CLINICAL SCIENCES Correlation Between Epithelial Ingrowth ...€¦ · neas correlates with basement membrane remodeling at the flap margin. To address this question, we performed

CLINICAL SCIENCES

Correlation Between Epithelial Ingrowth andBasement Membrane Remodeling in Human CorneasAfter Laser-Assisted In Situ KeratomileusisPierre R. Fournie, MD; Gabriel M. Gordon, PhD; Daniel G. Dawson, MD; Francois J. Malecaze, MD, PhD;Henry F. Edelhauser, PhD; M. Elizabeth Fini, PhD

Objective: To further investigate the hypothesis that epi-thelial ingrowth in human corneas after laser-assisted insitu keratomileusis (LASIK) correlates with basementmembrane remodeling, as suggested by the presence ofmatrix metalloproteinase 9 around epithelial cells in thelamellar scar.

Methods: Immunohistochemical analysis and trans-mission electron microscopy were applied to human post-mortem corneas with post-LASIK epithelial ingrowth.

Results: Epithelial ingrowth into the flap margin wasobserved in 8 of 18 corneas (44%). Matrix metallopro-teinase 9 immunolocalized around ingrown epitheliumin 6 of these 8 corneas (75%). There was a positive cor-relation between the presence of matrix metalloprotein-ase 9 at the wound margin and discontinuities in the base-

ment membrane, as determined by laminin and �4 integrinimmunofluorescence. Transforming growth factor �2 waspresent into the stroma of some corneas with epithelialingrowth and interrupted basement membrane, suggest-ing some degree of epithelial-stromal interaction. Trans-mission electron microscopy confirmed large areas of re-modeled basement membrane along ingrown epithelialcells.

Conclusions: The neo–basement membrane compo-nents underlying the ingrown cells in human corneas withepithelial ingrowth after LASIK appear to be partially dis-assembled. Epithelial-stromal interaction over time maybe related to prolonged wound healing remodeling, whichcalls into question the stability of the flap.

Arch Ophthalmol. 2010;128(4):426-436

E PITHELIAL INGROWTH UNDER-neath the flap after laser-assisted in situ keratomileu-sis (LASIK) is a condition inwhich epithelial cells, which

normally cover the surface of the cornea,grow beneath the flap. This condition iscommon and occurs to some extent in 1%to 15% of LASIK cases.1,2 One study of1013 eyes demonstrated a 14.7% inci-dence rate, with 1.7% of eyes requiring sur-gical removal of epithelial ingrowth be-cause it interfered with vision.1 In anotherstudy of 3786 eyes, significant epithelialingrowth occurred in 0.92% of primaryLASIK cases and 1.70% of retreatment(“enhancement”) cases.2 Epithelial in-growth into the flap margin was found in20 of 38 corneas (53%) examined bymeans of step sections under light micros-copy.3 The difference between clinical andhistopathologic rates of epithelial in-growth corresponds to the “infraclinical”cases detected by microscopy. This differ-

ence highlights the frequency of epithe-lial ingrowth in post-LASIK corneas.

For epithelial ingrowth to occur, eitherepithelial cells are implanted during theprocedure or, more commonly, there is apersistent track in an area where the pe-ripheral flap does not adequately adhere.As a result, cells of the surface epithe-lium can grow into the interface.4 Risk fac-tors for epithelial ingrowth include trauma,flap dislocation, LASIK retreatment, anda deficient technique that results in pe-ripheral epithelial defects, poor flap ad-hesion, or perforated corneal flap.2,5

Visually significant epithelial in-growth can appear as early as 1 to 2 dayspostoperatively, but it appears most of-ten 1 to 3 months postoperatively.4 Mostepithelial ingrowth does not affect visionand does not require treatment. How-ever, in approximately 1% to 2% of caseswith epithelial ingrowth,2 it is necessaryfor the surgeon to lift up the flap and re-move the ingrown cells. Left unattended,

Author Affiliations: BascomPalmer Eye Institute, Universityof Miami Miller School ofMedicine, Miami, Florida(Drs Fournie, Gordon, andFini); Institut Nationalde la Sante et de la RechercheMedicale (INSERM), U563,Universite Toulouse III,Paul Sabatier, and Serviced’Ophtalmologie, HopitalPurpan, CHU Toulouse,Toulouse, France (Drs Fournieand Malecaze); Emory EyeCenter, Emory University,Atlanta, Georgia (Drs Dawsonand Edelhauser); and Institutefor Genetic Medicine, KeckSchool of Medicine, Universityof Southern California,Los Angeles (Drs Gordonand Fini).

(REPRINTED) ARCH OPHTHALMOL / VOL 128 (NO. 4), APR 2010 WWW.ARCHOPHTHALMOL.COM426

©2010 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Non-Human Traffic (NHT) User on 11/01/2020

the ingrown cells can continue to grow, affecting visionas the cells encroach on the visual axis and cause distor-tion at the flap surface. Aggressive ingrowth can be as-sociated with flap edge melting.1,2,4

Members of the matrix metalloproteinase (MMP) fam-ily are involved in normal and pathologic tissue repairprocesses, including epithelial regeneration and failureto heal, fibrotic repair and scar remodeling, infection, andangiogenesis.6-26 In a previous study, our group re-ported immunolocalization of MMP-9 (also known asgelatinase B) around epithelial cells trapped in the la-mellar scar of post-LASIK corneas with epithelial in-growth.27 Matrix metalloproteinase 9 catalyzes cleavageof all types of denatured collagens as well as compo-nents of the native basement membrane.28 It is pro-duced in the corneal epithelium, migrating across the in-tact basement membrane after abrasion injury, and thendisappears once regeneration is complete.10 However, insituations involving chronic epithelial defects, MMP-9levels remain elevated, causally contributing to failureto heal.10,11 Similarly, a number of different MMPs are pro-duced by the fibrotic repair tissue deposited in responseto keratectomy that penetrates through the epitheliumand into the stroma.9 A causal relationship between over-expression of MMP-9 in the corneal epithelium andbasement membrane dissolution/failure to heal has beendemonstrated in experimental models.10,11 In these stud-ies, the corneal epithelium gave the appearance of aninvading front, dissolving basement membrane and sub-sequently penetrating the underlying stroma in its path.

Our group recently suggested that MMP-9 immuno-localization around ingrown epithelium in post-LASIKcorneas may correlate with basement membrane inter-ruption or irregularities.27 Previous studies10,11 empha-sized overexpression of MMP-9 in the corneal epithe-lium in cases in which epithelial-stromal interactions wereobserved. We undertook the present study to determinewhether epithelial ingrowth in human post-LASIK cor-neas correlates with basement membrane remodeling atthe flap margin. To address this question, we performedanalyses using immunohistochemistry and transmis-sion electron microscopy.

METHODS

POST-LASIK TISSUE SAMPLES

After approval by the Emory University Institutional ReviewBoard, 18 postmortem corneoscleral buttons from 10 cornealeye bank donors with a history of LASIK surgery were ob-tained from various eye banks in North America. A number wasassigned to each cornea according to the order of inclusion. Thespecimens were received in corneal storage medium (Optisol-GS; Bausch & Lomb Surgical, Irvine, California) within 6 daysof death (mean [SD] time of preservation, 3.51 [1.40] days).Preoperative, intraoperative, and postoperative clinical rec-ords were reviewed when available. As controls, 4 postmor-tem normal corneas from 2 patients stored in corneal storagemedium (mean time of preservation, 2.95 [0.35] days) wereobtained from the Georgia Eye Bank (Atlanta) and the LionsEye Bank (Miami, Florida).

The corneoscleral buttons were evaluated for signs of pre-vious LASIK surgery by identifying a gray hazy semicircular ring

evident at the LASIK flap margin. The corneoscleral buttonswere oriented with the hinge in the superior position and thentrisected. The central portion was immediately snap frozen inliquid nitrogen, embedded in optimal cutting temperature com-pound (Tissue-Tek-II; Miles Inc, Elkhart, Indiana), and storedat −70°C. Frozen specimens were sectioned in a cryostat mi-crotome (Leica 1850 cryostat; Leica Microsystems Inc, Deer-field, Illinois) and mounted on adhesive-coated glass slides forconventional and immunofluorescent histologic processing. Theremaining frozen portion was processed for analysis by trans-mission electron microscopy (TEM). Normal control corneaswere processed identically to the LASIK corneas.

LIGHT MICROSCOPY

Sections were fixed and stained with hematoxylin-eosin ac-cording to routine protocols. The peripheral lamellar woundat the flap margin, the central lamellar wound, and the over-lying and underlying stroma were analyzed by light micros-copy with an inverted microscope (Zeiss Axiovert 200M; CarlZeiss Meditec, Jena, Germany) coupled to a camera (ZeissAxioCam MRc5). Histopathologic findings were recorded. Epi-thelium was studied in detail with the use of step sections tofocus on epithelial ingrowth underneath the flap.

INDIRECT IMMUNOLOCALIZATION

Slides to be stained were air-dried for 20 minutes at room tem-perature and then fixed for 20 minutes in 100% cold acetone(−20°C). They were washed with phosphate-buffered saline (PBS)3 times for 5 minutes each time, then incubated for 1 hour in ahumidified level chamber in 10% normal donkey serum (D9663;Sigma-Aldrich Corp, St Louis, Missouri) or goat serum (G9023;Sigma-Aldrich Corp) in PBS to block nonspecific staining. Pri-mary antibodies were used at a dilution of 1:100 and incubatedovernight at 4°C. The following antibodies were purchased: rab-bit polyclonal antibody to gelatinase B (RP3MMP9) (Triple PointBiologics, Forest Grove, Oregon), rabbit polyclonal antibody totransforming growth factor �2 (TGF-�2) (Santa Cruz Biotech-nology, Santa Cruz, California); rabbit polyclonal antilaminin(L9393) (Sigma-Aldrich Corp); and mouse monoclonal anti-body to mucin 16 (MUC16) and rat monoclonal antibody to�4 integrin (Abcam, Cambridge, Massachusetts). After 3 addi-tional washes with PBS for 5 minutes each, secondary antibodieswere applied for 1 hour. The secondary antibodies used wereconjugated goat anti–mouse IgG (A-11001), donkey anti–rat IgG(A-21208), and donkey anti–rabbit IgG (A-21206) (Alexa Fluor488; Invitrogen Molecular Probes, Carlsbad, California). Sampleswere mounted with the use of mounting medium with 4�,6-diamidino-2-phenylindole (Vectashield; Vector Laboratories,Burlingame, California) for nuclear counterstaining. Negative con-trol sections were processed identically but incubated with strain-specific IgG as the primary antibody. Rabbit IgG, rat IgG, andmouse IgG were purchased (Chemicon, Temecula, California).

Samples were examined with an inverted fluorescence mi-croscope (Zeiss Axiovert 200M), and images were captured witha camera (Zeiss AxioCam MRc5) attached to the microscope.Camera and microscope settings were controlled by software(Axiovision version 4.1; Carl Zeiss Meditec). Four regions wereevaluated: (1) the LASIK flap wound margin, (2) the cornealstroma in the LASIK flap, (3) the paracentral and central la-mellar wound regions, and (4) the residual stromal bed. Nor-mal control corneas were evaluated in central, paracentral, andperipheral regions. The LASIK flap wound margin was also ex-amined with a confocal microscope (Leica TCS SP2; Leica Mi-crosystems Inc, Bannockburn, Illinois).




TRANSMISSION ELECTRON MICROSCOPY

Corneal portions frozen in liquid nitrogen were fixed over-night in cold 3% glutaraldehyde and 2% paraformaldehyde inPBS. The specimens were then postfixed in 1% osmium tetrox-ide for 1 hour, rinsed in PBS, dehydrated in a series of ethanoland propylene oxide solutions, and embedded in epoxy resin.Semithin (0.5-1.0 µm) and ultrathin (70-80 nm) sections werecut on a microtome (Porter Blum MT-2; Sorvall, Newtown, Con-necticut). Semithin sections were stained with toluidine blueand evaluated by light microscopy to identify regions of epi-thelial ingrowth or normal unaffected regions at the LASIK flapwound margin. The blocks were trimmed around these areasof interest and ultrathin sections were cut, placed in a coppergrid, double-stained with uranyl acetate and lead citrate, andexamined by TEM (CX-100; JEOL, Tokyo, Japan).

RESULTS

DEMOGRAPHICS OF LASIKAND CONTROL CASES

The demographics of the donors of the LASIK and con-trol specimens are given in the Table. The average (SD)age of the 10 LASIK donors was 50.3 (7.15) years (range,34-62 years), with an average time after LASIK of 4.7 (1.8)years (range, 2-8 years). Patient 10 underwent LASIK with2 enhancements (flap relifting) at 6 and 12 months afterthe initial surgery, and a third enhancement 4 years af-ter the initial surgery (flap recutting). The average ageof the 2 control donors was 53.0 (2.8) years (range, 51-55years). For patients whose clinical records were avail-able, we determined that epithelial ingrowth was not clini-cally significant. No visual disturbances were reported,and none of the patients required treatment.

INCIDENCE OF EPITHELIAL INGROWTHIN POST-LASIK CORNEAS

Our group previously reported an incidence rate of 33.3%for epithelial ingrowth in this series.27 In the present study,we reviewed these data by increasing the number of serialsections examined. By means of this procedure, epithelial

ingrowth into the flap margin was observed in 8 of the 18corneas examined (44%). Epithelial ingrowth was limitedto the peripheral edge of the flap in all cases. The amountof epithelial ingrowth into the lamellar wound varied be-tween corneas and also between different parts of the samecornea. In5corneas,epithelial ingrowthwasconnectedwiththe outer epithelial layer (Figure 1A and B). In 1 cornea,microscopic foci or islands of epithelial cells lay in the scar,and there was no evidence of a connection to the surface(Figure1C).Twocorneas showedbothepithelial ingrowthpatterns. These data show that epithelial ingrowth is a fre-quent pattern of healing at the margin of a LASIK flap.

MMP-9 IMMUNOLOCALIZATION

Our group previously localized MMP-9 in human post-LASIK corneas with epithelial ingrowth.27 To verify ourprevious findings, we repeated our experiments with theuse of serial sections. The presence of MMP-9 was de-tected around epithelial cells trapped in the lamellar scarin 6 of 8 corneas (75%) with epithelial ingrowth. Nor-mal control corneas and post-LASIK corneas without epi-thelial ingrowth did not stain for MMP-9 (not shown).Within the epithelial ingrowth subgroup, 6 of 7 corneas(86%) with epithelial ingrowth connected to the surfacewere immunoreactive for MMP-9 (Figure 2A-C). The1 patient with epithelial ingrowth not obviously con-nected to the surface showed no MMP-9 staining(Figure 2D). Within the 2 corneas with both epithelialingrowth patterns, only 1 had MMP-9 staining aroundfoci or islands of epithelial cells (Figure 2E). Therefore,staining for MMP-9 around epithelial cells trapped withinthe scar but unconnected to the surface was observed in1 of 3 corneas (33%). These results indicate that MMP-9typically immunolocalizes around ingrown epithelium.

BASEMENT MEMBRANE DISCONTINUITIES

Studies using experimental models have suggested a causalrelationship between overexpression of MMP-9 in the cor-neal epithelium and basement membrane dissolution/failure to heal.10,11 We investigated the distribution of base-ment membrane markers at the wound margin of post-LASIK corneas showing epithelial ingrowth with orwithout MMP-9 immunoreactivity. Corneas were stainedwith a polyclonal antibody against laminin and a mono-clonal antibody against �4 integrin subunit and viewedby immunofluorescence microscopy.

Laminin is a component of the basement membrane,whereas the heterodimeric complex of �6�4 integrin as-sociates with hemidesmosomes, which serve as the cells’main anchor to the basement membrane.29,30 In undam-aged corneas and post-LASIK corneas devoid of MMP-9staining, the corneal epithelial basement membrane ap-peared as a continuous distinct line between the basalsurface of the epithelial basal cells and the anterior stroma(Figure 3A-F). We consistently observed discontinui-ties or an apparent absence of localization of basementmembrane markers in regions showing MMP-9 stainingaround ingrown epithelium (Figure 3G-L). At some siteswe saw nonuniform staining, defined as a diffuse irregu-lar staining; this was different from the patchy staining

Table. Demographics of Human Post-LASIKand Control Corneas From Eye Banks

Patient No./Age, y

No. of CorneasHarvested

PostoperativeInterval, y

1/51 1 62/34 2 23/55 2 54/51 2 25/50 2 66/49 2 47/48 2 68/55 2 69/62 2 310/48 1 8 (1st flap)/4 (2nd flap)Control 1/51 2 NAControl 2/55 2 NA

Abbreviations: LASIK, laser-assisted in situ keratomileusis; NA, notapplicable.




characterizing the absence of staining in some areas. Theseresults suggest that MMP-9 immunolocalized around in-grown epithelial cells positively correlates with discon-tinuities in the basement membrane.

POLARITY OF TRAPPED EPITHELIAL CELLSIN EPITHELIAL INGROWTH

Basal cells of the corneal epithelium secrete the compo-nents necessary to form the basement membrane.31 Thepresence of the basement membrane between the basalepithelium and the underlying stroma determines the po-larity of epithelial cells. Therefore, loss of polarity in thetrapped ingrown cells may disrupt the basement mem-brane. Previous reports have localized the membrane-associated mucin MUC16 along the apical membrane ofthe apical and subapical cells in human ocular surfaceepithelium.32,33 Apical and subapical cells, however, showlow metabolic activity. On the other hand, basal cells arenot embedded with MUC16 mucin but secrete the com-ponents necessary to form the basement membrane. Weinvestigated the epithelial distribution of MUC16 mu-cin to determine whether ingrown cells express basal or

apical patterns. In all cases, MUC16 mucin immunolo-calized around the apical and subapical epithelial cells(Figure 4). No staining for MUC16 mucin was foundaround trapped cells, suggesting that ingrowing epithe-lium retains its polarity and appears to comprise basalor intermediate cells.

One major role that integrins fulfill as cell membranereceptors is to ensure anchorage dependence; this meansthat, under normal conditions, most cells must remainin their precise location because their detachment fromthe extracellular matrix usually leads to apoptosis.34,35

Staining for the �4 integrin subunit is seen only in basalcells, where the �6�4 complex interacts with hemides-mosomes.29,30 No staining was seen for �4 integrin aroundingrown cells not bound to the basement membrane. Thestaining for �4 integrin localized along the basal mem-brane of basal cells in contact with the basement mem-brane (Figure 3B, D, F, H, J, and L).

Taken together, these data suggest that the polarityof ingrown cells is maintained and that basement mem-brane discontinuities do not correlate with a loss in theability of epithelial cells to secrete the components nec-essary for formation of the basement membrane.

A B C

Figure 1. Varying types of epithelial ingrowth observed at the laser-assisted in situ keratomileusis (LASIK) wound margin. Epithelial ingrowth can be connectedwith the outer epithelial layer and follows the LASIK cut in linear (A [patient 1]) or saccular (B [patient 7, left eye]) patterns. Some post-LASIK corneas show foci(arrowheads) or islands (arrow) of epithelial cells in the scar that are not connected to the surface (C [patient 9, left eye]) (hematoxylin-eosin, originalmagnification �40).

A B C

D E

Figure 2. Immunolocalization of matrix metalloproteinase 9 (MMP-9) around ingrown epithelium in corneas after laser-assisted in situ keratomileusis (LASIK).MMP-9 was detected (arrowheads) around epithelial cells trapped in the lamellar scar, with (A [patient 1] and B [patient 7, left eye]) or without (E [patient 9, lefteye]) connection to the surface. For both patterns, however, some patients exhibited no staining for MMP-9 (C [patient 9, right eye] and D [patient 3, right eye])(original magnification �63 [A, E] and �40 [B-D]).




CORRELATION OF STROMAL STAININGFOR TGF-�2 WITH EPITHELIAL STAININGFOR MMP-9 AND WITH DISCONTINUITIES

OF EPITHELIAL BASEMENT MEMBRANE

Our group previously determined that TGF-�2, the ma-jor cytokine that mediates fibrotic repair in corneas, isconfined to the epithelium in the presence of basementmembrane.36,37 In the absence of basement membrane,

however, the stroma stains strongly for TGF-�2, mean-ing that basement membrane prevents fibrotic marker ex-pression by inhibiting TGF-�2 release into the stroma.Therefore, we hypothesized that the presence of TGF-�2in the stroma is further evidence of epithelial-stromal in-teraction with basement membrane discontinuities. In un-wounded corneas and post-LASIK corneas devoid ofMMP-9 staining, TGF-�2 was localized to the epithe-lium (Figure 5A-C). In 6 corneas with epithelial in-growth as well as MMP-9 staining around ingrown epi-thelium and discontinuities of the basement membrane,the stroma stained for TGF-�2 directly beneath the epi-thelium (Figure 5D and E). Staining was absent in theadjacent stroma and remained associated with the ante-rior stromal region along the apparently disassembledbasement membrane. These data indicate localized epi-thelial-stromal interactions at sites where the basementmembrane shows discontinuities.

TEM ANALYSIS

The immunofluorescence data suggest that the base-ment membrane may be lost in some post-LASIK cor-neas with epithelial ingrowth. To examine this questionmore carefully, TEM studies were conducted. The TEMdata shown come from areas with epithelial ingrowth.An apparently intact neo–basement membrane could bevisualized along most of the interface between epithe-lial ingrown cells and the overlying stroma. Neo–base-ment membrane underlying the ingrown cells demon-strated a basement membrane–like structure with a palelayer (the lamina lucida) immediately posterior to thecell membrane of the epithelial basal cells and an elec-tron-dense layer (the lamina densa). Numeroushemidesmosomes anchor the epithelial cells to thebasement membrane (Figure 6). The Bowman layerdid not regenerate even though, in some regions, a thin,acellular, membranelike zone separated the basementmembrane from the collagen fibers (Figure 6). Whenepithelial ingrowth was connected to the surface, in-

LamininPatient

Control

9R

3R

1

7L

9L

β4 Integrin

A B

C D

E F

G H

I J

K L

Figure 3. Localization of laminin and �4 integrin subunit confirming thebasement membrane discontinuities in corneas after laser-assisted in situkeratomileusis (LASIK) with matrix metalloproteinase 9 (MMP-9) localizedaround ingrown epithelium. Both basement membrane markers showed alinear, continuous staining pattern in the epithelial basement membrane in theunwounded corneas (A and B). In the post-LASIK corneas devoid of MMP-9staining, �4 integrin staining appeared more continuous than laminin staining,which exhibited an irregular pattern (C-F). In post-LASIK corneas with MMP-9localized around ingrown epithelium, staining for both basement membranemarkers appeared either irregular (H and J) or interrupted (G, I, K, and L)(arrowheads) (original magnification �40 [A-F and H-L], �63 [G]). R and Lnext to patient numbers indicate right and left eyes.

A B

C D

Figure 4. Maintained polarity of trapped epithelial cells in epithelial ingrowth.In control corneas (A) and in all corneas after laser-assisted in situkeratomileusis (B-D [patients 3, right eye; 7, left eye; and 9, left eye), MUC16mucin immunolocalized around the apical and subapical epithelial cells(arrowheads), with no staining around ingrown epithelial cells (originalmagnification �63).




grown cells looked viable. No pyknotic nuclear alter-ations were observed, whereas small vacuoles were ob-served within the cytoplasm. Cells were generallycompressed against each other. Viable foci or islands ofepithelial cells in the scar were close to the margin ofthe flap. In corneas that showed epithelial ingrowth to-gether with MMP-9 staining around ingrown epithe-lium and basement membrane discontinuities observedby immunofluorescence, the disappearance of specificcomponents of basement membrane was visualizedwith TEM with different patterns. In some cornealareas, the lamina lucida was highly disrupted or com-pletely absent, and the spacing between the basal cellmembrane and the apparent lamina densa (Figure 6Aand C and Figure 7A-E) was increased. Hemidesmo-somes were absent, and the electron-dense laminadensa exhibited occasional focal disruptions. The disas-

sembly of the basement membrane was substantial butincomplete: remnants of the lamina densa were ob-served at some locations. In other areas, the lamina lu-cida showed significant discontinuities and the laminadensa was almost entirely absent. In some places, nocomponents of the basement membrane were distin-guishable, and in its place were spaces filled with anelectron-dense granular material between ingrown cellsand the adjacent stroma (Figure 7F). No apparentlamina lucida and densa were visible. Despite the ab-sence of mature hemidesmosomes, the epithelial cellsappeared to be associated with the underlying stromathrough a fibrous capsule around the trapped cells.Thus, ultrastructural analysis of the basement mem-brane at the wound margin of post-LASIK corneas withepithelial ingrowth confirms large areas of remodelingof basement membrane.

A B C

D E

Figure 5. Finding of stromal staining for transforming growth factor �2 (TGF-�2) only in corneas treated with laser-assisted in situ keratomileusis (LASIK)exhibiting matrix metalloproteinase 9 (MMP-9) staining around ingrown epithelium and immunofluorescence basement membrane discontinuities.Immunoreactive protein for TGF-�2 was confined to the epithelium in control corneas (A) and in post-LASIK corneas devoid of MMP-9 staining (B [patient 3, righteye] and C [patient 9, right eye]). In post-LASIK corneas with MMP-9 staining around ingrown epithelium and basement membrane discontinuities detected byimmunofluorescence, TGF-�2 was localized in both the epithelium and the stroma adjacent to the ingrown cells (arrowheads) (D [patient 7, left eye], E [patient 9,left eye]) (original magnification �40 [A-C] and �63 [D and E]).

A B C

LD

EP EPEP

SS

SLL

Figure 6. Ultrastructural transmission electron micrograph studies show a basement membrane–like structure (A-C) beneath the epithelial ingrown cells (EP) withhemidesmosomes (arrowheads), a lamina lucida (LL), and a lamina densa (LD). The Bowman layer did not regenerate but, in some regions, a thin acellularmembranelike zone (*) separated the basement membrane from the collagen fibers of the underlying stroma (S) (A-C). Some areas showed increased spacingbetween the LL and the L/D and the absence of hemidesmosomes (white arrows) (A and C) (original magnification �5200 [A] and �8900 [B and C]).




MMP STAINING IN AREAS OF FIBROSIS

The corneal epithelium controls fibrotic activation. Whenthe basement membrane is preserved, changes in cor-neal stromal cells associated with fibrotic activation donot occur.36,37 In the absence of a basement membrane,TGF-�2 is released into the stroma: keratocytes becomeactive with a fibrotic phenotype and accumulation of ex-tracellular matrix.36,37 Therefore, we hypothesized thatthe epithelial-stromal interaction during epithelial in-growth, which induces stromal exposure to TGF-�2, maylead to fibrosis. The TEM confirmed fibrotic foci near theingrown epithelial cells, with deposition and remodel-ing of the extracellular matrix with disorganized stro-mal lamellae (Figure 8A and B). There was also clearevidence of a breakdown in some stromal lamellae(Figure 8C and D) and of spaces between the stromal la-mellae in some cases (Figure 8C).

To verify our previous findings, we investigated thedistribution of fibrotic MMPs. In addition to depositingextracellular matrix, activated keratocytes and contigu-ous epithelium turn on new synthesis of enzymes thatcan degrade extracellular matrix, demonstrating that thisnew stroma is being actively remodeled.38 Matrilysin(MMP-7) and stromelysin 1 (MMP-3) are members of thestromelysin family of MMP enzymes. These contributeto extracellular matrix degradation and may play a rolein the stromal degradation and matrix remodeling afterexcimer keratectomy.39 In all cases with epithelial in-growth, MMP-7 was immunolocalized to the ingrown epi-thelial cells (Figure 9B and D). In contrast to matrily-sin, MMP-3 was immunolocalized to some keratocytesaround ingrown cells, but not to the corneal epithelial

cells (Figure 9A and C). Normal control corneas and post-LASIK corneas without epithelial ingrowth did not stainfor MMP-3 (not shown). Normal and post-LASIK cor-neas without epithelial ingrowth displayed little stain-ing for MMP-7 (not shown).

Abnormal accumulation of fibrotic extracellular ma-trix components and MMPs near epithelial ingrowth sug-gests ongoing lysis and remodeling of corneal stroma.

COMMENT

Our previous study focused on the long-term MMP-9 lo-calization around epithelial cells trapped in the lamellarscar of post-LASIK corneas with epithelial ingrowth.27 Inthis article, we report the absence of continuous epithe-lial basement membrane around ingrown epithelial cells.With LASIK, the epithelium, basement membrane, andBowman layer remain intact except for the area in whichthe flap is cut. Advantages are quicker recovery and lesspostoperative haze because of the lack of epithelial-stromal interactions. One limitation of this study is thatit was performed in a cadaver eye model, specifically, ineyes with no clinical evidence of visually or functionallysignificant epithelial ingrowth. However, even in theseinfraclinical cases, LASIK with epithelial ingrowth ap-pears as a human postsurgical model featuring epithelial-stromal interaction during healing.

Basal cells of the corneal epithelium secrete the com-ponents necessary for formation of the basement mem-brane. New basement membrane components such aslaminin and type IV collagen are deposited underneathmigrating corneal epithelial cells10 and around corneal

A B C

D E F

EPEP

EP

EP

EP EP

S

SS

S

SS

Figure 7. Ultrastructural transmission electron micrograph studies confirming that the basement membrane–like structure separating epithelial ingrown cells (EP)from the adjacent stroma (S) exhibits areas of various discontinuities. The disassembly of the basement membrane took several features. Some increased spacing(*) between the lamina lucida and the lamina densa were observed (A-E) with (B and E) or without (A, C, and D) disappearance of hemidesmosomes(arrowheads). Lamina lucida was highly disrupted (A-D) or completely absent (E). Lamina densa (black arrows) exhibited areas of chaotic arrangements withscalloped features (A, B, C, and E) or interruptions (C and D). In some places, no components of the basement membrane were distinguishable, and anelectron-dense granular material between ingrown cells and the adjacent stroma formed a fibrous capsule (white arrows) (F) (original magnification �8900).




epithelial cells implanted in the stroma.40 Basement mem-brane–like structures should therefore be found aroundpost-LASIK epithelial ingrowth. Although there are sev-eral case studies on epithelial ingrowth after LASIK andeven more epidemiologic studies on the incidence andtreatment of this complication, histopathologic studiesare sparse. To our knowledge, only 7 such reports ex-ist.41-47 None of these articles shows immunofluores-cence data, whereas 3 show TEM results.42,46,47

Matrix metalloproteinase 9 is produced in the cor-neal epithelium as it regenerates across the intact base-ment membrane after an abrasion injury, and then it dis-appears once regeneration is complete.10 In situationsinvolving chronic epithelial defects, however, MMP-9 lev-els remain elevated and contribute to the failure to heal.10,11

Matrix metalloproteinase 9 has the capacity to cleave epi-thelial basement membrane components such as colla-gen types IV and VII and laminin.28,48 Experimental mod-els suggest a causal relationship between overexpressionof MMP-9 in the corneal epithelium and basement mem-brane dissolution/failure to heal.10,11 In these studies, thecorneal epithelium gave the appearance of an invadingfront, dissolving basement membrane and subsequentlypenetrating the underlying stroma lying in its path. Thepresence of MMP-9 associated with a disrupted base-ment membrane around ingrowing epithelium suggestsa similar long-term, ongoing process of remodeling andinvasion.

Both TEM and immunofluorescence microscopy wereused in the present study to evaluate the basement mem-brane in post-LASIK corneas. The disappearance of spe-

cific components of basement membrane was visual-ized by indirect immunolocalization methods. The probeswere chosen to detect components localized to 2 areasin the basement membrane: anti-laminin to detect thelamina lucida and anti–�4 integrin subunit, which to-gether with �6 integrin subunit associates with hemides-mosomes. Both of these components are considered thecells’ main anchors to the basement membrane. Both lami-nin and �4 integrin exhibit immunostaining discontinui-ties. This is consistent with the capacity of MMP-9, whenoverexpressed, to cleave epithelial basement compo-nents and reduce adhesion complex integrity.10,11 Our re-search group has previously shown that, in the presenceof basement membrane, TGF-�2 is absent from the stromaand is found only in the epithelium; in contrast, in theabsence of basement membrane, the stroma stains stronglyfor TGF-�2.36,37 Therefore, stromal staining for TGF-�2in areas of post-LASIK corneas with epithelial ingrowthconstitutes further evidence of the existence of base-ment membrane discontinuities. Because we observedstaining discontinuities of basement membrane mark-ers in only some areas around ingrown cells, additionalultrastructural studies were conducted. This analysisshowed basement membrane–like structures in someplaces. Other areas were devoid of such structures, con-firming the basement membrane discontinuities. Al-though some places appeared to lack specific compo-nents of the basement membrane, other places lacked anybasement membrane–like structures or even showed acapsule of connective tissue, similar to scar tissue, thatseparated the cells from the healthy stroma. These datamay explain differences in immunostaining findings thatshow epithelial-stromal interactions in some cases andtheir absence in others. This variability could be causedby a basement membrane–like structure or a capsule act-ing as a fibrous cocoon around the epithelium itself.49

A B

C D

EP

EPEP

EP

F

F

F

K

S

S

S

SS

Figure 8. Ultrastructural transmission electron micrograph studies showingfibrotic spots (F) within the stroma (S) adjacent to the epithelial ingrowncells (EP). Fibrotic areas exhibit a disorganized pattern of collagen fiberdeposition (*) (A). Keratocyte (K) activation is observed, as evidenced byincreased keratocyte density and adjacent collagen deposition (*) (B).Spaces between the stromal lamellae have formed in some cases(arrowheads) (C). Disorganized collagen deposition (black arrows) separatesthe neo–basement membrane from the collagen fibers, exhibiting clearevidence of a breakdown in the stromal lamellae (C), whereas others do notshow any spacing (white arrows) between the neo–basement membrane andthe stromal lamellae (D) (original magnification �8900 [A, C, and D] and�5200 [B]).

MMP-3Patient

1

9L

MMP-7

A B

C D

Figure 9. Immunolocalization of fibrotic matrix metalloproteinases (MMPs)around epithelial ingrowth with fibrotic spots. MMP-3 was localized aroundkeratocytes in the stroma adjacent to the ingrown cells (arrowheads)(A and C). MMP-7 was detected around epithelial cells at the surface of theedge of the laser-assisted in situ keratomileusis and around epithelial cellstrapped in the lamellar scar (B and D) (original magnification �63). L next topatient number indicates left eye.




In addition to the fibrotic environment, the presenceof stromelysins is further evidence of an ingrown epi-thelial-stromal interaction and that this adjacent stromais being actively remodeled. Matrix metalloproteinase 3was only immunolocalized to some keratocytes aroundepithelial cells trapped in the lamellar scar of post-LASIK corneas with epithelial ingrowth. This is consis-tent with the capacity of activated keratocytes to upregu-late the synthesis of MMP-3, whereas MMP-3 is notsynthesized by stromal cells in the uninjured cornea.9 Ma-trix metalloproteinase 7 was immunolocalized to the in-grown epithelial cells. This protein is expressed in epi-thelial-derived dividing cells. The corneal epithelium isone of the few nontumor sites of MMP-7 expression,15

but its role in the cornea has not been fully elucidated.Its natural substrates are proteoglycans, elastin, and gly-coproteins such as fibronectin and laminin.50,51 In-creased epithelial-stromal interactions have been impli-cated as a cause of scarring after removal of a basementmembrane layer.36,37 Matrilysin may play a role in base-ment membrane degradation in the cornea and in facili-tation of epithelial-stromal interactions,15 whereas strome-lysin 1 may facilitate the corneal fibrosis throughkeratocyte activation. Girard et al,9 however, found moreexpression of stromelysin 1 in the repair-adjacent stromathan in the repairing stroma 1 week after injury. In ad-dition, stromelysin 1 may play a role in reactivating col-lagenase.28 Stromelysin 1 may thus help clear the sub-epithelial scarring, a process that is often observed severalmonths after injury.

Two reasonable mechanisms can be proposed to ex-plain pathologic epithelial growth under a corneal flap.The first suggests that the cells are implanted under theflap during the surgery.52 The implantation may be sec-ondary to the microkeratome blade mechanically drag-ging the cells into the lamellar interface during keratec-tomy. Epithelial cells from the corneal surface may alsofloat into the interface during irrigation of the stromalbed after the ablation. It is believed that such cells havea very low potential to spread because they do not havemitotic potential, inasmuch as they are disconnected fromthe limbal stem cells and transient amplifying cells. Theseepithelial clusters, which are unconnected to the flapedge, may therefore resolve within months. Indeed, inthe present study, 1 cornea had an isolated colony of epi-thelial cells located away from the edge of the flap. In-grown cells may experience the corneal stroma as a “hos-tile” hypoxic environment, lacking basic nutrients fromthe tear film. This may explain the presence of epithelialcell remnants within the interface at the edge of the flapand the absence of MMP-9 release or basement mem-brane markers. We hypothesize that enzymes are likelyto be released soon after LASIK surgery but may not havea strong influence on the normal epithelial-stromal cyto-kine interaction because the amounts released are prob-ably limited.

The second theory states that the corneal epitheliuminvades the interface through a defect in the flap or inareas where the flap borders adhere poorly to thestroma.2,4,46,53 In most cases, continuity between the epi-thelium in the interface and the surface epithelium sug-gests that postoperative invasion is the most likely cause.

It is appropriate to view epithelial ingrowth as an epi-thelial fistula underneath the flap. Majo et al54 have shownthat the epithelial cells in the interface communicate withthe limbal stem cells. Asano-Kato et al,47 in a histopatho-logic study of 5 specimens of epithelial ingrowth, re-ported that early epithelial ingrowth consists of multi-layered squamous epithelial cells resembling thedifferentiated corneal epithelium; late epithelial in-growth, in contrast, was made up of clumps containingamorphous materials with scarce cellular elements. Asin our present study, those authors observed basementmembrane–like structure in some places and the ab-sence of basement membrane in other places. Thus, weprobed further to test whether loss of basement mem-brane in some places may be due to a lack of its forma-tion by epithelial cells or to its excessive degradation byenzymatic proteolysis.

The data presented herein and in the literature42,46,47

show that a basement membrane–like structure sur-rounds epithelial ingrown cells. In addition, it has beenreported that cells involved in epithelial ingrowth afterLASIK have short generation times with stem cell char-acteristics54 and a pluristratified appearance.47,54 It maybe that the loss of polarity of trapped ingrown cells actsto reduce basement membrane synthesis. We failed to de-tect the membrane-associated mucin MUC16 constitu-tively expressed by the human ocular surface epithelia.Taken together, these data suggest that cells within thelesion may differentiate with polarity as in the cornealsurface epithelium, which would explain their ability tosecrete basement membrane components.

In the normal cornea, epithelial cells undergo rela-tively rapid mitosis with differentiation toward the api-cal side and exfoliation of superficial flattened squa-mous cells.55 In the case of epithelial ingrowth, however,differentiated cells cannot exfoliate and consequently ac-cumulate. These physically limited conditions may ex-plain the observed local cell degeneration, phagocytosiswith local detersion of cell debris, and accumulation ofamorphous material. Lack of oxygen and various nutri-ents from the tear film necessary for maintaining cellu-lar metabolism may also explain some metabolic changesand the inability of ingrown cells to synthesize base-ment membrane.

Epithelial ingrowth soon after surgery may also causemelting of the flap edge.1,2,4,52 This peripheral loss of flaptissue has been connected to the release of proteases inthe area of epithelial ingrown cells. We did see MMP lo-calization. Matrix metalloproteinase 9 is capable of de-grading components of the basement membrane. Suchdegradation may account for the partial loss of base-ment membrane over time, and for the alterations in bothits structure and its function. Immunolocalization ofTGF-�2 in the region directly beneath the epithelium isfurther evidence of the epithelial-stromal interaction. Fi-brosis occurred as a result of abnormal extracellular ma-trix remodeling and is thought to be dependent on spe-cific cytokines such as TGF-�2.36,37 Extracellular matrixdeposits accumulated during wound healing are often re-moved over time by specific proteinases expressed in theaffected areas.38 We found cells overexpressing MMP-9,MMP-3, and MMP-7, which can degrade extracellular ma-




trix and basement membrane components. These datasuggest the occurrence of ongoing proteolysis, possiblyas an attempt of corneal stromal and epithelial cells tolocally remodel the stroma and remove fibrotic depos-its. Fibrosis resorption by extracellular proteolysis maytake several months to several years. This process doesnot seem to require myofibroblasts. Myofibroblasts arenecessary at the early stages of fibrotic scar formation;at later stages and in chronic scars, active stromal re-modeling may be continued by activated keratocytes inthe fibrotic scar area. These results suggest that epithe-lial ingrown cells after LASIK initiate a stromal woundhealing response consisting of an acute response in theearly stages and a more chronic, progressive response inlater stages. A new extracellular matrix may develop insome places and grow to form a fibrous capsule aroundthe trapped cells. A basement membrane–like structureor this fibrous cocoon may thus explain why most casesof epithelial ingrowth are self-limiting and asymptom-atic over time. Nevertheless, our data suggest an ongo-ing, low level of remodeling around some ingrown epi-thelial cells.

Therefore, either hypothesis—lack of synthesis or ex-cessive degradation—may account for the observed lossof basement membrane components. Even though a syn-thesis defect is possible in light of the data, the presenceof MMP-9 favors the degradation hypothesis.

The pathogenesis and progression of epithelial in-growth are unknown. Studies do show, however, thatLASIK enhancement by flap relifting increases the riskof epithelial ingrowth.2 Even if most patients are asymp-tomatic, progressive epithelial ingrowth can encroach onthe central visual axis or can result in flap melting andirregular astigmatism.1,2,4,52 Treatment in these cases con-sists of lifting the flap and scraping the epithelial cells.

Most cases of epithelial ingrowth are mild and self-limiting and require only careful observation. The diag-nosis of epithelial ingrowth is not necessarily an indica-tion for immediate removal. Many isolated nests ofingrown cells will regress within a few months with noadverse consequences. However, some epithelial in-growth may persist over time. Schmack et al56 studiedthe cohesive tensile strength of human LASIK cornealwounds and found that the weakest wound margin scarsshowed epithelial cell ingrowth. Even if there is no clini-cal evidence of an increased risk of flap displacement inpatients with epithelial ingrowth after LASIK, the pres-ence of MMP-9, which is associated with a disrupted base-ment membrane around ingrown epithelium, suggestsan ongoing remodeling process. This fact may help ex-plain the previous findings that identified the weakestwound margin scars as those with epithelial ingrowth.The potential consequences of such long-term remod-eling on the edge of the flap remain unknown.

We cannot say with certainty whether epithelial in-growth and the ongoing repair process it involves are thecause or the consequence of the presence of MMP. How-ever, we propose that minor defects in the surgically cre-ated flap prevent perfect alignment with the underlyingcornea after surgery. This would make slippage likely,creating space for epithelial ingrowth and a long-term re-quirement for fibrotic repair. As in other long-term re-

pair processes, MMP expression may become excessivebecause of ever-amplifying feedback loops, and this couldgradually contribute to the failure to heal. In such situ-ations, judicious and timely use of appropriate MMP in-hibitors may be beneficial by inhibiting epithelial in-growth and enabling fibrotic repair tissue to accumulatesufficiently to “tack” the flap in place.

In the present study, we showed immunolocaliza-tion of MMP-9 and basement membrane discontinuitiesaround ingrown epithelial cells in post-LASIK corneas.The epithelial-stromal interaction observed after LASIKover time, together with chronic epithelial ingrowth, maybe related to the prolonged remodeling of the adjacentstroma and calls into question the potential long-termconsequences for the edge of the flap. A more completeunderstanding of epithelial cell–matrix interactions af-ter LASIK may enable us to prevent or treat significantepithelial ingrowth more successfully.

Submitted for Publication: July 19, 2009; final revisionreceived October 20, 2009; accepted October 29, 2009.Correspondence: Pierre R. Fournie, MD, Department ofOphthalmology–Purpan Hospital, Place du Dr Baylac,31059 Toulouse CEDEX 9, France ([email protected]).Financial Disclosure: None reported.Funding/Support: This study was supported by Na-tional Eye Institute grants R01-EY012651 (Dr Fini), P30-EY014801 (Dr Fini), and R01-EY00933 (Dr Edel-hauser); unrestricted grants from Research to PreventBlindness (to Bascom Palmer Eye Institute, University ofMiami Miller School of Medicine, and Emory Eye Cen-ter); a Senior Scientific Investigator Award (Dr Fini); andthe Walter G. Ross Foundation (Dr Fini).

REFERENCES

1. Stulting RD, Carr JD, Thompson KP, Waring GO III, Wiley WM, Walker JG.Complications of laser in situ keratomileusis for the correction of myopia.Ophthalmology. 1999;106(1):13-20.

2. Wang MY, Maloney RK. Epithelial ingrowth after laser in situ keratomileusis. AmJ Ophthalmol. 2000;129(6):746-751.

3. Dawson DG, Kramer TR, Grossniklaus HE, Waring GO III, Edelhauser HF. His-tologic, ultrastructural, and immunofluorescent evaluation of human laser-assisted in situ keratomileusis corneal wounds. Arch Ophthalmol. 2005;123(6):741-756.

4. Asano-Kato N, Toda I, Hori-Komai Y, Takano Y, Tsubota K. Epithelial ingrowthafter laser in situ keratomileusis: clinical features and possible mechanisms. AmJ Ophthalmol. 2002;134(6):801-807.

5. Jabbur NS, Chicani CF, Kuo IC, O’Brien TP. Risk factors in interface epithelial-ization after laser in situ keratomileusis. J Refract Surg. 2004;20(4):343-348.

6. Fini ME, Stramer BM. How the cornea heals: cornea-specific repair mechanismsaffecting surgical outcomes. Cornea. 2005;24(8)(suppl):S2-S11.

7. Sivak JM, Fini ME. MMPs in the eye: emerging roles for matrix metalloprotein-ases in ocular physiology. Prog Retin Eye Res. 2002;21(1):1-14.

8. Matsubara M, Girard MT, Kublin CL, Cintron C, Fini ME. Differential roles for twogelatinolytic enzymes of the matrix metalloproteinase family in the remodelingcornea. Dev Biol. 1991;147(2):425-439.

9. Girard MT, Matsubara M, Kublin C, Tessier MJ, Cintron C, Fini ME. Stromal fi-broblasts synthesize collagenase and stromelysin during long-term tissueremodeling. J Cell Sci. 1993;104(pt 4):1001-1011.

10. Matsubara M, Zieske JD, Fini ME. Mechanism of basement membrane dissolu-tion preceding corneal ulceration. Invest Ophthalmol Vis Sci. 1991;32(13):3221-3237.

11. Fini ME, Parks WC, Rinehart WB, et al. Role of matrix metalloproteinases in fail-ure to re-epithelialize after corneal injury. Am J Pathol. 1996;149(4):1287-1302.




12. Azar DT, Pluznik D, Jain S, Khoury JM. Gelatinase B and A expression after laserin situ keratomileusis and photorefractive keratectomy. Arch Ophthalmol. 1998;116(9):1206-1208.

13. Ye HQ, Maeda M, Yu FS, Azar DT. Differential expression of MT1-MMP (MMP-14) and collagenase III (MMP-13) genes in normal and wounded rat corneas.Invest Ophthalmol Vis Sci. 2000;41(10):2894-2899.

14. Mulholland B, Tuft SJ, Khaw PT. Matrix metalloproteinase distribution duringearly corneal wound healing. Eye (Lond). 2005;19(5):584-588.

15. Lu PC, Ye H, Maeda M, Azar DT. Immunolocalization and gene expression of ma-trilysin during corneal wound healing. Invest Ophthalmol Vis Sci. 1999;40(1):20-27.

16. Williams JM, Fini ME, Cousins SW, Pepose JS. Corneal responses to infection.In: Krachmer JH, Mannis MJ, Holland EJ, eds. The Cornea, Volume I: Funda-mentals of Cornea and External Disease. St Louis, MO: Mosby; 1997:129-162.

17. Fini ME, Cook JR, Mohan R. Proteolytic mechanisms in corneal ulceration andrepair. Arch Dermatol Res. 1998;290(suppl):S12-S23.

18. West-Mays JA, Strissel KJ, Sadow PM, Fini ME. Competence for collagenase geneexpression by tissue fibroblasts requires activation of an interleukin 1 alpha au-tocrine loop. Proc Natl Acad Sci U S A. 1995;92(15):6768-6772.

19. Mohan R, Rinehart WB, Bargagna-Mohan P, Fini ME. Gelatinase B/lacZ trans-genic mice: a model for mapping gelatinase B gene expression during develop-mental and injury-related tissue remodeling. J Biol Chem. 1998;273(40):25903-25914.

20. Mohan R, Sivak J, Ashton P, et al. Curcuminoids inhibit the angiogenic re-sponse stimulated by fibroblast growth factor-2, including expression of matrixmetalloproteinase gelatinase B. J Biol Chem. 2000;275(14):10405-10412.

21. Mohan R, Chintala SK, Jung JC, et al. Matrix metalloproteinase gelatinase B (MMP-9)coordinates and effects epithelial regeneration. J Biol Chem. 2002;277(3):2065-2072.

22. Sivak JM, West-Mays JA, Yee A, Williams T, Fini ME. Transcription factors Pax-6and AP-2� interact to coordinate corneal epithelial repair by controlling expres-sion of matrix metalloproteinase gelatinase B (MMP-9). Mol Cell Biol. 2004;24(1):245-257.

23. Jung JC, Huh MI, Fini ME. Constitutive collagenase-1 synthesis through MAPKpathways is mediated, in part, by endogenous IL-1� during fibrotic repair in cor-neal stroma. J Cell Biochem. 2007;102(2):453-462.

24. Lyu J, Joo CK. Wnt-7a up-regulates matrix metalloproteinase-12 expression andpromotes cell proliferation in corneal epithelial cells during wound healing. J BiolChem. 2005;280(22):21653-21660.

25. Berman M, Dohlman CH, Gnadinger M, Davison P. Characterization of colla-genolytic activity in the ulcerating cornea. Exp Eye Res. 1971;11(2):255-257.

26. Brown D, Chwa M, Escobar M, Kenney MC. Characterization of the major matrixdegrading metalloproteinase of human corneal stroma: evidence for an enzyme/inhibitor complex. Exp Eye Res. 1991;52(1):5-16.

27. Fournie PR, Gordon GM, Dawson DG, Edelhauser HF, Fini ME. Correlations oflong-term matrix metalloproteinase localization in human corneas after success-ful laser-assisted in situ keratomileusis with minor complications at the flap margin.Arch Ophthalmol. 2008;126(2):162-170.

28. Woessner JF Jr. The matrix metalloproteinase family. In: Parks WC, MechamRP, eds. Matrix Metalloproteinases. San Diego, CA: Academic Press; 1998:300-356.

29. Quaranta V, Jones JC. The internal affairs of an integrin. Trends Cell Biol. 1991;1(1):2-4.

30. Gipson IK, Spurr-Michaud S, Tisdale A, Elwell J, Stepp MA. Redistribution of thehemidesmosome components �6�4 integrin and bullous pemphigoid antigensduring epithelial wound healing. Exp Cell Res. 1993;207(1):86-98.

31. Ohji M, SundarRaj N, Hassell JR, Thoft RA. Basement membrane synthesis byhuman corneal epithelial cells in vitro. Invest Ophthalmol Vis Sci. 1994;35(2):479-485.

32. Gipson IK, Inatomi T. Mucin genes expressed by the ocular surface epithelium.Prog Retin Eye Res. 1997;16(1):81-98.

33. Argüeso P, Spurr-Michaud S, Russo CL, Tisdale A, Gipson IK. MUC16 mucin is

expressed by the human ocular surface epithelia and carries the H185 carbohy-drate epitope. Invest Ophthalmol Vis Sci. 2003;44(6):2487-2495.

34. Meredith JE Jr, Fazeli B, Schwartz MA. The extracellular matrix as a cell survivalfactor. Mol Biol Cell. 1993;4(9):953-961.

35. Frisch SM, Francis H. Disruption of epithelial cell-matrix interactions inducesapoptosis. J Cell Biol. 1994;124(4):619-626.

36. Stramer BM, Zieske JD, Jung JC, Austin JS, Fini ME. Molecular mechanisms con-trolling the fibrotic repair phenotype in cornea: implications for surgical outcomes.Invest Ophthalmol Vis Sci. 2003;44(10):4237-4246.

37. LaGier AJ, Yoo SH, Alfonso EC, Meiners S, Fini ME. Inhibition of human cornealepithelial production of fibrotic mediator TGF-�2 by basement membrane–likeextracellular matrix. Invest Ophthalmol Vis Sci. 2007;48(3):1061-1071.

38. Fini ME. Keratocyte and fibroblast phenotypes in the repairing cornea. Prog Re-tin Eye Res. 1999;18(4):529-551.

39. Azar DT, Hahn TW, Jain S, Yeh YC, Stetler-Stevensen WG. Matrix metallopro-teinases are expressed during wound healing after excimer laser keratectomy.Cornea. 1996;15(1):18-24.

40. Chen L, Kato T, Toshida H, Nakamura S, Murakami A. Immunohistochemical char-acterization of epithelial cells implanted in the flap-stroma interface of the cornea.Jpn J Ophthalmol. 2005;49(2):79-83.

41. Latvala T, Barraquer-Coll C, Tervo K, Tervo T. Corneal wound healing and nervemorphology after excimer laser in situ keratomileusis in human eyes. J RefractSurg. 1996;12(6):677-683.

42. Wright JD Jr, Neubaur CC, Stevens G Jr. Epithelial ingrowth in a corneal grafttreated by laser in situ keratomileusis: light and electron microscopy. J CataractRefract Surg. 2000;26(1):49-55.

43. Sun CC, Chang S, Tsai R. Traumatic corneal perforation with epithelial ingrowthafter laser in situ keratomileusis. Arch Ophthalmol. 2001;119(6):907-909.

44. Grupcheva CN, Malik TY, Craig JP, McGhee CN. In vivo confocal microscopy ofcorneal epithelial ingrowth through a laser in situ keratomileusis flap button-hole.J Cataract Refract Surg. 2001;27(8):1318-1322.

45. Anderson NJ, Edelhauser HF, Sharara N, et al. Histologic and ultrastructural find-ings in human corneas after successful laser in situ keratomileusis. ArchOphthalmol. 2002;120(3):288-293.

46. Naoumidi I, Papadaki T, Zacharopoulos I, Siganos C, Pallikaris I. Epithelial in-growth after laser in situ keratomileusis: a histopathologic study in human corneas.Arch Ophthalmol. 2003;121(7):950-955.

47. Asano-Kato N, Toda I, Hori-Komai Y, Takano Y, Dogru M, Tsubota K. Histopatho-logical findings of epithelial ingrowth after laser in situ keratomileusis. Cornea.2005;24(2):130-134.

48. Nagase H, Visse R, Murphy G. Structure and function of matrix metalloprotein-ases and TIMPs. Cardiovasc Res. 2006;69(3):562-573.

49. Waring GO III. Epithelial ingrowth after laser in situ keratomileusis. Am J Ophthalmol.2001;131(3):402-403.

50. McDonnell S, Wright JH, Gaire M, Matrisian LM. Expression and regulation ofstromelysin and matrilysin by growth factors and oncogenes. Biochem Soc Trans.1994;22(1):58-63.

51. Birkedal-Hansen H. Proteolytic remodeling of extracellular matrix. Curr Opin CellBiol. 1995;7(5):728-735.

52. Helena MC, Meisler D, Wilson SE. Epithelial growth within the lamellar interfaceafter laser in situ keratomileusis. Cornea. 1997;16(3):300-305.

53. Probst L, Machat J. Epithelial ingrowth following LASIK. In: Probst L, ed. TheArt of LASIK. Thorofare, NJ: Slack Inc; 1998:427-433.

54. Majo F, Rochat A, Nicolas M, Abou Jaoude G, Barrandon Y. Oligopotent stemcells are distributed throughout the mammalian ocular surface. Nature. 2008;456(7219):250-254.

55. Hanna C, Bicknell DS, O’Brien JE. Cell turnover in the adult human eye. ArchOphthalmol. 1961;65:695-698.

56. Schmack I, Dawson DG, McCarey BE, Waring GO III, Grossniklaus HE, Edel-hauser HF. Cohesive tensile strength of human LASIK wounds with histologic,ultrastructural, and clinical correlations. J Refract Surg. 2005;21(5):433-445.




CLINICAL SCIENCES Correlation Between Epithelial Ingrowth ...€¦ · neas correlates with basement membrane remodeling at the flap margin. To address this question, we performed

Documents