In Vivo Confocal Microscopic Evaluation of Corneal Wound Healing After Epi-LASIK

In Vivo Confocal Microscopic Evaluation of Corneal Wound Healing after Epi-LASIK

Wei-Li Chen, Huai-Wen Chang, and Fung-Rong Hu

PURPOSE. To study the healing of corneal wounds after epikera- The reduced depth of treatment lessens the loss of adequate tome laser-assisted in situ keratomileusis (epi-LASIK). corneal tectonic strength. In addition, the lack of incision into METHODS. Twenty-seven patients who had undergone epi- the corneal nerves limits the severity of dry eye. Flap-related LASIK for the treatment of myopia or myopic astigmatism in 46 complications, such as epithelial ingrowth, flap striae, disloca- 5–8 eyes were enrolled. A single intraoperative application of tion, or loss can also be avoided. Epi-LASIK is reported to 0.02% mitomycin C (MMC) for 20 to 30 seconds was used in 24 preserve a viable epithelial sheet after replacement within at eyes with a refraction of no less than Ϫ6.0 D (MMC group). least 24 hours after treatment.9–10 It also provides lower levels MMC was not given to eyes with myopia less than Ϫ6.0 D of transforming growth factor-␤1 in tears compared with (non-MMC group). The eyes were examined by in vivo confo- LASEK and reduces the incidence of haze formation.11 Many cal microscopy at 1, 3, and 7 days after surgery and then patients have reported shortened durations of pain and satis- weekly during the first month and once each at 3 and 6 factory surgical outcomes,12–14 although contradictory results months. Selected images of the corneal basal–apical surface have also been reported.15,16 epithelia and stromal reactions quantified by z-scan profile To date, only a few investigators have studied the histopa- were evaluated. thology of the corneal epithelial flaps created with an epikera- RESULTS. In vivo confocal microscopy showed that cells in most tome, and most specimens were obtained within 24 hours after 9,10,16 of the epithelial flaps were damaged during the first few days surgery. To our knowledge, no researchers have sequen- after surgery and were rapidly replaced by new growing cells. tially observed the wound-healing process (namely, epithelial In the MMC and non-MMC groups, the corneal basal epithelial flaps and stromal reactions) microscopically after epi-LASIK. cells returned to their preoperative morphology in 0% and The purpose of this study was to assess by in vivo confocal 13.6% of the eyes after 1 week, 37.5% and 36.4% after 2 weeks, microscopy the healing of corneal wounds during the 6 and 87.5% and 86.3% after 1 month, respectively. The corneal months after epi-LASIK. apical surface epithelial cells in the MMC and non-MMC groups recovered their squamous morphology in 12.5% and 13.6% of the eyes at 2 weeks, 37.5% and 54.5% at 1 month, and 52.4% and 57.9% at 6 months, respectively. There was no difference METHODS in the stromal reaction between the groups at 1, 3, and 6 months after surgery. Patients CONCLUSIONS. Damage of the epithelial flaps after epi-LASIK was The study procedure was approved by the Institutional Review Board observed by in vivo confocal microscopy. MMC usage may for Human Studies of the National Taiwan University Hospital and cause more damage to the epithelial flaps. There was no dif- adhered to the guidelines in the Declaration of Helsinki for research in ference in stromal reaction between the groups with and human subjects. Twenty-seven patients (mean age Ϯ SD, 29.2 Ϯ 5.4 without MMC. (Invest Ophthalmol Vis Sci. 2008;49:2416–2423) years; age range, 21–47; 46 eyes) who underwent epi-LASIK surgery DOI:10.1167/iovs.07-1085 between March 2005 and May 2006 for the treatment of myopia or myopic astigmatism were enrolled. All patients provided informed lthough laser-assisted in situ keratomileusis (LASIK) is the consent for the surgery and the in vivo confocal microscopic study. Amost popular surgical procedure for treating refractive Inclusion criteria were Ͼ18 years of age, stable refraction, no previous errors, flap-related problems may still lead to undesirable con- 1–4 refractive surgery, no ocular or systemic disease that could affect sequences. Surface-ablation techniques, such as photore- epithelial healing, tear break-up time of no less than 10 seconds, and fractive keratectomy (PRK), laser-assisted subepithelial kera- Schirmer test with anesthesia of no less than 5 mm before surgery. tectomy (LASEK), and epikeratome laser-assisted in situ Slit-lamp biomicroscopy and in vivo confocal microscopy were per- keratomileusis (epi-LASIK) have certain advantages over LASIK. formed before surgery to rule out corneal disease. Two physicians performed all the epi-LASIK procedures with an epikeratome (Centurion Epi Edge; Norwood Abbey, Melbourne, VIC, From the Department of Ophthalmology, National Taiwan Univer- Australia) and an excimer laser (Technolas 217; Bausch & Lomb, sity Hospital, Taipei, Taiwan. Rochester, NY) and attempted to achieve emmetropia. Patients with Presented at the annual meeting of the American Academy of incomplete flaps, free flaps, or poor adhesion of the epithelial flaps Ophthalmology, Las Vegas, Nevada, November 11–14, 2006. Supported by the Department of Medical Research at the National after surgery were excluded from the analysis. Taiwan University Hospital. During surgery, a 0.02% mitomycin C (MMC) solution was applied Submitted for publication August 18, 2007; revised January 6 and for 20 or 30 seconds to eyes with refraction of Ϫ6.0 to Ϫ8.0 D or more February 24, 2008; accepted April 15, 2008. than Ϫ8.0 D, respectively. MMC was not given to eyes with myopia Disclosure: W.-L. Chen, None; H.-W. Chang, None; F.-R. Hu, less than Ϫ6.0 D. Chilled physiologic saline (BSS Plus; Alcon Labora- None tories Inc, Ft. Worth, TX) was used to wash out any retained MMC. The publication costs of this article were defrayed in part by page A therapeutic bandage contact lens was inserted immediately after charge payment. This article must therefore be marked “advertise- ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact. surgery and was removed within 7 days after surgery. Artificial tears Corresponding author: Fung-Rong Hu, Department of Ophthal- and 0.1% fluorometholone were applied 4 times daily for 1 month and mology, National Taiwan University Hospital, 7 Chung-Shan South were gradually tapered over 4 months. Gentamicin (0.3%) was applied Road, Taipei, Taiwan; [email protected]. four times daily for 1 week.

Downloaded from iovs.arvojournals.org on 09/30/2021 IOVS, June 2008, Vol. 49, No. 6 In Vivo Confocal Microscopy and Epi-LASIK 2417

In Vivo Confocal Microscopy In vivo confocal microscopy was performed before surgery and on days 1, 3, 5, and 7 after surgery. Evaluations were repeated weekly in the ﬁrst month and once each at 3 and 6 months. Before the therapeutic bandage contact lenses were removed, the patients were examined with the lenses in place. After the contact lenses were discontin- ued, examinations were performed without contact lenses. Before in vivo confocal microscopic examination, one drop of 0.5% propara- caine solution and artiﬁcial tears was instilled into the lower conjunc- tival sac. The patient was then seated at an examination table with the head in a headrest. The center of the cornea was examined with a confocal microscope (Confoscan 3.4.1; Nidek Technologies, Padova, Italy) equipped with a standard 40ϫ water-immersion front lens. The microscope was set to the automatic mode. A scan of the full thickness of the cornea was automatically performed in each participant; the examinations lasted 1 to 3 minutes. Each scan recorded 350 images at a distance of 4 ␮m, on a z-axis.

Image Analysis of Corneal Basal Epithelial Cells In vivo confocal microscopy was used to classify the corneal basal epithelium as follows: (1) intact cellular border without a visible nucleus (Fig. 1A); (2) elongated shape with intact cellular borders but no visible nucleus (Fig. 1B); (3) amorphous and poorly identified appearance (Fig. 1C); (4) patchy whitening of the cellular sheet in which some areas showed intact borders without visible nuclei, whereas some areas demonstrated unidentified cellular morphology (Fig. 1D); (5) basal cells with prominent nuclei and high nucleus/ cytosol (N/C) ratios but no identifiable cellular borders (Fig. 1E); or (6) cellular borders and low N/C ratios (Fig. 1F). FIGURE 1. In vivo confocal photomicrographs showing the different Image Analysis of Corneal Apical Surface Cells patterns of basal epithelial cells. (A) Basal epithelial cells with an intact cellular border without visible nuclei. This pattern was seen in all eyes Cells above the basal cells were defined as corneal apical surface cells. before surgery and in cells thought to be in a late differentiation phase In the early postoperative period, the severe epithelial cellular damage during wound healing (image obtained before surgery). (B) Basal cells makes it difficult to identify the corneal epithelial cells precisely at with an elongated shape and an intact cellular border but no visible different depths. However, it is easy to identify basal epithelial cells nuclei. This pattern was seen only during the first few days after because of their typical appearance, as shown in Figure 1, and their epi-LASIK (image from postoperative day 1). (C) Basal epithelial cells with amorphous and poorly identified morphology. This finding was location immediately superficial to the corneal stroma. Thus, those though to be caused by cellular necrosis or high reflectivity due to cells that were located above the basal cells were defined as corneal superficial cellular pathology and was seen only during the first few apical surface cells in this study. Apical cells cannot exist if no basal weeks after epi-LASIK (image from postoperative day 3). (D) Basal layer is visualized under this definition. It is also impossible to identify epithelial cells with patchy white or necrotic areas (large arrows)in the origin of the different layers of cells in this study by in vivo confocal the cellular sheet. Small arrows: intact cellular borders without prom- microscopy. inent nuclei. This pattern was seen only during the first few days after Results were classified as cells showing large and flat superficial epi-LASIK (image from postoperative day 1). (E) Regenerated basal epithelia with small nuclei (Fig. 2A); elongated superficial epithelial cells with prominent nuclei but no cellular borders. The N/C ratio was cells (Fig. 2B); cells with high N/C ratios but without the normal large, high. This pattern was seen only during the first few weeks after epi-LASIK (image from postoperative day 5). (F) Further regenerating squamous appearance (Fig. 2C); large, flat epithelial cells with multi- cells with a decreased N/C ratio and newly formed cellular borders. directional elongation (Fig. 2D); exfoliating superficial epithelial cells This pattern was seen only during the first few weeks after epi-LASIK, resembling sloughing or necrotic cells (Fig. 2E); or amorphous and and usually appeared later than the pattern shown in (E). The image is poorly identified cells (Fig. 2F). from postoperative week 2. Image Analysis of the Stromal Reaction The stromal reaction was evaluated by stromal scatter based on the determine corneal epithelial thickness, which was the distance be- reflectivity (Confoscan 3; Nidek Technologies) tied to the light inten- tween the innermost basal epithelium and the most superficial apical sity.17 The z-scan system (a graphic showing the depth coordinate on surface epithelial cells. the z-axis and the level of reflectivity on the y-axis) with the profile of scattered light assessment was used. Three consecutive measurements Statistical Analysis were performed on the same examination, to quantify the light inten- Data regarding corneal epithelial thickness and stromal reactions were sity. Three parameters were used to represent the stromal reaction: the calculated as means Ϯ standard deviations. Statistical analyses of the peak of the light intensity in the whole stroma, the average light results were performed by Student’s t-test for postoperative stromal intensity of the anterior stroma within the anterior 50 ␮m of depth, reaction. P Ͻ 0.05 was considered statistically significant. and the average light intensity of the whole stromal layer.

Measurement of Corneal Epithelial Thickness RESULTS All areas of the z-scan curve where the basal epithelium and the apical Patients who completed the 1-, 3-, and 6-month follow-up surface epithelial cells could be clearly recognized were recorded. The examinations numbered 27 (24 and 22 eyes in the groups with depth values on the z-axis indicated by the software were used to and without MMC treatment, respectively), 25 (22 and 20 eyes

Downloaded from iovs.arvojournals.org on 09/30/2021 2418 Chen et al. IOVS, June 2008, Vol. 49, No. 6

both the groups during the ﬁrst few weeks after surgery. Cells with cellular borders and low N/C ratios (Fig. 1F) were also noted in the ﬁrst few weeks in both groups. However, this pattern usually appeared later than the pattern shown in Figure 1E.

Healing of Apical Surface Epithelial Cells Observed by In Vivo Confocal Microscopy Table 2 summarizes the distribution of morphologic features on the apical surface epithelial cells. Large, flat apical surface epithelia with small cell nuclei were observed in eyes before surgery and at 2 weeks or longer after surgery (Fig. 2A). The corneal apical surface epithelial cells in the MMC and non- MMC groups recovered to this pattern (Fig. 2A) in 12.5% and 13.6% of the eyes at 2 weeks, 37.5% and 54.5% at 1 month, 52.4% and 57.9% at 6 months, respectively. During the first few days after surgery, elongated superficial epithelial cells (Fig. 2B) and large flat epithelial cells with multidirectional elongation (Fig. 2D) were observed in the MMC and non-MMC groups. The cause of these patterns (Figs. 2B and 2D) was thought to be mechanical stretching of epithelial flaps during surgery. Exfoliating superficial epithelial cells (Fig. 2E) and amorphous cells (Fig. 2F) were believed to be caused by severe pathologic changes (most likely cell death) and were found only in the first few weeks after surgery in the groups treated with and without MMC. Cells with a high N/C ratio but without normal large, squamous appearance were FIGURE 2. In vivo confocal photomicrographs showing the different considered active regenerative cells without full differentiation patterns of corneal apical surface epithelial cells. (A) Normal squamous (Fig. 2C). Such cells could be found after postoperative day 3, superficial cells. This pattern was found in preoperative eyes and in the and were seen in most eyes between 3 weeks and 3 months late healing stages, mostly after 3 weeks (image obtained before surgery). (B) Elongated superficial epithelial cells. This pattern was seen after surgery. only during the first few days after epi-LASIK (image from postoperative day 1). (C) Nonsquamous epithelial cells with a high N/C ratio. Stromal Reaction Observed by In Vivo These cells were thought to be regenerating cells without full differ- Confocal Microscopy entiation. This pattern was found after postoperative day 3 and was seen in most eyes between postoperative week 3 and month 3 (image The peak of the light intensity in the whole stroma was found from postoperative week 3). (D) Apical surface epithelial cells with within the anterior 12 ␮m of depth in all eyes. Table 3 presents multidirectional elongation. This pattern was seen only at postopera- the comparison of the stromal reactions between the MMC and tive day 1. (E) Exfoliating superficial epithelial cells, which were non-MMC treatment groups. In all three parameters which believed to be sloughing or necrotic. This pattern was seen only during represent stromal reaction, there were no differences observed the first few weeks after epi-LASIK (image from postoperative day 1). between the groups at 1, 3, and 6 months after surgery (P Ͼ (F) Amorphous cells. Difficulty in identifying these cells was thought to 0.05). be caused by the high reflectivity caused by cellular necrosis. This pattern was seen only during the first few weeks after epi-LASIK (image from postoperative day 1). Corneal Epithelial Thickness The total corneal epithelial thicknesses in the MMC and non- MMC groups were, respectively, 54.2 Ϯ 4.5 and 56.1 Ϯ 5.1 ␮m with and without MMC), and 22 (21 and 19 eyes with and at 1 month (P Ͼ 0.05), 56.1 Ϯ 6.7 and 55.2 Ϯ 4.9 at 3 months without MMC). (P Ͼ 0.05), and 58.1 Ϯ 5.1 and 57.3 Ϯ 7.2 at 6 months (P Ͼ 0.05). Healing of Basal Cells Observed by In Vivo Confocal Microscopy Representative Cases Table 1 shows the distribution of basal epithelial morphologies Patient 1 (case 1) was a 24-year-old man who had a preopera- in the treatment groups. Before surgery, all eyes had intact tive refractive error in his right eye of Ϫ4.0 D. Figure 3 shows cellular borders without visible nuclei (Fig. 1A). This pattern the sequential confocal microscopic findings in the basal epi- was also observed in cells thought to be in a late differentiating thelial cells after epi-LASIK. The cells looked normal on day 1 phase during wound healing. In the MMC and non-MMC treat- but demonstrated notable pathologic changes (amorphous and ment groups, the corneal basal epithelial cells returned to this poorly identified basal cells) on days 3 and 5. On day 7, small pattern (Fig. 1A) in 0% and 13.6% of the eyes after 1 week, cells with prominent nuclei but no identifiable cellular borders 37.5% and 36.4% after 2 weeks, and 87.5% and 86.3% after 1 were found. On day 14, the basal cells had cellular borders and month, respectively. a lower N/C ratio. Normal basal cells with intact cellular bor- During the first few days after epi-LASIK, elongated cells ders but no visible nuclei appeared at 1 month after surgery. with intact cellular borders but no visible nuclei (Fig. 1B) and Patient 2 (case 2) was a 37-year-old woman who underwent patchy whitening of the cellular sheet (Fig. 1D) were observed epi-LASIK in her left eye, which had a preoperative refractive in the MMC and non-MMC groups. Amorphous and poorly error of Ϫ5.5 D. Figure 4 shows the sequential results. The identified cells (Fig. 1C) were noted in the first few weeks in basal epithelial cells appeared healthy throughout the observa- both groups. Cells with prominent nuclei and high N/C ratios tional period. However, pathologic changes were seen in the but no identifiable cellular borders (Fig. 1E) were observed in apical surface cells on days 1 and 7. After 2 weeks, these cells

Downloaded from iovs.arvojournals.org on 09/30/2021 IOVS, June 2008, Vol. 49, No. 6 In Vivo Confocal Microscopy and Epi-LASIK 2419

TABLE 1. Different Patterns of Basal Epithelial Cells Classiﬁed by In Vivo Confocal Microscopy

Patchy Border (؉) Whitening of Border (؊) Border (؉) Total Time after Surgery Nucleus (؊) Elongated Amorphous Cell Sheet Nucleus (؉) Nucleus (؉) Number

Eyes with MMC treatment 1 d 8.3 (2) 8.3 (2) 62.5 (15) 20.8 (5) 0 (0) 0 (0) 24 3 d 0 (0) 4.2 (1) 50 (12) 4.2 (1) 37.5 (9) 4.2 (1) 24 7 d 0 (0) 0 (0) 25 (6) 0 (0) 41.6 (10) 33.3 (8) 24 2 wk 37.5 (9) 0 (0) 8.3 (2) 0 (0) 0 (0) 54.2 (13) 24 3 wk 66.7 (16) 0 (0) 0 (0) 0 (0) 0 (0) 33.3 (8) 24 1 mo 87.5 (21) 0 (0) 0 (0) 0 (0) 0 (0) 12.5 (3) 24 3 mo 100 (22) 0 (0) 0 0 (0) 0 0 (0) 22 6 mo 100 (21) 0 (0) 0 0 (0) 0 0 (0) 21 Eyes without MMC treatment 1 d 31.8 (7) 4.5 (1) 45.5 (10) 18.2 (4) 0 (0) 0 (0) 22 3 d 9.1 (2) 0 (0) 50 (11) 0 (0) 36.4 (8) 4.6 (1) 22 7 d 13.6 (3) 0 (0) 22.7 (5) 0 (0) 31.8 (7) 31.8 (7) 22 2 wk 36.4 (8) 0 (0) 4.5 (1) 0 (0) 9.1 (2) 50 (11) 22 3 wk 72.7 (16) 0 (0) 0 (0) 0 (0) 0 (0) 27.3 (6) 22 1 mo 86.3 (19) 0 (0) 0 (0) 0 (0) 0 (0) 13.7 (3) 22 3 mo 100 (20) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 20 6 mo 100 (19) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 19

Data are the percentage (number) of eyes. MMC: MMC treatment during operation; border (ϩ) nucleus (Ϫ), intact cellular borders without visible nuclei; elongated: elongated basal epithelial cells with intact cellular borders, but no visible nuclei; amorphous: amorphous and poorly identified basal epithelial cells; patchy whitening of cell sheet: cells in some areas showed normal appearance, whereas cells in other areas demonstrated patchy whitening, and cellular morphology was not identifiable; border (Ϫ) nucleus (ϩ), basal cells with prominent nuclei but no identifiable cellular borders (a high N/C ratio was found); border (ϩ) nucleus (ϩ): cells with cellular border and low N/C ratio.

had high N/C ratios but not the normal, large, squamous hemidesmosomes are lost and provisional attachment com- appearance. A normal, large, flat epithelium with small nuclei plexes, called focal contacts, form.20–22 The epithelial cells was observed after 1 month. Among all the eyes, only two in flatten, migrate, and slide as a single-layered intact sheet to the group with no MMC treatment were found to have a cover the denuded surface.23 During the second phase, cells morphologically healthy basal epithelium, as seen in case 2 distal to the original wound proliferate to repopulate the throughout the observational period. wound area, and cell stratification and differentiation occur.19,20 In the third phase, hemidesmosomes reform, and 18,19 DISCUSSION extracellular matrix is synthesized and reassembled. After epi-LASIK, healing of the corneal epithelium may dif- Normal wound healing after epithelial debridement involves fer from normal epithelial healing. The replaced epithelial flap, three distinct but continuous phases.18,19 In the first phase, whether dead or alive, stays in front of the leading edge, and a

TABLE 2. Different Patterns of Apical Surface Cells Classiﬁed by In Vivo Confocal Microscopy

Nonsquamous Normal with Large Multidirectional Total Time after Surgery Squamous Elongated N/C Elongated Exfoliating Amorphous Number

Eyes with MMC treatment 1 d 0 (0) 16.7 (4) 0 (0) 8.3 (2) 33.3 (8) 41.7 (10) 24 3 d 0 (0) 8.3 (2) 8.3 (2) 0 (0) 33.3 (8) 50 (12) 24 7 d 0 (0) 4.2 (1) 20.8 (5) 0 (0) 37.5 (9) 37.5 (9) 24 2 wk 12.5 (3) 0 (0) 54.1 (13) 0 (0) 16.7 (4) 16.7 (4) 24 3 wk 25 (6) 0 (0) 54.1 (13) 0 (0) 0 (0) 20.8 (5) 24 1 mo 37.5 (9) 0 (0) 62.5 (15) 0 (0) 0 (0) 0 (0) 24 3 mo 50 (11) 0 (0) 50 (11) 0 (0) 0 (0) 0 (0) 22 6 mo 52.4 (11) 0 (0) 47.6 (10) 0 (0) 0 (0) 0 (0) 21 Eyes without MMC treatment 1 d 0 (0) 13.6 (3) 0 (0) 4.5 (1) 31.8 (7) 50 (11) 22 3 d 0 (0) 9.1 (2) 13.6 (3) 0 (0) 31.8 (7) 45.5 (10) 22 7 d 0 (0) 0 (0) 18.2 (4) 0 (0) 36.4 (8) 45.4 (10) 22 2 wk 13.6 (3) 0 (0) 36.4 (8) 0 (0) 22.7 (5) 27.3 (6) 22 3 wk 31.6 (7) 0 (0) 54.5 (12) 0 (0) 0 (0) 11.6 (3) 22 1 mo 54.5 (12) 0 (0) 45.5 (10) 0 (0) 0 (0) 0 (0) 22 3 mo 50 (10) 0 (0) 50 (10) 0 (0) 0 (0) 0 (0) 22 6 mo 57.9 (11) 0 (0) 42.1 (8) 0 (0) 0 (0) 0 (0) 19

Data are the percentage (number) of eyes. MMC, mitomycin C treatment during operation; normal squamous: large, ﬂat epithelium with low N/C ratio; elongated: elongated apical surface cells; nonsquamous with large N/C: cells with large N/C ratio but without normal large, squamous appearance; exfoliating: apical surface cells that look like sloughing or necrotic cells; amorphous: high reﬂectivity and amorphous patterns hindering detailed observation.

Downloaded from iovs.arvojournals.org on 09/30/2021 2420 Chen et al. IOVS, June 2008, Vol. 49, No. 6

TABLE 3. Stromal Reaction Represented by Intensity of Light (z-Scan Proﬁle)

Time after Surgery With MMC Without MMC P

Peak of stromal light intensity 1mo 65.6 Ϯ 22.9 71.3 Ϯ 27.8 NS 3mo 61.0 Ϯ 24.2 66.8 Ϯ 34.2 NS 6 mo 56.25 Ϯ 21.6 58.5 Ϯ 17.1 NS Average light intensity of the anterior 50-␮m depth of the stroma 1mo 59.8 Ϯ 11.4 64.2 Ϯ 20.8 NS 3mo 56.6 Ϯ 14.7 62.7 Ϯ 16.3 NS 6mo 52.3 Ϯ 7.1 52.2 Ϯ 6.7 NS Average light intensity of the whole stroma 1mo 48.3 Ϯ 12.5 52.2 Ϯ 18.4 NS 3mo 47.6 Ϯ 19.2 49.7 Ϯ 13.6 NS 6mo 41.5 Ϯ 12.1 41.5 Ϯ 19.3 NS

MMC, mitomycin C treatment during surgery. NS, nonsigniﬁcant by Student’s t-test.

denuded surface is lacking. In addition, preservation of the Discrepant results from different studies on the survival of basement membrane under the epithelial flap may exert spe- epikeratome-created epithelial flaps have been reported. Pal- cific effects.9,10 Moreover, laser treatment of the stromal bed likaris et al.10,29 initially concluded that the epikeratome pro- may induce an epithelial–stromal interaction and interfere with duces epithelial flaps with a sharply separated basement mem- wound healing.24–26 Finally, pharmacologic agents used during brane and intact basal cells, although occasional focal surgery may contribute to wound healing. disruption of the basal lamina was also observed. Their more We examined specific epithelial wound healing after epi- recent study showed that most of the epithelial cells were LASIK by using in vivo confocal microscopy. This method has morphologically normal with only minor cell degeneration.9 been used to observe corneal epithelial healing in various Tanioka et al.16 demonstrated that most of the basal cells in conditions. Cho et al.27 found that healing superficial cells in epithelial flaps created with different epikeratome devices rabbit corneas after epithelial debridement were initially ellip- were PI-positive dead cells. Because of the difficulty in obtain- tical and large. By the end of the second week, the cells were ing human corneal epithelial sheets for evaluation, these stud- almost the same size and shape as normal cells. In limbal stem ies observed only flap survival in the first few days after epi- cell–deficient corneas, healing epithelial cells were smaller and LASIK.9,10,16 Our long-term results are more similar to those of more variable in size than were normal corneal cells.27 In Tanioko et al.16 In the group without MMC treatment, 31.8% of post-PRK rabbit corneas, small superficial corneal epithelial the eyes had basal cells with normal morphology (Fig. 1A) on cells with high N/C ratios were found at 1 week after surgery. postoperative day 1; the incidence decreased dramatically on The corneal epithelium recovered normal cellular morphology days 3 and 7. Those cells with morphology as shown in Figures at 2 weeks.28Although these studies provided important infor- 1B, 1C, and 1D were thought to have undergone severe patho- mation about epithelial wound healing, the authors did not logic changes—most likely, cell death. The healing basal epi- describe observations in different layers of corneal epithelial thelial cells during the first few days after surgery showed high cells. N/C ratios without visible cellular borders (Fig. 1E). Visible

FIGURE 3. Case 1: In vivo confocal photomicrographs show changes in the corneal basal epithelial cells after surgery. On day 1, morphologically intact cells had easily identified cellular borders. On days 3 and 5, high reflectivity hindered observation of the cellular morphology, as it did in most eyes. However, loss of intact cellular borders associated with to- tally disorganized structure were still observed. On day 7, compact basal cells had prominent nuclei and a high N/C ratio, but no cellular border was found. On day 14, regenerating cells had prominent nuclei and newly forming cellular borders. On day 28, basal cells had recovered their preoperative status. Nuclei could not be seen, although cellular borders were easily identified.

Downloaded from iovs.arvojournals.org on 09/30/2021 IOVS, June 2008, Vol. 49, No. 6 In Vivo Confocal Microscopy and Epi-LASIK 2421

FIGURE 4. Case 2. In vivo confocal photomicrographs show changes in the corneal basal and apical surface epithelial cells after surgery. On postoperative days 1 and 7, at week 2, and at month 1, morphologically intact basal epithelial cells were found. Amorphous apical surface cells were observed on postoperative days 1 and 7. At 2 weeks after surgery, apical surface cells were compact and small with prominent nuclei and a high N/C ratio. However, no well- differentiated squamous pattern was found. At 1 month after surgery, well- differentiated, squamous cells were seen on the apical surface.

cellular borders usually reappeared at 1 to 2 weeks after sur- The exact process of epithelial healing after epi-LASIK is still gery (Fig. 1F). At that time, basal cellular nuclei were usually not clearly understood despite the information provided by in seen, but the N/C ratios markedly decreased. After 3 weeks, vivo confocal microscopy. Gradual sloughing of the damaged 72.7% of the eyes had no visible nuclei. To our knowledge, we epithelial sheets may have occurred, followed by replacement are the first to report the appearance and subsequent disap- with multilayered, highly active, new and growing cells. Be- pearance of basal cellular nuclei, combined with the disappear- cause no denuded surface area was seen ahead of the leading ance and subsequent reappearance of basal cellular borders edge (as noted in corneal debridement or PRK), we can easily during wound healing, evaluated by in vivo confocal micros- explain why we saw no single layer of migrated cells, as is copy. Although not proved, a large and visible nucleus may observed in phase I of normal epithelial healing. Features represent new, actively growing cells that are metabolically indicating the highly metabolic and undifferentiated status of active and unstable. The reestablishment of cellular borders the new growing cells were the increased N/C ratios of the may represent reconstruction of the cellular junction, which basal and apical surface cells, the disappearance of cellular implies a return to normal and stable conditions. borders in basal cells, and the disappearance of the large The recovery of normal morphology took longer in apical squamous cells on the apical surface. However, we cannot rule surface cells than in basal cells. In both the MMC and non-MMC out the possibility that new growing cells from the periphery treatment groups, most eyes had abnormal apical surface epi- of the cornea may have mixed with surviving cells on the flap thelial cells during the first few weeks. Abnormalities included and that together they composed the final cellular sheet. elongated, amorphous, multidirectional, or unidentifiable mor- Several of our findings need further clarification. First, our phology. Abnormal cells were thought to have undergone observational period and methods differed from those of pre- severe pathologic changes—most likely, cell death. In the vious studies. In published reports about epi-LASIK, healthy intermediate stage of healing between 3 weeks and 3 months, corneal epithelial sheets were observed only in the first 24 most eyes had small and compact apical surface cells, with hours, and immunohistochemistry and electron microscopy high N/C ratios but no squamous appearance (Fig. 2C). These were used for the examination.9,10,16 Our cross-sectional views cells were thought to be in higher metabolic, earlier differen- of the epithelial sheets, long observation times, and in vivo tiation stages than normal squamous cells. observations may add important information. Second, use of

Downloaded from iovs.arvojournals.org on 09/30/2021 2422 Chen et al. IOVS, June 2008, Vol. 49, No. 6

different epikeratomes could lead to different cutting levels their normal morphology, and apical surface cells needed more and affect the survival of the cellular sheet, although Tanioka et time than basal cells. Finally, the stromal reaction was the same al.16 reported dead cells in epithelial flaps prepared with dif- in the eyes with higher myopia treated with MMC as in the eyes ferent epikeratomes. Third, because no similar study has been with lesser myopia without MMC treatment. Because cellular published, we have justified the grading scheme of corneal conditions cannot be properly assessed by morphologic anal- epithelial wound healing on the basis of the presumed corre- ysis alone, sequential evaluations using apoptosis assays, im- lation with severity just by morphologic interpretation. Further munohistochemical staining, and/or electron microscopy are histologic studies are needed to confirm our hypothesis. needed to provide further information about the corneal Fourth, during postoperative week 1, most examinations were wound-healing process after epi-LASIK. performed with therapeutic bandage contact lenses in place. To rule out the possibility that contact lens fitting might change the in vivo microscopic appearance of corneal epithe- References lial cells, we did a preliminary test in rabbit eyes and found that 1. Binder PS. Ectasia after laser in situ keratomileusis. J Cataract placement of a contact lens did not change the appearance of Refract Surg. 2003;29:2419–2429. the healing basal and apical surface epithelial cells (data not 2. Wilson SE. Clinical practice: use of lasers for vision correction of shown). We also compared the appearance of the basal and nearsightedness and farsightedness. N Engl J Med. 2004;351:470– apical cells in the patients before and after removal of the 475. contact lenses, and found no difference in apical cellular ap- 3. Randleman JB. Post-laser in-situ keratomileusis ectasia: current pearance (data not shown). Fifth, the stromal reaction based understanding and future directions. Curr Opin Ophthalmol. on the reflectivity tied to the light intensity in this study should 2006;17:406–412. be interpreted carefully. Potentially strong confounding factors 4. Schallhorn SC, Amesbury EC, Tanzer DJ. Avoidance, recognition, may exist and make the results difficult to analyze. Sixth, to and management of LASIK complications. Am J Ophthalmol. inhibit potential haze, we used MMC and a chilled physiologic 2006;141:733–739. saline solution in patients with high degrees of myopia. MMC 5. Fagerholm P. Wound healing after photorefractive keratectomy. J has been widely accepted as an adjunctive therapy for corneal Cataract Refract Surg. 2000;26:432–447. haze after higher myopic corrections by PRK and LASEK.30–34 6. Pallikaris IG, Katsanevaki VJ, Kalyvianaki MI, Naoumidi II. Ad- It has been hypothesized that MMC inhibits fibroblast prolifer- vances in subepithelial excimer refractive surgery techniques: epi- ation and differentiation, consequently blocking myofibroblast LASIK. Curr Opin Ophthalmol. 2003;14:207–212. formation through its various potent effects.35,36 Although 7. Taneri S, Zieske JD, Azar DT. Evolution, techniques, clinical out- epi-LASIK is intended to create a cellular sheet with an intact comes, and pathophysiology of LASEK: review of the literature. basement membrane, the morphologically unhealthy epithelial Surv Ophthalmol. 2004;49:576–602. cells we found may still be able to induce an epithelial–stromal 8. Netto MV, Wilson SE. Indications for excimer laser surface abla- reaction as in PRK or LASIK. There are only limited studies tion. J Refract Surg. 2005;21:734–741. reported on the use of MMC in epi-LASIK. In this study, we 9. Katsanevaki VJ, Naoumidi II, Kalyvianaki MI, Pallikaris IG. Epi- found that higher myopic groups (ϾϪ6.0 D) treated with LASIK: histological findings of separated epithelial sheets 24 hours MMC may have similar stromal reactions compared with the after treatment. J Refract Surg. 2006;22:151–154. low myopic group (ϽϪ6.0 D) without MMC treatment. Our 10. Pallikaris IG, Naoumidi II, Kalyvianaki MI, Katsanevaki VJ. Epi- LASIK: comparative histological evaluation of mechanical and al- results support the widely accepted view that MMC can pre- cohol-assisted epithelial separation. J Cataract Refract Surg. 2003; vent corneal haze formation in surface ablation for higher 29:1496–1501. myopic corrections. We also found that the group with MMC 11. Long Q, Chu R, Zhou X, et al. Correlation between TGF-beta 1 in treatment had a higher incidence of early corneal basal epithe- tears and corneal haze following LASEK and Epi-LASIK. J Refract lial damage compared with the group without MMC treatment. Surg. 2006;22:708–712. In addition, the only two eyes that had morphologically intact 12. Dai J, Chu R, Zhou S, et al. One year outcomes of epi-LASIK for basal epithelial morphology during the entire observational myopia. J Refract Surg. 2006;22:589–595. period were without MMC treatment. Although we tried to 13. Kalyvianaki MI, Katsanevaki VJ, Kavroulaki DS, et al. Comparison avoid direct contact between the MMC solution and the epi- of corneal sensitivity and tear function following Epi-LASIK or laser thelial sheets and we protected the epithelial sheet during in situ keratomileusis for myopia. Am J Ophthalmol. 2006;142: vigorous irrigation of the stromal bed with the chilled saline 669–671. solution, mechanical or pharmacologic cellular damage caused 14. Matsumoto JC, Chu YS. Epi-LASIK update: overview of techniques by MMC or irrigation itself may still have occurred during and patient management. Int Ophthalmol Clin. 2006;46:105–115. surgery. Because almost all eyes in the groups with and with- 15. Torres LF, Sancho C, Tan B, Padilla K, Schanzlin DJ, Chayet AS. out MMC treatment have severe cellular damage of the apical Early postoperative pain following Epi-LASIK and photorefractive surface epithelial cells within 2 weeks after surgery, it is diffi- keratectomy: a prospective, comparative, bilateral study. J Refract cult to evaluate the effect of MMC on these cells. We found no Surg. 2007;23:126–132. differences in corneal epithelial thickness between the two 16. Tanioka H, Hieda O, Kawasaki S, Nakai Y, Kinoshita S. Assessment groups at 1, 3, and 6 months after surgery. The healing patterns of epithelial integrity and cell viability in epithelial flaps prepared of basal and apical cells were also similar at 3 and 6 months with the epi-LASIK procedure. J Cataract Refract Surg. 2007;33: after surgery. The effect of MMC on corneal epithelial wound 1195–1200. healing was thus thought to occur only in the early postoper- 17. Pisella PJ, Auzerie O, Bokobza Y, Debbasch C, Baudouin C. Eval- uation of corneal stromal changes in vivo after laser in situ kera- ative period. tomileusis with confocal microscopy. Ophthalmology. 2001;108: In conclusion, our study demonstrated several important 1744–1750. new findings. First, we believe this is the first human study in 18. Zieske JD, Gipson IK. Agents that affect corneal wound healing: which in vivo confocal microscopy was used to study early and modulation of structure and function. In: Albert DM, Jakobiec FA, late wound healing after surface ablation, specifically epi- eds. Principles and Practice of Ophthalmology. 2nd ed. Vol. 1. LASIK. Second, the corneal epithelial flaps after epi-LASIK were Philadelphia: WB Saunders; 2000;364–372. not as healthy as previously thought, as cells underwent severe 19. Gipson IK, Spurr-Michaud SJ, Tisdale AS. Anchoring fibrils form a pathologic changes within or after 24 hours. Third, at least complex network in human and rabbit cornea. Invest Ophthalmol several weeks were required for the epithelial cells to return to Vis Sci. 1987;28:212–220.

Downloaded from iovs.arvojournals.org on 09/30/2021 IOVS, June 2008, Vol. 49, No. 6 In Vivo Confocal Microscopy and Epi-LASIK 2423

20. Klatte DH, Kurpakus MA, Grelling KA, Jones JC. Immunochemical 28. Moller-Pedersen T, Li HF, Petroll WM, et al. Confocal microscopic characterization of three components of the hemidesmosome and characterization of wound repair after photorefractive keratec- their expression in cultured epithelial cells. J Cell Biol. 1989;109: tomy. Invest Ophthalmol Vis Sci. 1998;39:487–501. 3377–3390. 29. Pallikaris IG, Kalyvianaki MI, Katsanevaki VJ, Ginis HS. Epi-LASIK: 21. Okada Y, Saika S, Shirai K, et al. Disappearance of desmosomal preliminary clinical results of an alternative surface ablation pro- components in rat corneal epithelium during wound healing. Oph- cedure. J Cataract Refract Surg. 2005;31:879–885. thamologica. 2001;215:61–65. 30. Schipper I, Suppelt C, Gebbers JO. Mitomycin C reduces scar 22. Crosson CE, Klyce SD, Beuerman RW. Epithelial wound closure in formation after excimer laser (193 nm) photorefractive keratec- the rabbit cornea: a biphasic process. Invest Ophthalmol Vis Sci. tomy in rabbits. Eye. 1997;11:649–655. 1986;27:464–473. 31. Sadeghi HM, Seitz B, Hayashi S, LaBree L, McDonnell PJ. In vitro 23. Wilson SE, Liu JJ, Mohan RR. Stromal-epithelial interactions in the effects of mitomycin C on human keratocytes. J Refract Surg. 1998;14:534–540. cornea. Prog Retin Eye Res. 1999;18:293–309. 32. Majmudar PA, Forstot SL. Dennis RF et al. Topical mitomycin C for 24. Wilson SE, Mohan RR, Ambrosio R Jr, et al. The corneal wound subepithelial fibrosis after refractive corneal surgery. Ophthalmol- healing response: cytokine-mediated interaction of the epithelium, ogy. 2000;107:89–94. stroma, and inflammatory cells. Prog Retin Eye Res. 2001;20:625– 33. Carones F, Vigo L, Scandola E, Vacchini L. Evaluation of the 637. prophylactic use of mitomycin C to inhibit haze formation after 25. Nakamura K, Kurosaka D, Bissen-Miyajima H, et al. Intact corneal photorefractive keratectomy. J Cataract Refract Surg. 2002;28: epithelium is essential for the prevention of stromal haze after 2088–2095. laser assisted in situ keratomileusis. Br J Ophthalmol. 2001;85: 34. Argento C, Cosentino MJ, Ganly M, Comparison of laser epithelial 209–213. keratomileusis with and without the use of mitomycin C. J Refract 26. Javier JA, Lee JB, Oliveira HB, et al. Basement membrane and Surg. 2006;22:782–786. collagen deposition after laser subepithelial keratomileusis and 35. Lee JS, Oum BS, Lee SH. Mitomycin C influence on inhibition of photorefractive keratectomy in the leghorn chick eye. Arch Oph- cellular proliferation and subsequent synthesis of type I collagen thalmol. 2006;124:703–709. and laminin in primary and recurrent pterygia, Ophthalmic Res. 27. Cho BJ, Djalilian AR, Holland EJ. Tandem scanning confocal mi- 2001;33:140–146. croscopic analysis of differences between epithelial healing in 36. Mohan RR, Hutcheon AE, Choi R, et al. Apoptosis, necrosis, pro- limbal stem cell deficiency and normal corneal reepithelization in liferation, and myofibroblast generation in the stroma following rabbits. Cornea. 1998;17:68–73. LASIK and PRK. Exp Eye Res. 2003;76:71–87.

Downloaded from iovs.arvojournals.org on 09/30/2021