BC%20%EC%88%98%EC%88%A0%20%ED%9B%84%20%EA%B2%B0%EA%B3%BC%EC%97%90%20%EB%AF%B8%EC%B9%98%EB%8A%94%20%EC%98%81%ED%96%A5(%EC%9B%90%EB%AC%B8).pdf