Keratocyte density of central human cornea after laser in situ keratomileusis

Keratocyte Density of Central Human Cornea After Laser In Situ Keratomileusis KATSUYA MITOOKA, MD, MANUEL RAMIREZ, MD, LEO J. MAGUIRE, MD, JAY C. ERIE, MD, SANJAY V. PATEL, MD, JAY W. MCLAREN, PHD, DAVID O. HODGE, MS, AND WILLIAM M. BOURNE, MD

● PURPOSE:

To determine changes in keratocyte density in the first year after laser in situ keratomileusis (LASIK). ● DESIGN: Prospective interventional cohort study. ● METHODS: Seventeen eyes of 11 patients received LASIK with a planned 180-␮m flap to correct refractive errors between ⴚ2.0 diopters and ⴚ11.0 diopters. Images of the full-thickness cornea were obtained by using confocal microscopy in vivo before LASIK and at 1 week, 1, 3, 6, and 12 months after LASIK. Bright objects (that resembled keratocytes) in images without motion blur were manually counted by one observer. Cell densities were determined in anterior and posterior halves of the stromal flap, anterior and posterior halves of the layer 100 ␮m-thick immediately deep to the ablation (retroablation layer), and in the posterior third of the stroma. The region of stroma that was ablated (as measured 1 month after LASIK) was omitted from the preoperative analysis. ● RESULTS: Keratocyte density in the anterior flap was 28,978 ⴞ 5849 cells/mm3 (mean ⴞ SD) pre-LASIK, and was decreased at all postoperative examinations, but the difference was not significant until 12 months after LASIK (22% decrease). Keratocyte densities in the posterior flap were 20,397 ⴞ 4215 cells/mm3 pre-LASIK and were decreased by 20%– 40% at all postoperative examinations 1 week to 1 year after LASIK. Keratocyte densities in the anterior half of the retroablation layer were 16,605 ⴞ 3595 cells/mm3 pre-LASIK and decreased by 16%–30% between 3 and 12 months after LASIK. Keratocyte densities in the posterior half of the retroablation layer and posterior stroma did not change. Accepted for publication Dec 13, 2001. From the Department of Ophthalmology (K.M., M.R., L.J.M., J.C.E., S.V.P., J.W.M., W.M.B.) and Section of Biostatistics (D.O.H.) Mayo Clinic, Rochester, Minnesota. Supported in part by NIH Grant EY02037, Research to Prevent Blindness, Inc., New York, NY, and Mayo Foundation, Rochester, MN. Proprietary interests: none. Reprint requests to William M. Bourne, MD, Mayo Clinic, 200 1st Street SW, Rochester, MN, 55905; fax: 507-284-8566; e-mail: bourne@ mayo.edu 0002-9394/02/$22.00 PII S0002-9394(01)01421-0

©

2002 BY

● CONCLUSIONS:

Keratocyte densities in the posterior flap and anterior retroablation layer (regions adjacent to the lamellar cut) decrease at 1 week and 3 months, respectively, after LASIK and remain decreased in these regions at 12 months after LASIK. In the anterior flap, keratocyte density decreases 1 year after LASIK. The long-term effects of these cellular deficits, if any, require further study. (Am J Ophthalmol 2002;133: 307–314. © 2002 by Elsevier Science Inc. All rights reserved.)

A

FTER PHOTOREFRACTIVE KERATECTOMY (PRK), A

superficial central corneal scar (haze) develops in many patients.1–3 This haze is associated with the activation of keratocytes in the anterior corneal stroma after treatment.4 –13 After laser in situ keratomileusis (LASIK), however, corneal haze is extremely rare, possibly because the Bowman layer is preserved centrally or because the LASIK treatment is 130 –180 ␮m deep to the epithelial surface. This deeper treatment does not disrupt the molecular and cellular interactions between corneal epithelial cells and stromal keratocytes as PRK does.14 The deep keratocytes in human corneas treated by LASIK have somewhat different morphologic and histologic features than the superficial keratocytes treated by PRK and may respond differently to the healing process.2,15–16 In fact, histologic studies have failed to find activated keratocytes after LASIK in rabbits,10,12,13 although cells resembling activated keratocytes were present in one human cornea removed 8 days after LASIK.17 Of more concern, however, is the apparent decrease in keratocytes in the anterior flap more than 6 months after LASIK in humans, as Vesaluoma and associates 18 discovered by using confocal microscopy in a cross-sectional study (each eye examined only once). In the present study, we examined keratocyte density in the flap and throughout the stroma before and at intervals after LASIK for up to 1 year. This longitudinal design allowed us to determine changes within subjects over time in keratocyte morphology, keratocyte density, interface particle density, and subbasal nerve fiber bundle morphology after LASIK.

ELSEVIER SCIENCE INC. ALL

RIGHTS RESERVED.

307

thelium. Digital images were stored on a computer workstation (Indy, Silicon Graphics Inc., Mountain View, CA) at 30 frames/s. Each image represented a coronal section approximately 475 ␮m ⫻ 350 ␮m (horizontal ⫻ vertical) and was separated from adjacent images by approximately 2.6 ␮m.19 A full-thickness scan required 7–9 seconds to complete. The video camera (VE-1000 SIT; Dage-MTI, Michigan City, IN) acquired images in one of two modes, a fixed-gain mode in which the video gain, voltage, and black level were constant and an automatic-gain mode in which these parameters were automatically adjusted by the camera to maintain optimal exposure throughout the acquisition. Four to eight full-thickness scans were made (two to four in each mode) and the scan in each mode with the least transverse movement and with no apparent anterior–posterior movement of the objective relative to the cornea was used for analysis.

METHODS ● DESIGN:

We conducted a prospective interventional cohort study in which patients undergoing LASIK procedures gave informed consent to have confocal microscopic examinations before and at intervals after their operations. Our Institutional Review Board approved the study.

● PATIENTS:

Seventeen eyes of eleven patients were studied (the other 5 eyes in these patients underwent reoperations for undercorrections and were excluded). All had normal anterior ocular segments, intraocular pressures (ⱕ22 mm Hg), and fundi. Contact lens wear was discontinued 2 weeks (soft lenses) or 3 weeks (hard lenses) before the operation. Patients were excluded if they had diabetes mellitus, glaucoma, or currently used any topical ocular medication. Mean patient age was 33.2 ⫾ 9.2 years (mean ⫾ SD, range 20 – 46 years); there were 10 females and 1 male. Preoperative spheroequivalent refraction was ⫺ 6.56 ⫾ 2.44 diopters (range, ⫺2.00 to ⫺11.00 diopters) and planned ablation depth was 62.8 ⫾ 26.2 ␮m (range, 18 –110 ␮m). No patient wore contact lenses after LASIK.

● KERATOCYTE MORPHOLOGY:

The appearance of keratocytes after LASIK was compared with their appearance before LASIK. “Activated” keratocytes or repair fibroblasts20 were defined as those bright enlarged objects (assumed to be keratocytes) in which the cell bodies and portions of the processes could be seen as well as the nuclei, which are normally the only visible portions of the cells. The presence of activated keratocytes was noted and the thickness of stroma that contained activated keratocytes was determined from the number of video frames that contained the activated keratocytes.

● SURGICAL PROCEDURE AND POSTOPERATIVE TREATMENT: All eyes were treated for myopia or myopic astig-

matism with a VISX Star excimer laser (VISX, Santa Clara, CA). A flap with a planned thickness of 180 ␮m was created by using a Hansatome microkeratome (Chiron Vision Corporation, Claremont, CA). Patients fixated on a target during the ablation. The stromal bed was irrigated with balanced salt solution at room temperature before and after flap replacement to eliminate residual debris. The flap was allowed to dry in place for at least 3 minutes to facilitate its adhesion at the end of the operation. After LASIK, topical medications consisted of fluorometholone 0.1% (FML, Allergan Inc., Irvine, CA) 4 times/day for 1 week and tapered over 2 weeks and ofloxacin 0.3% (Ocuflox, Allergan Inc., Irvine, CA) 4 times/day for 5 days.

● STROMAL LAYERS:

Keratocyte density was estimated in six layers through the depth of the stroma (Figure 1): (1) anterior half of the stromal flap, (2) posterior half of the stromal flap, (3) anterior half of the layer 100 ␮m-thick immediately behind the ablation (retroablation layer), (4) posterior half of the retroablation layer, (5) posterior 66%–90% of the stroma, and (6) posterior 91%–100% of the stroma. The flap interface was identified as described under interface particle density measurement. The thicknesses of the stromal flap (distance from the most anterior keratocyte to the flap interface) and the stromal bed (distance from the flap interface to the endothelium) determined from the 1-month post-LASIK scan were used to delimit the corresponding anterior and posterior stromal layers in the preoperative cornea. This left a stromal gap between the flap and bed in the preoperative cornea. This gap represented the ablated zone for that cornea and was not analyzed. After LASIK, the posterior 66%–90% and 91%–100% of stroma were determined from the preLASIK stromal thickness (distance between the most anterior keratocytes and the endothelium). We selected two images with the least motion blur from each layer from images recorded with the camera in the automatic-gain mode. Bright objects (that resembled keratocyte nuclei) in the image were manually counted by one masked observer to calculate keratocyte density (cells/mm3).19 The images

● CONFOCAL MICROSCOPY:

A tandem scanning confocal microscope (Tandem Scanning Corporation, Reston, VA) was used to examine corneas preoperatively and at 1 week, 1, 3, 6 and 12 months after LASIK by using a technique described previously.19 The microscope objective lens was disinfected with 70% isopropyl alcohol wipes before and after the examination on every eye. A drop of 2.5% hydroxypropyl methylcellulose (Goniosol, CIBA Vision Ophthalmics, Atlanta, GA) was placed on the tip of the objective lens as an optical coupling medium and the lens was manually advanced until the medium contacted the surface of the central cornea. A full-thickness scan, consisting of a series of confocal images, was recorded as the focal plane was advanced at approximately 78 ␮m/s from anterior to the epithelium to posterior to the endo-

308

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

MARCH 2002

FIGURE 1. Schematic representation of anteroposterior stromal layers. Six layers of the stroma after LASIK were compared with the corresponding layers before LASIK. (1) Anterior half of the stromal flap. (2) Posterior half of the stromal flap. (3) Anterior half of the retroablation layer (100 ␮m-thick layer immediately behind the ablation). (4) Posterior half of the retroablation layer. (5) Posterior 66% to 90% of the stroma. (6) Posterior 91% to 100% of the stroma.

were presented in random order to the observer for evaluation. Mean cell density in each layer after LASIK was compared with mean cell density in the corresponding pre-LASIK layer. The full-thickness density was estimated by multiplying the number of cells in a full-thickness stromal column of 1 mm2 cross-sectional area by the stromal thickness. The number of cells in the column was estimated from the mean density in each analyzed frame weighted by the distance between frames.19 ● IMAGE CONTRAST:

Image contrast, the difference between object and background intensities expressed as a percentage of the background intensity, was calculated. Mean intensities of objects and background were determined by using an algorithm that automatically identified boundaries of objects in the same nonblurred images that were used to estimate density manually.21

● INTERFACE IDENTIFICATION AND INTERFACE PARTICLE DENSITY: Small bright objects (presumably metal particles) were visible in the anterior stroma in 78 of the 85 post-LASIK scans used for cell analysis. These objects were used to identify the interface in postoperative corneas. In the seven scans without visible interface particles, an abrupt change in intensity between adjacent images, usually associated with a peak on the intensity diagram, was noted and assumed to represent the interface. Particles were manually counted by one observer in confocal images of the interface. Density was expressed as particles/mm2.

VOL. 133, NO. 3

FIGURE 2. Confocal images of subbasal nerve fiber bundles (NFBs). (Top left) group 1 (no nerves). White arrow shows microfold of the Bowman layer, which is easily differentiated from a nerve by viewing adjacent images in the scan. (Top right) group 2 (only short [< 200 ␮m] NFBs without interconnections). White arrow shows nerve. (Bottom left) group 3 (long NFBs [> 200 ␮m] without interconnections). (Bottom right) group 4 (long NFBs with interconnections).

FIGURE 3. Confocal images of the retroablation layer of the same eye before and after LASIK. At 1 week (center), keratocytes had bright nuclei and visible cell processes (activated keratocytes). At 3 months (right), there were no activated keratocytes and the appearance was similar to that before LASIK (left).

● SUBBASAL NERVE FIBER MORPHOLOGY:

Confocal images at each corneal examination were grouped into four different categories, as devised by Linna and associates,22 according to the appearance of subbasal nerve fiber bundles (NFBs): (1) no nerves, (2) only short (⬍ 200 ␮m) NFBs without interconnections, (3) long NFBs (ⱖ 200 ␮m) without interconnections, (4) long NFBs with interconnections (Figure 2). We used images from all confocal scans for characterizing NFBs and for determining the presence of microfolds in the Bowman layer. ● STATISTICAL ANALYSIS:

Keratocyte densities in the various stromal layers were compared by using a one-factor repeated measures analysis of variance for normally distributed data or the Friedman test for non-normally distributed data. Significant differences were adjusted for multiple comparisons by using the Student–Newman–Keuls procedure. The Spearman correlation coefficient (rs) was used to evaluate the relationship between the planned ablation depth and the thickness of stroma that contained activated keratocytes. To take into account the potential correlation

KERATOCYTE DENSITY AFTER LASIK

309

TABLE 1. Keratocyte Density Density (Mean ⫾ SD, n ⫽ 17) Stromal Layer

Pre-op

Anterior flap 28,978 ⫾ 5849 Posterior flap 20,397 ⫾ 4215 Anterior retroablation 16,605 ⫾ 3595 layer (0–50 ␮m) Posterior retroablation layer 15,606 ⫾ 4690 (50–100 ␮m) Posterior 66%–90% 17,824 ⫾ 3705 Posterior 91%–100% 17,384 ⫾ 4587 Full thickness 18,336 ⫾ 4277

1 Week

1 Month

3 Months

6 Months

12 Months

P Value*

23,342 ⫾ 4001 28,555 ⫾ 7874 24,848 ⫾ 5928 27,049 ⫾ 6430 22,563 ⫾ 5449 0.02†# 16,182 ⫾ 3171 16,351 ⫾ 3719 13,626 ⫾ 4752 16,114 ⫾ 5337 12,272 ⫾ 2533 ⬍0.001‡ 14,879 ⫾ 3929 15,708 ⫾ 3222 12,424 ⫾ 3211† 13,914 ⫾ 3996 11,629 ⫾ 3435 ⬍0.001
16,994 ⫾ 2878 17,350 ⫾ 3406 18,196 ⫾ 2826

17,028 ⫾ 2020 15,657 ⫾ 3079

0.25†

17,113 ⫾ 4243 17,993 ⫾ 2755 17,773 ⫾ 3775 16,842 ⫾ 4512 18,010 ⫾ 3222 16,876 ⫾ 4857 17,418 ⫾ 2671 18,524 ⫾ 2455 17,467 ⫾ 2450

18,027 ⫾ 2454 16,944 ⫾ 3543 18,281 ⫾ 3516 17,231 ⫾ 3427 18,095 ⫾ 2367 16,065 ⫾ 2112

0.83 0.73 0.62

*One factor repeated measures analysis of variance, except as noted. Significant differences are adjusted for multiple comparisons using the Student–Newman–Keuls procedure. † Friedman’s test. # Pre-op is significantly different from 12 months. ‡ Pre-op is significantly different from all other time points. One year is significantly different from 1 week, 1 month, and 6 months.
Pre-op and 1 month are significantly different from 3 months and 1 year. One week is significantly different from 1 year.

between eyes from the same patient, comparisons were estimated using generalized estimated equation (GEE) models. The GEE results were very similar to the standard tests; therefore, only the results of the standard tests were reported. A probability of less than 0.05 was considered statistically significant.

tively, and remained below preoperative densities at 12 months. Keratocyte densities in the posterior half of the retroablation layer and posterior stroma did not change. Image contrast 12 months after LASIK was not decreased in the full-thickness cornea or in any of the anteroposterior layers compared with preoperatively (Table 2). After LASIK, the layers on either side of the interface had the lowest keratocyte density as compared with other layers (Figure 4). Figure 4 also demonstrates that keratocyte density in the flap and anterior half of the retroablation layer tended to decrease with time after LASIK, with the lowest values recorded at 12 months.

RESULTS ● KERATOCYTE MORPHOLOGY:

After LASIK, keratocyte nuclei were visible as bright oval objects without visible processes, similar to the pre-LASIK appearance (Figure 3, left). The only exception was in the anterior portion of the retroablation layer, where some of the keratocytes showed bright nuclei and visible cell processes (Figure 3, middle). These highly visible cells, presumably activated keratocytes,1,3,6 – 8,18,20,23 were present in 11 of 17 corneas (64.7%) at 1 week and 1 of 17 corneas (5.9%) at 1 month. When present at 1 week, these cells appeared immediately posterior to the interface in a layer that extended posteriorly 22.7 ⫾ 13.0 ␮m (range 5.2– 46.8 ␮m). We found no correlation between the planned ablation and the depth of activated keratocytes at 1 week post LASIK (rp⫽0.27, P ⫽ .28).

● INTERFACE PARTICLES:

The density of the interface particles (Figure 5) was 19.8 ⫾ 24.6/mm2 (range 0 –90.2/ mm2) at 1 week and did not change significantly thereafter. ● SUBEPITHELIAL NERVE PLEXUS:

The subepithelial nerve plexus was not visible in 11 of 17 corneas (64.7%) at 1 week, but was visible in all eyes at 6 and 12 months. At 12 months, we found long nerve fiber bundles with interconnections in more than half of corneas (Table 3). ● MICROFOLDS OF THE BOWMAN LAYER:

Microfolds of the Bowman layer were visible in 16 of 17 corneas (94%) at 1 week, although the number with microfolds decreased thereafter. At 12 months, 8 of 17 corneas (47%) showed microfolding.

● KERATOCYTE DENSITY:

Keratocyte densities in the full thickness stroma and in the various layers before and after LASIK are given in Table 1. All 17 corneas were examined preoperatively and at each postoperative interval. Full thickness keratocyte density did not change. Keratocyte densities in the anterior flap decreased at 12 months. Keratocyte densities in the posterior flap and anterior half of the retroablation layer decreased immediately after LASIK and at 3 months after LASIK, respec-

310

AMERICAN JOURNAL

DISCUSSION THIS LONGITUDINAL STUDY CONFIRMS THE FINDING BY

Vesaluoma and associates18 of decreased keratocyte density OF

OPHTHALMOLOGY

MARCH 2002

TABLE 2. Image Contrast Contrast (%, Mean ⫾ SD, n ⫽ 17)

Stromal Layer

Pre-op

12 Months Post-op

P*

Anterior flap Posterior flap RAZ (0–50 ␮m) RAZ (50–100 ␮m) Posterior (66%–99%) Posterior (90%–100%)

11.1 ⫾ 3.0 9.5 ⫾ 2.5 7.6 ⫾ 1.7 7.3 ⫾ 1.7 8.7 ⫾ 1.7 9.6 ⫾ 3.2

10.2 ⫾ 1.3 9.8 ⫾ 1.4 9.8 ⫾ 1.6 8.8 ⫾ 1.2 9.4 ⫾ 1.5 9.9 ⫾ 1.7

0.28 0.80 ⬍0.01 0.02 0.12 0.78

*Paired t test.

FIGURE 4. Mean keratocyte density (12 sublayers) vs. stromal depth (n ⴝ 17). After LASIK, keratocyte density was lowest in the sublayers near the interface (2b and 3a) as compared with other layers. Image contrast was not different near the interface as compared with other layers. The six stromal layers (see Figure 1) and 12 stromal sublayers were as follows: (1) Anterior flap, (1a) first frame with countable cells, (1b) 25%–50% of flap. (2) Posterior flap, (2a) 50%–75% of flap, (2b) most posterior frame of flap with countable cells. (3) Anterior half of retroablation layer, (3a) most anterior frame with countable cells, (3b) 25–50 ␮m posterior to interface. (4) Posterior half of retroablation layer, (4a) 50 –75 ␮m from interface, (4b) 75–100 ␮m from interface. (5) 66%–90% of stromal thickness, (5a) 66%–78%, (5b) 79%–90%. (6) 91%– 100% of stromal thickness, (6a) 91%–95%, (6b) last frame with countable cells.

in the anterior flap more than 6 months after LASIK. In 9 corneas from 6 months to 2 years after LASIK, they noted cell densities in the most anterior layer of keratocytes that were 43% less than those in 6 preoperative corneas, whereas we noted a decrease of 21% in 17 corneas from preoperative to 1 year postoperative (Figure 3, layer 1a). It is not known why cell density decreases, although Vesaluoma and associates18 suggested that denervation of the flap may play a role. The nerves that supply the flap, except those in the hinge, are cut by the microkeratome during the LASIK procedure.24 Observations by Mu¨ ller and assoVOL. 133, NO. 3

FIGURE 5. Confocal images of the interface of the same eye before and after LASIK. There were no interface particles pre-LASIK (left). After LASIK (1 week, center; 12 months, right), interface particles were visible (white arrows). The density of the interface particles did not change between 1 week and 12 months.

ciates 25 point to direct innervation of individual keratocytes and the lack of this trophic influence may affect the most anterior keratocytes and decrease their density. Conversely, cell density in the anterior flap was significantly diminished only after 6 months when the nerves in the flap had regenerated.22 The remarkable decrease in keratocyte density on both sides of the lamellar cut after LASIK is consistent with, although not as dramatic as, the acellular zone after LASIK that has been reported in animals12 and in humans.18 Cellular damage in the retroablation zone posterior to the lamellar interface might be mediated in part by photoacoustic effects,26 secondary radiation,27 or hydroxyl radical formation.28 In experimental models of LASIK and PRK, cells died by apoptosis induced by cytokines such as Interleukin-1.14 Indeed, apoptotic cells have been noted on each side of the lamellar cut after LASIK in rabbits.29 Our results seem to indicate, in humans at least, that the keratocyte deficit that results after stromal injury may not be fully reconstituted by cell division or migration. This finding is consistent with the decreased keratocyte densities found in transplanted human corneas.30 This long-term decrease in cell density is different from the transient decrease followed by a later increase reported after PRK.3,5,7,12 This difference in response may be related to the different regions of the stroma that are ablated, the anterior stroma in PRK, and the mid stroma in LASIK. Keratocytes from these regions have somewhat different cellular morphology2,15,31 and numbers of mitochondria,15 in addition to producing different ratios of keratan sulfate to dermatan sulfate.32 Møller-Pedersen and associates have hypothesized that the integrity of the first layer of keratocytes may be responsible for the development of fibrosis, myofibroblast transfomation, and haze formation after refractive surgery.11 Different attributes of the superficial and deep keratocytes may give the keratocytes that remain after ablation by PRK somewhat different recovery properties than those that remain after ablation by LASIK. Indeed, stronger stromal reaction and more stromal regeneration has been noted in rabbits after PRK than after LASIK.10,12,13 At 1 week, many cells with prominent processes were visible in the anterior half of the retroablation layer and we

KERATOCYTE DENSITY AFTER LASIK

311

TABLE 3. Number of Corneas With Subbasal Nerve Fiber Bundles (NFBs) Time After LASIK Group

Pre-op

1 Week

1 Month

3 Months

6 Months

12 Months

No NFBs Short NFBs Long NFBs Long interconnected NFBs

0 0 0 17

11 3 2 1

10 1 3 3

7 3 3 4

0 4 7 6

0 4 3 10

assumed that these were activated keratocytes.1,3,6 – 8,18,20,23 This finding is consistent with a histopathologic study of a human cornea that showed activated keratocytes near the interface.17 Such activated keratocytes have been associated with the healing process in primates after PRK.6,8 The incidence of activation diminished after 1 week, a duration considerably shorter than activation after PRK, which typically peaks from 3 weeks to 4 months.1,3,5,7,9,33 To our surprise, we saw more activated keratocytes after LASIK than we saw in a similar series of corneas after PRK.34 In both of these studies, PRK and LASIK were performed with the same excimer laser and observed by using the same confocal microscope. However, the stroma was cooled immediately after PRK by rinsing with cold balanced salt solution while the stroma was not rinsed after LASIK. This difference may have affected the subsequent keratocyte activation. This strong activation after LASIK may also be related to the use of corticosteroids and a difference in their effect in the two situations. Intense local or systemic corticosteroid treatment is known to reduce the numbers of fibroblasts around corneal wounds, to reduce the deposit of connective tissue, and to delay wound healing.33 After PRK, corticosteroids were applied directly to the stromal wound because there was no epithelium for several days. After LASIK, the stromal flap covered the wound and access to the injured tissue by the corticosteroids was restricted, perhaps weakening the response. Also, keratocytes in the different regions of ablation, superficial after PRK, and deeper stroma after LASIK, may respond differently to laser ablation or to corticosteroid treatment. Subepithelial nerves were visible in all of our patients by 6 months after LASIK. This is consistent with the work of others who have found corneal sensitivity to be normal by 6 months after LASIK.22,35 The nerves that were visible early after LASIK (Table 2) may have been those that entered the flap through the superior hinge and were not damaged during the procedure. In general, the pattern of postoperative innervation that we found is similar to that reported by Linna and associates22 in their cross-sectional study. In two other cross-sectional confocal studies, however, nerves were not seen in the central flap until 8 weeks36 and 4 months37 after LASIK. In the former study, however, corneas were not examined before 8 weeks postoperatively.36 Neither study gave details of the LASIK 312

AMERICAN JOURNAL

procedure or whether a hinged flap was used or its position, which may affect the persistence of viable nerves. Bright particles have consistently been visible by confocal microscopy after LASIK in human corneas,18 although the nature of this interface debris remains unknown. Kaufman and associates described bright particles in rabbit corneas after LASIK and attributed them to surface debris on the microkeratome blades, perhaps sterile wax or an oil-like material. They also related the debris to the Sands of the Sahara Syndrome (diffuse lamellar keratitis).38 Hirst and associates attributed interface deposits after LASIK to exposure of the methylcellulose sponges to the excimer laser beam during the stromal ablation,39 which did not occur in the procedures in the present study. We observed fewer particles than Vesaluoma and associates18 and we did not observe a decrease in particle numbers with time as they did. The significance of microfolds in the Bowman layer in 94% of our corneas at 1 week after LASIK is not clear. Their appearance diminished with time and by 1 year after LASIK, they were visible in only 47% of corneas. Vesaluoma and associates18 observed microfolds in 97% (60 of 62) of patients from 3 days to 2 years after LASIK and they did not find a decrease with time. We did not specifically examine the Bowman layer, but we relied solely upon the 4 – 8 full-thickness scans; microfolds may have been present in some corneas in areas of the Bowman layer outside the region of the scans. Nevertheless, the chance appearance of microfolds in these scans decreased with time after LASIK and it is reasonable to conclude that the total number of microfolds in the central corneas decreased with time also. In summary, we have demonstrated a decrease in cell density in the posterior half of the stromal flap and in the anterior half of the retroablation layer after LASIK, layers adjacent to the cut. These decreases are evidence that keratocytes lost from stromal injury are not completely replaced in human corneas. In addition, the keratocyte density in the anterior flap was decreased 12 months after LASIK. This finding may be related to changes in trophic factors from the diminished innervation, although it appeared well after we noted return of nerves to this region in most eyes. Although the cell density at 12 months was diminished as much as 40% from preoperative densities in some regions, the clinical significance of this change is OF

OPHTHALMOLOGY

MARCH 2002

unknown. The long-term effects of LASIK on keratocyte density and the clinical importance of any changes on the cornea will be determined only after longer longitudinal studies.

REFERENCES 1. 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. 2. 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. 3. Frueh BE, Cadez R, Bohnke M. In vivo confocal microscopy after photorefractive keratectomy in humans. A prospective, long-term study. Arch Ophthalmol 1998;116:1425–1431. 4. Gaster R, Binder P, Coalwell K, Berns M, McCord R, Burstein N. Corneal surface ablation by 193 nm excimer laser and wound healing in rabbits. Invest Ophthalmol Vis Sci 1989:30:90 –98. 5. Hanna KD, Pouliquen Y, Waring GO, Savoldelli M, Cotter J, Morton K, Menasche M. Corneal stromal wound healing in rabbits after 193-nm excimer laser surface ablation. Arch Ophthalmol 1989:107:895–901. 6. Del Pero RA, Gigstad JE, Roberts AD, Klintworth GK, Martin CA, L’Esperance FA, Taylor DM. A refractive and histopathologic study of excimer laser keratectomy in primates. Am J Ophthalmol 1990;109:419 – 429. 7. Fantes FE, Hanna KD, Waring GO, Pouliquen Y, Thompson KP, Savoldelli M. Wound healing after excimer laser keratomileusis (photorefractive keratectomy) in monkeys. Arch Ophthalmol 1990;108:665– 675. 8. Wu WC, Stark WJ, Green WR. Corneal wound healing after 193-nm excimer laser keratectomy. Arch Ophthalmol 1991: 109:1426 –1432. 9. Beuerman R, McDonald M, Shofner R, et al. Quantitative histological studies of primate corneas after excimer laser photorefractive keratectomy. Arch Ophthalmol 1994;112: 1103–1110. 10. Amm M, Wetzel W, Winter M, Uthoff D, Duncker GI. Histopathological comparison of photorefractive keratectomy and laser in situ keratomileusis in rabbits. J Refract Surg 1996;12:758 –766. 11. 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. 12. Wachtlin J, Langenbeck K, Schru¨ nder S, Zhang E-P, Hoffmann F. Immunohistology of corneal wound healing after photorefractive keratectomy and laser in situ keratomileusis. J Refract Surg 1999;15:451– 458. 13. Park CK, Kim JH. Comparison of wound healing after photorefractive keratectomy and laser in situ keratomileusis in rabbits. J Cataract Refract Surg 1999;25:842– 850. 14. Wilson SE. Molecular cell biology for the refractive corneal surgeon: programmed cell death and wound healing. J Refract Surg 1997;13:171–175. 15. Mu¨ ller LJ, Pels L, Vrensen GFJM. Novel aspects of the ultrastructural organization of human corneal keratocytes. Invest Ophthalmol Vis Sci 1995;36:2557–2567.

VOL. 133, NO. 3

16. Hahnel C, Somodi S, eiss DG, Guthoff RF. The keratocyte network of human cornea: a three-dimensional study using confocal laser scanning fluorescence microscopy. Cornea 2000;19:185–193. 17. Latvala T, Barraquer-Coll C, Tervo K, Tervo T. Corneal wound healing and nerve morphology after excimer laser in situ keratomileusis in human eyes. J Refract Surg 1996;12: 677– 683. 18. Vesaluoma M, Pe´ rez-Santonja J, Petroll WM, Linna T, Alio´ J, Tervo T. Corneal Stromal Changes Induced by Myopic LASIK. Invest Ophthalmol Vis Sci 2000;41:369 –376. 19. Patel SV, McLaren JW, Hodge DO, Bourne WM. Normal human keratocyte density and corneal thickness measurement by using confocal microcopy in vivo. Invest Ophthalmol Vis Sci 2001;42:333–339. 20. Fini ME. Keratocyte and fibroblast phenotypes in the repairing cornea. Prog Retinal Eye Res 1999;18:529 –551. 21. Patel SV, McLaren JW, Camp JJ, Nelson LR, Bourne WM. Automated quantification of keratocyte density by using confocal microscopy in vivo. Invest Ophthalmol Vis Sci 1999;40:320 –326. 22. Linna TU, Vesaluoma MH, Pe´ rez-Santonja JJ, Petroll WM, Alio´ JL, Tervo TMT. Effect of myopic LASIK on corneal sensitivity and morphology of subbasal nerves. Invest Ophthalmol Vis Sci 2000;41:393–397. 23. Møller-Pedersen T, Li HF, Petroll WM, Cavanagh HD, Jester JV. Confocal microscopic characterization of wound repair after photorefractive keratectomy. Invest Ophthalmol Vis Sci 1998;39:487–501. 24. Linna TU, Pe´ rez-Santonja JJ, Tervo KM, Sakla HF, Alio´ JL, Tervo TMT. Recovery of corneal nerve morphology following laser in situ keratomileusis. Exp Eye Res 1998;66:755– 763. 25. Mu¨ ller I, Pels L, Vrensen GFJM. Ultrastructural organization of human corneal nerves. Invest Ophthalmol Vis Sci 1996; 37:476 – 488. 26. Kermani O, Lubatschowski H. Structure and dynamics of laser-induced acoustic shockwaves in 193 nm excimer laser photoablation of the cornea. Fortschr Ophthalmol 1991:88: 748 –753. 27. Lubatschowski H, Kermani O, Otten C, Haller A, Schmiedt KC, Ertmer W. ArF-excimer laser-induced secondary radiation in photoablation of biological tissue. Lasers Surg Med 1994;14:168 –177. 28. Shimmura S, Masumizu T, Nakai Y, et al. Excimer laserinduced hydroxyl radical formation and keratocyte death in vitro. Invest Ophthalmol Vis Sci 1999;40:1245–1249. 29. Helena MC, Baerveldt F, Kim WJ, Wilson SE. Keratocyte apoptosis after corneal surgery. Invest Ophthalmol Vis Sci 1998;39:276 –283. 30. Bourne WM. Cellular changes in transplanted human corneas. Cornea 2001;20:560 –569. 31. Poole CA, Brookes NH, Clover GM. Confocal imaging of the keratocyte network in porcine cornea using the fixable dye 5-chloromethylfluorescein diacetate. Curr Eye Res 1996; 15:165–174. 32. Castoro JA, Bettelheim AA, Bettelheim FA. Water gradients across bovine cornea. Invest Ophthalmol Vis Sci 1988; 29:963–968. 33. Tuft SJ, Zabel RW, Marshall J. Corneal repair following keratectomy; comparison between conventional surgery and laser photoablation. Invest Ophthalmol Vis Sci 1989;30: 1769 –1777. 34. McLaren JW, Erie JC, Patel SV, Maguire LJ, Ramirez M, Bourne WM. Keratocyte density 1 day, 5 days, and 1 month after photorefractive keratectomy (PRK). Invest Ophthalmol Vis Sci (ARVO Abstract). 1999;40:S108.

KERATOCYTE DENSITY AFTER LASIK

313

corneal alterations following laser in situ keratomileusis by confocal slit scanning microscopy. German J Ophthalmol 1997;5:526 –531. 38. Kaufman SC, Maitchouk DY, Chiou AG and Beuerman RW. Interface inflammation after laser in situ keratomileusis. Sands of the Sahara syndrome. J Cataract Refract Surg 1998;24:1589 –1593. 39. Hirst LW, Vandeleur Jr KW. Laser in situ keratomileusis interface deposits. J Refract Surg 1998;14:653– 654.

35. Pe´ rez-Santonja JJ, Sakla HF, Cardona C, Chipont E, Alio JL. Corneal sensitivity after photorefractive keratectomy and laser in situ keratomileusis for low myopia. Am J Ophthalmol 1999;127:497–504. 36. Kauffmann T, Bodanowitz S, Hesse L, Kroll P. Corneal reinnervation after photorefractive keratectomy and laser in situ keratomileusis: an in vivo study with a confocal videomicroscope. German J Ophthalmol 1997;5:508 –512. 37. Slowik C, Somodi S, Richter A, Guthoff R. Assessment of

314

AMERICAN JOURNAL

OF

OPHTHALMOLOGY

MARCH 2002