Laser in situ keratomileusis (LASIK)

Corneal Stromal Changes Induced by Myopic LASIK Minna Vesaluoma, 1 Juan Pérez Santonja, 2 W. Matthew Petroll, 3 Tuuli Linna, 1 Jorge Alió, 2 and Timo Tervo 1 PURPOSE. Despite the rapidly growing popularity of laser in situ keratomileusis (LASIK) in correction of myopia, the tissue responses have not been thoroughly investigated. The aim was to characterize morphologic changes induced by myopic LASIK in human corneal stroma. METHODS. Sixty-two myopic eyes were examined once at 3 days to 2 years after LASIK using in vivo confocal microscopy for measurement of flap thickness, keratocyte response zones, and objective grading of haze. RESULTS. Confocal microscopy revealed corneal flap interface particles in 100% of eyes and microfolds at the Bowman s layer in 96.8%. The flaps were thinner (112 25 m) than intended (160 m). The keratocyte activation in the stromal bed was greatest on the third postoperative day. Patients with increased interface reflectivity due to abnormal extracellular matrix or activated keratocytes at 1 month (n 9) had significantly thinner flaps than patients with normal interface reflectivity (n 18; 114 12 versus 132 22 m, P 0.027). After 6 months the mean density of the most anterior layer of flap keratocytes was decreased. CONCLUSIONS. Keratocyte activation induced by LASIK was of short duration compared with that reported after photorefractive keratectomy. The flaps were thinner than expected, and microfolds and interface particles were common complications. The new findings such as increased interface reflectivity associated with thin flaps and the apparent loss of keratocytes in the most anterior flap 6 months to 2 years after surgery may have important clinical relevance. (Invest Ophthalmol Vis Sci. 2000;41:369 376) Laser in situ keratomileusis (LASIK) 1 is a new technique for the correction of moderate to high myopia, with explosively increasing popularity worldwide. A hinged flap (consisting of the surface epithelium, Bowman s layer, and anterior stroma) is first created using a microkeratome. The flap is folded back, and the exposed stroma is photoablated using an excimer laser. The flap is then returned into place to cover the treated area. In another refractive surgical procedure, photorefractive keratectomy (PRK), the epithelium is first removed, the exposed stroma is photoablated, and the epithelial defect then heals in 2 to 4 days. After both procedures, flattening of the central corneal curvature due to tissue removal results in decreased refractive power (myopic correction). Neither of the procedures has proven superior in terms of efficacy outcomes. 2 4 However, LASIK offers advantages (such as minimal postoperative pain and faster clinical and functional recovery, as well as less regression and haze formation). 2,5 An increasing number of studies are being published on clinical outcome after LASIK, 2,4,6 11 but few reports address the biological changes associated with LASIK. 12 18 Recently, in vivo confocal microscopy has been introduced as a tool for the evaluation of wound healing after refractive surgery in humans. 12,15,19 22 In addition, the confocal microscopy through focusing technique (CMTF) has been developed for measurement of corneal sublayer thickness and estimation of the intensity of postoperative haze. 21,23 Our aim was to characterize changes in corneal keratocyte and stromal morphology induced by myopic LASIK in humans, using in vivo confocal microscopy. Special attention was paid to the reactions at the corneal flap interface. From the 1 Department of Ophthalmology, Helsinki University Central Hospital, Helsinki, Finland; the 2 Refractive Surgery and Cornea Unit, Alicante Institute of Ophthalmology, University of Alicante, School of Medicine, Alicante, Spain; and the 3 Department of Ophthalmology, University of Texas, Southwestern Medical Center, Dallas, Texas. Supported by The Instrumentarium Scientific Foundation (MV, TL), Alicante Institute of Ophthalmology, Alicante, Spain (MV, WMP, TL); the Finnish Medical Council (TT, WMP); Helsinki University Central Hospital (TT, MV); the Finnish Eye and Tissue Bank Foundation (MV, TL); the Finnish Eye Foundation (MV, TL, TT), the Finnish Medical Foundation (MV), and Ella and Georg Ehrnrooth Foundation (MV). Submitted for publication May 28, 1999; revised August 30, 1999; accepted September 23, 1999. Commercial relationships policy: N. Corresponding author: Minna Vesaluoma, Helsinki University Eye Hospital, Eye Bank, P. O. Box 220, FIN-00029 HUCH, Finland. minna.vesaluoma@huch.fi METHODS Patients Sixty-two eyes of 62 patients (38 females and 24 males; age 34.6 8.9 years, mean SD) who had undergone myopic LASIK were examined once after surgery using in vivo confocal microscopy. The preoperative spherical equivalent (SE) of refraction was 6.83 3.10 D (range, 1.75 to 15.00 D). The Ethical Review Committee of Helsinki University Eye Hospital approved the research plan, and it followed the tenets of the Declaration of Helsinki. Each patient gave an informed consent. Fourteen patients were examined on the third postoperative day; other time points were 1 to 2 weeks (n 16), 1 to 2 months (n 12), 3 months (n 11), and 6 months to 2 years ( 6 months, n 9). Six patients were also examined preoperatively. Investigative Ophthalmology & Visual Science, February 2000, Vol. 41, No. 2 Copyright Association for Research in Vision and Ophthalmology 369

370 Vesaluoma et al. IOVS, February 2000, Vol. 41, No. 2 corticosteroid (Fluorometholone 0.1%, FML; Allergan S.A., Madrid, Spain) eyedrops were instilled four times a day for the first 10 days. FIGURE 1. Clinical findings after LASIK. Metal particles (A) at the flap interface. Striae (B) are more much commonly encountered in the flap than thicker folds (C). LASIK Procedure LASIK procedures were performed using the Automated Corneal Shaper microkeratome (ALK-E; Chiron Vision, Irvine, CA) to create the flap, and the Technolas 217 C-Lasik excimer laser (Chiron Technolas GmbH, Dornach, Germany) equipped with the PlanoScan program (version 2.998; n 32) or the VISX 20/20 excimer laser (VISX, Santa Clara, CA) equipped with the multi-zone ablation algorithm (version 4.02c; n 30) for photoablation. The two excimer laser groups did not differ from each other with respect to age or ablation depth (P 0.05). The LASIK procedure was performed under topical anesthesia with 0.4% oxybuprocaine. The flap diameter was 8.5 mm and the intended thickness 160 m. Eyes were not occluded after surgery. Antibiotic (tobramycin 0.3%, Tobrex; Alcon Iberhis S.A., Madrid, Spain) and Slit-Lamp Examination and In Vivo Confocal Microscopy Each patient was examined on slit-lamp by two independent ophthalmologists. The findings were presented as drawings in patient charts or photographs. A tandem scanning confocal microscope (TSCM; model 165A; Tandem Scanning, Reston, VA) was used for examining the central cornea of the patients at the Alicante Institute of Ophthalmology, Spain. The setup and operation of the confocal microscope has been described previously. 21,23,24 Briefly, a 24, 0.6 NA variable working distance objective lens was used. The field-of-view with this lens is 450 360 m, and the z-axis resolution is 9 m. Images were detected using a Dage VE1000 low-light level camera and recorded on SVHS tape. In addition, confocal microscopy through-focus scans (CMTF) were obtained as previously described. 21,23 Video images of interest were digitized using a PC-based imaging system with custom software (University of Texas Southwestern Medical Center at Dallas) and printed using an Epson Stylus Color 800 printer (Seiko Epson, Nagano, Japan) without image enhancement. The central cell density of the most anterior keratocyte layer was calculated by hand in the area of 205 190 m and reported as counts per square millimeter (counts/mm 2 ). Similarly, interface particle density was also determined. Interface particles were at most 25 m 2 in size and, thus, remarkably smaller than what was regarded as keratocyte nuclei. Furthermore, interface particles were usually brighter than keratocyte nuclei. Using the custom software, the CMTF data were digitized onto the PC, and intensity profile curves were calculated. 23 From each scan, the flap thickness was measured (defined as the distance between the surface epithelium and the flap interface characterized by accumulation of interface particles), as well as the thickness of the pre- and post-interface acellular zones and post-interface keratocyte activation zone (defined as bright keratocyte nuclei and visible keratocyte processes). A quantitative estimate of the increased backscattering (CMTF-haze) 21 around the flap interface was obtained by calculating the area below the CMTF profile corresponding to peaks originating from the structures of interest (i.e., interface particles, increased extracellular matrix [ECM] reflection, or activated keratocytes). CMTF-haze values were not given for the preoperative corneas, because they did not show any extra peaks deviating from the baseline. An average of three CMTF scans was performed per each eye. In five eyes (one eye at 1 2 weeks, 1 2 months, and 3 months and two eyes at 6 months), an acceptable CMTF profile was not produced because of the patients inability to fixate steadily; these scans were not included in the analysis. Average values of the measurements were used for all statistical calculations. Statistical Analyses Statistical analyses were performed using SPSS for Windows (version 7.0). Normality was tested using Shapiro Wilk test, and ANOVA or nonparametric Kruskal Wallis H and Mann Whitney tests were performed, respectively, for comparison of the groups. Pearson correlation coefficients (r) were used to evaluate the correlations between continuous variables. Data are given as mean SD, and the differences were considered statistically significant when P 0.05.

IOVS, February 2000, Vol. 41, No. 2 Corneal Morphology after Myopic LASIK 371 FIGURE 2. The CMTF intensity profiles of a preoperative cornea (A) and of a cornea 3 days after LASIK with a correction of 11.50 D (spherical equivalent; B). Images corresponding to selected intensity peaks are shown below. The distance from the corneal surface is given in parentheses. (A) a surface epithelium (0 m); b basal epithelial cells (26 m); c branching subbasal nerve fiber bundles (50 m); d the most anterior keratocyte layer (71 m); e more posterior keratocytes (160 m); f endothelial cell layer (583 m). (B) a surface epithelium (0 m, image not shown); b basal epithelial cells (39 m, image not shown); c faint subbasal nerve fiber bundles (arrows; 70 m); d the most anterior keratocyte layer with moderately reflecting keratocyte nuclei (arrows) and small particles (presumably degenerative microdots, arrowhead; 80 m); e the flap interface with some keratocyte nuclei (arrow) and particles of variable sizes (arrowhead; 103 m); f the activated post-interface keratocytes with highly reflective cell nuclei (arrow) and prominent processes (arrowhead; 132 m); g stromal keratocytes (images not shown); and h endothelial cell layer (465 m, image not shown). The CMTF-haze of this cornea is 1274 U. Scale bar, 100 m. RESULTS Biomicroscopy of the Corneas Particles were detected at the flap interface in 24 of 62 eyes (38.7%) by slit-lamp biomicroscopy. Metal particles were observed in 10 eyes (16.1%; Fig. 1A), lipid in 9 (14.5%), diffuse infiltration in 1 (1.6%), and single undefined spots in 4 eyes (6.5%). Fine striae of the Bowman s layer (Fig. 1B) were detected in 25 eyes (40.3%), whereas 3 eyes (4.8%) presented with thicker folds (Fig. 1C). Epithelial ingrowth was observed in 6 eyes (9.7%). One eye (1.6%) developed a small area of flap melting at 6 months postoperatively. Epithelial ingrowth and flap melting were in all cases peripheral disorders, and did not affect visual acuity, or corneal astigmatism. FIGURE 3. Subepithelial microfolds. Folding in the Bowman s layer (A) and extending into the level of the anterior keratocytes (B). Scale bar, 100 m.

372 Vesaluoma et al. IOVS, February 2000, Vol. 41, No. 2 FIGURE 4. Interface particles. (A) The mean density of interface particles per square millimeter at various time points after LASIK. The highest density was calculated at 3 days postoperatively. The black areas represent SDs. Number of patients in each group is given in parentheses. (B) High density of presumably metal particles at corneal interface 3 days after LASIK (flap thickness 97 m). Scale bar, 100 m. Confocal Microscopy Examples of CMTF profiles are shown in Figure 2. Most of the eyes (60/62, 96.8%) presented with microfolding of Bowman s layer by confocal microscopy. In some cases folding appeared as unevenness in Bowman s layer (Fig. 3A), whereas in others the microfolds appeared as more prominent wrinkles, where keratocytes and basal epithelial cells could be identified at the same sagittal corneal level (Fig. 3B). Interspersed particles of variable size and reflectivity were observed at the interfaces of all eyes (Figs. 2B, panel e, and 4). The highest density of particles was calculated at 3 days postoperatively (683 990 particles/mm 2 ), and the lowest at 6 months (135 122 particles/mm 2, Kruskal Wallis H test, P 0.001). The number of particles did not correlate with the ablation depth (Pearson correlation coefficient, r 0.086, P 0.518) or the flap thickness (r 0.043, P 0.753), but a weak negative correlation was found with the time after surgery (r 0.252, P 0.049). It was impossible to define the nature of the particles using confocal microscopy, except in the case of metallic particles, which had an unusually strong light reflection (Fig. 4B). However, it cannot be excluded that they were salt crystals. In some cases these particles appeared scattered on both sides of the flap interface, extending to the level of the most anterior keratocytes, although most of them were located at the interface. We did not identify layers of inflammatory cells in operated corneas, although some of the interface particles could have been inflammatory cells. The intended flap thickness in all eyes was 160 m. However, confocal microscopy revealed that the flap interface was located at 112 25 m (range, 62 165 m) below the surface epithelium (Fig. 5). The thinnest flaps were found at 3 days (95 16 m), and the thickest flaps were observed at 6 months (144 17 m; ANOVA, P 0.001). Soon after LASIK, the keratocytes disappeared from both sides of the lamellar cut (Fig. 6). At 3 days 85.7% (12/14) of the eyes presented with a clear keratocyte-free zone in the stromal bed and 78.5% (11/14) in the pre-interface area. Beyond 3 days keratocytes were generally observed in the immediate proximity of the keratome cut. The most anterior keratocyte layer was chosen for analysis of the keratocyte density in the corneal flap, because the first corneal keratocytes behind the Bowman s layer formed an easily identifiable landmark in every cornea. In preoperative corneas we calculated an average of 1060 468 keratocyte nuclei/mm 2, and the number remained close to this level up to 3 months after surgery, whereas at 6 months the density was approximately 60% of that of the preoperative corneas (603 194 keratocyte nuclei/mm 2 ; ANOVA, P 0.007; Fig. 7). No significant association was found between the anterior keratocyte counts and the flap thickness (r 0.189, P 0.167). The morphology of the first keratocytes observed behind the flap interface was dramatically different in eyes 3 days postoperatively compared with the preoperative eyes (Figs. 2B, panel f, and 8A). The oval and brightly reflecting keratocyte nuclei appeared larger than preoperative nuclei, and the processes could be easily visualized, suggesting that the cells were activated. 21 The keratocytes were stellate and appeared to form an interwoven meshwork. Processes were still detected in 75% (12/16) of the corneas at 1 to 2 weeks postoperatively (Fig. 8B), and in two corneas at 1 to 2 months and 3 months. In general, the post-interface keratocyte processes were no longer distinguishable beyond 2 weeks (Fig. 8C). The thicknesses of the keratocyte activation zones are shown in Figure 8D. The keratocytes anterior to the keratome cut did not show such reactive responses. The keratocytes posterior to the postinterface activation zone also appeared quiet at all time points, and their morphology was not remarkably different from that of normal posterior keratocytes. The CMTF-haze was produced by the increased interface reflectivity from the accumulated particles, activated postinterface keratocytes, or increased ECM. It peaked at 3 days postoperatively (Kruskal Wallis H test, P 0.046; Figs. 2B and FIGURE 5. The mean flap thickness of all patients was 112 25 m (range, 62 165 m), whereas the intended flap thickness was 160 m. The flaps showed the tendency of being thinner at the early time points after LASIK. One patient (at 6 mo) was excluded because the determination of the flap interface was not reliable due to variation in the different scans. The black areas represent SDs. Number of patients in each group is given in parentheses.

IOVS, February 2000, Vol. 41, No. 2 Corneal Morphology after Myopic LASIK 373 FIGURE 6. The acellular zones. The acellular zone bordering into the flap interface was considerably thinner on the flap side (A, pre-interface acellular zones) than on the stromal bed side (B, post-interface acellular zones). The acellular zones were most prominent 3 days after LASIK. The black areas represent SDs. Number of patients in each group is given in parentheses. 9A). The CMTF-haze was positively correlated with the thickness of the keratocyte activation zone (r 0.618, P 0.001). The number of interface particles was not related to the haze estimate (r 0.022, P 0.869). In morphologic evaluation, increased reflectivity from ECM or activated keratocytes (Figs. 9B, 9C) was observed at the central flap interface in 11 of 32 (34.4%) eyes at 1 month. CMTF scans were obtained in 9 of 11 of these eyes. These corneas had higher haze estimates than the corneas without such morphologic changes (421 353 versus 141 107 U; Mann Whitney test, P 0.033). The densities of interface particles in these two patient groups were 181 188 and 278 504 particles/mm 2 (Mann Whitney test, P 0.584). Interestingly, the patients with morphologically increased interface reflectivity caused by abnormal ECM or activated keratocytes had significantly thinner flaps than the patients with morphologically normal stromal keratocyte and matrix reflectivity (114 12 versus 132 22 m; ANOVA, P 0.027). In addition, 4 of 5 patients with flaps thinner than 90 m at 1 to 2 weeks presented with a pronounced interface response (Fig. 9D). The eyes operated with the two different excimer lasers were compared, but no significant differences were observed with respect to density of interface particles, flap thickness, pre- or post-interface acellular zones, keratocyte activation zone, or CMTF haze estimate at any time points (Mann Whitney test, P 0.05). DISCUSSION The tremendous increase in the popularity of LASIK as a method of modern refractive surgery is based on good clinical results in myopia up to 12 D with minimal pain and a short visual recovery time. 2,4,6 11 It is performed worldwide with increasing frequency despite the absence of thorough data on the healing response and long-term complications at the tissue level. Clinically visible complications such as flap folds or striae, accumulation of flap interface debris, bacterial keratitis, or interface abscess, noninfectious interface infiltrates, epithelial ingrowth, and subsequent flap melting are relatively well known, 8,9,25 30 but the underlying cell biology of these phenomena is less well understood. Confocal microscopy revealed microfolds in almost every eye (93.8%). Microfolding appeared in two forms: as wavy unevenness of Bowman s layer or as more prominent folds, which extended into the anterior stroma. Microfolding may result from stretching of the flap during surgery, or from FIGURE 7. The most anterior corneal keratocytes. The preoperative cornea (A) presented with a keratocyte nuclei count of 1620/mm 2 (image 68 m from the corneal surface), and the cornea operated 2 years earlier (B) with 608 nuclei/mm 2 (image depth, 71 m). Scale bar, 100 m. The density of keratocyte nuclei appeared significantly decreased at 6 months after LASIK (C). One patient (at 1 2 weeks) had an extremely thin flap, and the interface was located in the Bowman s layer. Therefore, the most anterior keratocytes could not be counted, and this cornea was excluded. The black areas represent SDs. Number of patients in each group is given in parentheses.

374 Vesaluoma et al. IOVS, February 2000, Vol. 41, No. 2 FIGURE 8. The first post-interface keratocytes. The keratocytes showed signs of activation (brightly reflecting nuclei and prominent processes) at 3 days (A) and 1 week (B) after LASIK, but at 1 month the keratocytes had returned back to dendritic form (C). Scale bar, 100 m. For comparison, see preoperative keratocytes at the depth of 160 m in Figure 1A. The thickness of the post-interface keratocyte activation zone was most prominent at 3 days to 1 to 2 weeks after LASIK (D). One patient (at 3 days) presented with large numbers of panstromal microdots, probably due to previous contact lens wear, and a reliable definition of keratocyte activation zone was impossible. Therefore, this patient was excluded. The black areas represent SDs. Number of patients in each group is given in parentheses. impaired compatibility of the flap to the reformed stromal bed. Slowik et al. reported folds in the flap of a patient with two retreatments. 15 However, none of our patients had undergone retreatments. The clinical significance of slight microfolding appears negligible. However, deeper and more extensive folding might affect the topography of the corneal surface, resulting in irregular astigmatism. The localization of the flap interface was easy because each eye showed reflective interface particles. The potential origin of the material in the flap interface includes sources such as metal from the microkeratome blade, 31 cotton from the swabs, lipids or inflammatory cells from the tear fluid, or epithelial remnants carried to the interface with the microkeratome. The density of interface particles was not associated with increased CMTF-haze or increased reflectivity due to abnormal ECM or prolonged keratocyte activation at the interface. The possible clinical significance of persisting interface particles remains to be studied. Techniques such as rinsing of the flap interface should be improved to eliminate the presence of harmful particles in the interface. Our measurements confirmed the earlier finding, which was based on intraoperative pachymeter recordings, that the flaps are much thinner than expected. 9 At 3 days thin flaps might be explained by a shrinkage of the flap tissue, due to dehydration or retraction of the collagen lamellae. Interestingly, the flaps tended to be thicker at later time points. Whether this reflected a change in corneal hydration, epithelial hyperplasia, or true tissue regeneration as shown after PRK 32 could be assessed in a prospective follow-up study. FIGURE 9. The intensity of the haze estimate (CMTF-haze) was at it strongest at 3 days and 1 to 2 weeks after LASIK, corresponding mostly to the visualization of the activated post-interface keratocytes (A). At 1 month increased reflectivity of the flap interface due to abnormal ECM or activated keratocytes was associated with thin flaps (B and C; flap thicknesses, 116 and 108 m, respectively). At 1 to 2 weeks a strong interface reaction was related with flaps cut at the level of the first anterior keratocytes (D; flap thickness, 87 m). Scale bar, 100 m.

IOVS, February 2000, Vol. 41, No. 2 Corneal Morphology after Myopic LASIK 375 One initial response to LASIK was the creation of thin keratocyte-free zones on both sides of the lamellar cut. Keratocyte death has been described in experimental models after LASIK and PRK. 16,32 34 Apoptosis has been considered as the mechanism underlying the disappearance of keratocytes, and the difference in haze formation after PRK and LASIK may be due to the difference in keratocyte apoptosis triggered by epithelial cytokines. 16,34 The significance of keratocyte apoptosis as an initiator of haze formation was recently questioned, because the haze intensity was shown to correlate positively with the volume of the photoablated tissue rather than the thickness of the keratocyte death zone in rabbits subjected to transepithelial PRK. 33 These keratocyte death zones at 1 week were considerably thicker than the acellular zones in our study. Similar data on human PRK, or rabbit LASIK, are not yet available for comparison. A surprising and novel finding was the apparent loss of cells in the most anterior keratocyte layer beginning at 6 months after surgery. The reason for this, as well as the potential consequences, is unknown. However, a direct innervation of keratocytes by stromal nerve fibers has recently been suggested. 35 During LASIK surgery most of the stromal nerve trunks are cut, and only those in the hinge area are spared, 17 so that most of the keratocytes in the central flap lose their neural input. In fact, the sensitivity of the central cornea is reduced for more than 6 months after LASIK and is lower than that observed after PRK. 36 In rabbits, the recovery of the anterior stromal nerves requires at least 5 months. 17 Based on these data, it can be speculated that lack of communication with the sensory nerves is involved in the loss of the most anterior keratocytes. However, this theory may be inconsistent in that keratocyte loss is not observed until after 6 months, and the innervation is, for the most part, restored by 6 months. Furthermore, because corneal keratocytes are connected by gap junctions, 37 disruption of communication with more posterior keratocytes and the keratocytes surrounding the flap may also affect the integrity of the anterior keratocyte layer. Our findings on the post-interface corneal keratocyte morphology confirmed the profound differences between LASIK and PRK in the extent and duration of the initial keratocyte response. 21 The first changes in keratocyte morphology were characterized by visualization of oval brightly reflective keratocyte nuclei and thick cell processes behind the flap interface by 3 days. These changes in keratocyte reflectivity were still present at 1 to 2 weeks, although the processes became thinner over time. We believe that these cells are activated keratocytes, because after PRK, cells with similar morphology have been shown in the human corneal midstroma at 1 month. 21 However, it should be noted that changes in stromal hydration may also affect the visibility of corneal keratocyte processes. The interpretation of the CMTF-haze was difficult in our study, because deep folding of the Bowman s layer, interface particles, increased ECM reflection, and the activated keratocytes contributed to the backscattering of light. In addition, the CMTF-haze peaks were relatively low in most of the eyes. Morphologically characterized ongoing keratocyte activation or increased reflectivity from the ECM was more frequently associated with thin flaps. Møller Pedersen et al. (1998) have hypothesized that the integrity of the most anterior keratocyte layer may control the myofibroblast transformation and haze formation after refractive surgery. 33 Our results support this hypothesis by suggesting that lamellar cut at the level of the most anterior keratocyte layer predisposes to pronounced keratocyte activation. In addition to a higher cell density, the anterior keratocytes are also morphologically different from the more posterior keratocytes. 21,38,39 Thus, these keratocytes may also present with functional differences, especially during corneal healing. Our present study brings out the following new findings: Microfolding is an almost unavoidable complication of LASIK surgery and serves as a challenge for improvement of the surgical technique and instruments, particles of variable size and reflectivity can always be observed at the flap interface, the keratocytes initially disappear from both sides of the lamellar cut, keratocyte activation after LASIK is visible up to 1 to 2 weeks postoperatively, the integrity of the most anterior keratocyte layer in the flap is jeopardized in corneas 6 months after surgery, and prolonged keratocyte activation or increased reflectivity from abnormal ECM is associated with thin flaps. Acknowledgment We thank Seppo Sarna for his invaluable advice concerning the statistical analysis of the data. References 1. Pallikaris IG, Papatzanaki ME, Stathi EZ, et al. Laser in situ keratomileusis. Lasers Surg Med. 1990;10:463 468. 2. Hersh PS, Brint SF, Maloney RK, et al. Photorefractive keratectomy versus laser in situ keratomileusis for moderate to high myopia. Ophthalmology. 1998;105:1512 1523. 3. Azar DT, Farah SG. Laser in situ keratomileusis versus photorefractive keratectomy: an update on indications and safety. Ophthalmology. 1998;105:1357 1358. 4. Shah S, Chatterjee A, Smith RJ. Predictability of spherical photorefractive keratectomy for myopia. Ophthalmology. 1998;105:2178 2185. 5. Farah SG, Azar DT, Gurdal C, Wong J. Laser in situ keratomileusis: literature review of a developing technique. J Cataract Refract Surg. 1998;24:989 1006. 6. Fiander DC, Tayfour F. Excimer laser in situ keratomileusis in 124 myopic eyes. J Refract Surg. 1995;11(suppl):S234 S238. 7. Salah T, Waring GO III, El Maghraby A, et al. Excimer laser in situ keratomileusis under a corneal flap for myopia of 2 to 20 diopters. Am J Ophthalmol. 1996;121:143 155. 8. Condon PI, Mulhern M, Fulcher T, et al. Laser intrastromal keratomileusis for high myopia and myopic astigmatism. Br J Ophthalmol. 1997;81:199 206. 9. Pérez Santonja JJ, Bellot J, Claramonte P, Ismail MM, Alió JL. Laser in situ keratomileusis to correct high myopia. J Cataract Refract Surg. 1997;23:372 385. 10. Maldonado Bas A, Onnis R. Results of laser in situ keratomileusis in different degrees of myopia. Ophthalmology. 1998;105:606 611. 11. Knorz MC, Wiesinger B, Liermann A, et al. LASIK for moderate and high myopia and myopic astigmatism. Ophthalmology. 1998;105: 932 940. 12. Nagel S, Wiegand W, Thaer AA. Hornhautveränderungen und korneale Heilungsvorgänge nach Keratomileusis in situ: in-vivo-untersuchungen mittels der konfokalen Spalt-Scanning-Mikroskopie. Ophthalmologe. 1995;92:397 401. 13. Latvala T, Barraquer Coll C, Tervo K, Tervo T. Corneal wound healing and nerve morphology after excimer laser in situ keratomileusis (LASIK) in human eyes. J Refract Surg. 1996;12:677 683. 14. Amm M, Wetzel W, Winter M, Uthoff D, Duncker GIW. Histopathological comparison of photorefractive keratectomy and laser in situ keratomileusis. J Refract Surg. 1996;12:758 766. 15. Slowik C, Somodi S, Richter A, Guthoff R. Assessment of corneal alterations following laser in situ keratomileusis by confocal slit scanning microscopy. Ger J Ophthalmol. 1997;5:526 531.

376 Vesaluoma et al. IOVS, February 2000, Vol. 41, No. 2 16. Helena MC, Baerveldt F, Kim W J, Wilson SE. Keratocyte apoptosis after corneal surgery. Invest Ophthalmol Vis Sci. 1998;39:276 283. 17. Linna TU, Pérez Santonja JJ, Tervo KM, Sakla HF, Alió JL, Tervo TMT. Recovery of corneal nerve morphology following laser in situ keratomileusis. Exp Eye Res. 1998;66:755 763. 18. Pérez Santonja JJ, Linna TU, Tervo KM, Sakla HF, Alió JL, Tervo TMT. Corneal wound healing after laser in situ keratomileusis. J Refract Surg. 1998;14:602 609. 19. Cavanagh HD, Petroll WM, Alizadeh H, He Y G, McCulley JP, Jester JV. Clinical and diagnostic use of in vivo confocal microscopy in patients with corneal diseases. Ophthalmology. 1993;100: 1444 1454. 20. Corbett MC, Prydal JI, Verma S, Oliver KM, Pande M, Marshall J. An in vivo investigation of the structures responsible for corneal haze after photorefractive keratectomy and their effect on visual function. Ophthalmology. 1996;103:1366 1380. 21. Møller Pedersen T, Vogel M, Li HF, Petroll WM, Cavanagh HD, Jester JV. Quantification of stromal thinning, epithelial thickness, and corneal haze after photorefractive keratectomy using in vivo confocal microscopy. Ophthalmology. 1997;104:360 368. 22. Linna T, Tervo T. Real-time confocal observations on human corneal nerves and wound healing after excimer laser photorefractive keratectomy. Curr Eye Res. 1997;16:640 649. 23. Li HF, Petroll WM, Møller Pedersen T, Maurer JK, Cavanagh HD, Jester JV. Epithelial and corneal thickness measurements by in vivo confocal microscopy through focusing (CMTF). Curr Eye Res. 1997;16:214 221. 24. Petroll WM, Jester JV, Cavanagh HD. Quantitative 3-dimensional confocal imaging of the cornea in situ and in vivo: system design and calibration. Scanning. 1996;18:45 49. 25. Gimbel HV, Penno EE, van Westenbrugge JA, Ferensowics M, Furlong MT. Incidence and management of intraoperative and early postoperative complications in 1000 consecutive laser in situ keratomileusis cases. Ophthalmology. 1998;105:1839 1847. 26. Knorz MC, Liermann A, Seibert V, Steiner H, Wiesinger B. Laser in situ keratomileusis to correct myopia of 6.00 to 29.00 diopters. J Refract Surg. 1996;12:575 584. 27. Aras C, Ozdamar A, Bahcecioglu H, Sener B. Corneal interface abscess after excimer laser in situ keratomileusis. J Refract Surg. 1998;14:156 157. 28. Smith RJ, Maloney RK. Diffuse lamellar keratitis: a new syndrome in lamellar refractive surgery. Ophthalmology. 1998;105:1721 1726. 29. Helena MC, Meisler D, Wilson SE. Epithelial ingrowth within the lamellar interface after laser in situ keratomileusis (LASIK). Cornea. 1997;16:300 305. 30. Castillo A, Diaz Valle D, Gutierrez AR, Toledano N, Romero F. Peripheral melt of flap after laser in situ keratomileusis. J Refract Surg. 1998;14:61 63. 31. Kaufman SC, Maitchouk DY, Chiou AGY, Beuerman RW. Interface inflammation after lasik in situ keratomileusis: Sands of the Sahara syndrome. J Cataract Refract Surg. 1998;24:1589 1593. 32. Møller Pedersen T, Cavanagh HD, Petroll WM, Jester JV. Neutralizing antibody to TGF modulates stromal fibrosis but not regression of photoablative effect following PRK. Curr Eye Res. 1998; 17:736 747. 33. Møller Pedersen T, Cavanagh HD, Petroll WM, Jester JV. Corneal haze development after PRK is regulated by volume of stromal tissue removal. Cornea. 1998;17:627 639. 34. Wilson SE, Kim WJ. Keratocyte apoptosis: implications on corneal wound healing, tissue organization, and disease. Invest Ophthalmol Vis Sci. 1998;39:220 226. 35. Müller L, Pels L, Vrensen GFJM. Ultrastructural organization of human corneal nerves. Invest Ophthalmol Vis Sci. 1996;37:476 488. 36. Pérez Santonja JJ, Sakla HF, Cardona C, Chipont E, Alió JL. Corneal sensitivity after photorefractive keratectomy (PRK) and laser in situ keratomileusis (LASIK) for low myopia. Am J Ophthalmol. 1999;127:497 504. 37. Watsky MA. Keratocyte gap junctional communication in normal and wounded rabbit corneas and human corneas. Invest Ophthalmol Vis Sci. 1995;36:2568 2576. 38. Müller LJ, Pels L, Vrensen GFJM. Novel aspects of the ultrastructural organization of human corneal keratocytes. Invest Ophthalmol Vis Sci. 1995;36:2557 2567. 39. Poole CA, Brookes NH, Clover GM. Keratocyte networks visualised in the living cornea using vital dyes. J Cell Sci. 1993;106: 685 691.