BritishJournal ofOphthalmology 1994; 78: 863-870 863 Improved preservation ofhuman corneal basement membrane following freezing of donor tissue for Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from epikeratophakia

Robert D Young, W John Armitage, Paul Bowerman, Stuart D Cook, David L Easty

Abstract States, good results continue to be achieved by Current methods for the production of the small number ofBritish surgeons performing lenticules for epikeratophakia involve rapid the technique.4 However, no comprehensive freezing, cryolathing, and slow warming of the account of its long term outcome has yet been donor . We have found that this pro- published. cedure causes structural damage to the Several complications resulting in the failure epithelial basement membrane in the donor of epikeratophakia have been reported, includ- cornea which may subsequently contribute to ing infection, graft dehiscence, persistent inter- poor postoperative re-epithelialisation of the face haze or opacity, ulceration, and imperfect implant, leading to graft failure. Endeavouring re-epithelialisation. Among these, the failure of to overcome these problems, the effects of host epithelial cells to migrate over and re- cryoprotection of donor cornea were investi- surface the anterior face of the grafted tissue gated, using dimethyl sulphoxide, in conjunc- continues to be the major reason for the removal tion with different cooling and warming rates ofepikeratophakia lenticules.'-'0 as part of the protocol for cryolathing. The Epithelial healing is itselfa complex phenome- structural integrity of the epithelial basement non involving mitosis of host cells at the graft membrane zone (BMZ) was then assessed by periphery, centripetal migration, and attach- electron microscopy and by immunofluores- ment. In all of these processes the nature of the cence microscopy using antibodies to types IV underlying substratum must be considered to be and VII coliagen, components of the basal of profound importance, as shown by studies of lamina and anchoring fibrils respectively, and cell growth and attachment in vivo and in an antibody to a component of the anchoring vitro." 12 The integrity ofthe epithelial basement filaments. No differences in the pattern of membrane in the cryolathed donor cornea, immunostaining for these components were which becomes the presumptive substrate for detected, indicating that the composition of host epithelial cells, may therefore be an import- the BMZ was unaltered by the different treat- ant factor influencing the recovery of an intact ment regimens applied. However, electron and durable epithelial sheet over the newly http://bjo.bmj.com/ microscopy showed that preservation of base- grafted lenticule. ment membrane ultrastructure was markedly In preliminary observations both of cryo- improved when cornea was warmed rapidly lathed prepared for epikeratophakia and rather than slowly, both in cryoprotected and of failed grafts, we identified several structural non-cryoprotected tissue. Epithelial cell reten- alterations to the epithelial basement membrane tion and preservation of stromal architecture zone of the (BMZ), including disruption basal on September 26, 2021 by guest. Protected copyright. appeared superior in cryoprotected samples, lamina and disorganisation of anchoring fibril while keratocyte structure was heterogeneous complexes and anterior Bowman's layer. These throughout the experimental groups. Further changes appeared to be attributable to freezing University ofBristol, work is in progress to assess the efficacy of rather than to the lathing procedure itself and Department of these protocols in the preservation of kerato- were sufficiently extensive to warrant considera- Ophthalmology, Muscle cyte viability in association with improved and Collagen Research tion as potential inhibitors ofepithelial healing in Group, Langford, Avon basement membrane structure in donor tissue the immediate postoperative period. The few R D Young for epikeratophakia. studies carried out on the effects of freezing (Br3 Ophthalmol 1994; 78: 863-870) corneal tissue the Bristol Hospital, during preparation ofepikera- Lower Maudlin Street, tophakia lenticules have focused mainly on the Bristol, UK viability of stromal keratocytes.'3 1' Factors such W J Armitage Epikeratophakia was originally introduced in as cooling and warming rates and the presence of S D Cook 1980' as a reversible and simplified approach to D L Easty cryoprotective agents (for example, dimethyl corneal refractive for , although sulphoxide), not only affect the survival of cells United Kingdom it soon also found application in the surgical during freezing and thawing, but influence the Transplant Support treatment of keratoconus2 and .3 The Service Authority, location and quantity of ice, and thus the degree Bristol, UK technique involves reshaping a frozen disc of of structural disruption within tissues.'5 We P Bowerman donor cornea on a cryolathe, to form a lenticule, therefore investigated these factors to determine Correspondence to: which is then transplanted onto the de-epithelia- whether the of structural Dr R D Young, Muscle and preservation integrity Collagen Research Group, lised surface ofBowman's layer. ofthe anterior face ofthe prospective graft could Churchill Building, University The results ofepikeratophakia since its incep- be improved. The ultrastructural organisation of of Bristol, Langford, Avon tion have traditionally been considered good,'7 the BMZ was assessed by transmission electron BS18 7DY. and Accepted for publication while, in recent years, the popularity of the microscopy and aspects of its composition were 1 July 1994 procedure may have declined in the United studied by immunofluorescence microscopy 864 Young, Armitage, Bowerman, Cook, Easty

with monoclonal antibodies directed against Treatment oftissue samplesfrom paired human corneas

specific components of the basal lamina and Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from anchoring fibril structures.

Materials and methods

COOLING AND WARMING RATES 4 7 8 3 Rapid cooling Samples were cooled at about 20°C/min to - 50°C on the cryolathe. The tissue sample was attached by its epithelial surface to an aluminium arbour, which was clamped in the cryolathe chuck. (3) rapid cool, slow warm (without lathing); (4) rapid cool, rapid warm. Slow cooling Samples 5-8 were from the right cornea, and Samples were cooled at 1°C/min to -50°C in a were incubated in the cryoprotectant medium: programmable controlled rate freezer (Planer (5) addition and removal of cryoprotectant Kryo 10-16). without freezing; (6) slowcool, rapidwarmwithcryoprotectant; (7) rapidcool, slowwarmwithcryoprotectant; Rapid warming (8) rapid cool, rapid warm with cryopro- Frozen samples were immersed directly into tectant. medium at approximately 22°C (ambient The samples were all processed for microscopy temperature), which gave a warming rate of after these treatments. >50°C/min.

IMMUNOFLUORESCENCE MICROSCOPY Slow warming Samples from each of the corneal sectors (1-8) Frozen samples were allowed to thaw still were mounted in Tissue-tek medium on attached to the cryolathe chuck, which gave a aluminium stubs and frozen by plunging into warming rate of <20'C/min. liquid nitrogen cooled isopentane. Frozen Corneas routinely processed for epikerato- sections, 7 ,tm thick, were cut on a Bright phakia were cooled and warmed on the cryo- cryostat and mounted on slides coated with lathe - that is, rapid cool/slow warm. Biobond adhesive. Sections were exposed to monoclonal antibodies to type IV collagen (goat anti-human), a component of the basal lamina; CRYOPROTECTANT ADDITION AND REMOVAL type VII collagen (LH7.2, mouse anti-human), a http://bjo.bmj.com/ Samples were incubated in 10% (v/v) dimethyl component of the anchoring fibrils; or to an sulphoxide (DMSO) in Eagle's minimal essential antibody to a component of the anchoring fila- medium (MEM) for 15 minutes at 22°C. The ments (LH39, mouse anti-human). Type IV cryoprotectant was removed by dilution by antibody was an affinity purified reagent sup- transferring corneal tissue into 0-5 mol/litre plied by Southern Biotechnology Associates. sucrose in MEM at 22°C. The sucrose acted as an LH7.2 and LH39 were generously donated by osmotic buffer to limit the cell that Professor I Leigh, Department of Experimental swelling on September 26, 2021 by guest. Protected copyright. would have been induced by the reduction in Dermatology, London Hospital Medical DMSO concentration.'6 After 15 minutes, the College. After 2 h incubations in the primary samples were placed into MEM alone. monoclonal antibody and washing in 1% bovine serum albumin (BSA) in phosphate buffered saline (PBS), or in PBS/BSA alone in the case EXPERIMENTAL TREATMENTS of control preparations, the sections were The corneas were obtained within 23 h post incubated with secondary antibodies conjugated mortem from a 60-year-old man, with no to fluorescein isothiocyanate (FITC), followed ocular disease, who -had died from Hodgkin's by further washing and mounting in Citifluor. lymphoma, which is a medical contraindication They were then viewed using a Leitz microscope to transplantation of the corneas. Corneoscleral equipped for FITC fluorescence and phase discs were excised and placed in Eagle's MEM. optics. The corneas were then dissected free of sclera, cut into four equal quadrants, and subjected to the following treatments (summarised in the ELECTRON MICROSCOPY diagram): Samples from each treated corneal sector were Samples 1-4 were from the left cornea, and fixed at room temperature for 2-3 h in 2 5% were not incubated in the cryoprotectant glutaraldehyde in 0 1 M sodium cacodylate medium: buffer containing 4 mM calcium chloride at pH (1) processed for microscopy without further 7-2. After rinsing in buffer, they were post-fixed treatment; in buffered 1% osmium tetroxide, stained en (2) rapid cool, lathing to 0 35 mm thickness, bloc in 0-5% aqueous uranyl acetate and slow warm (that is, similar to routine processing dehydrated in an ethanol series. Sections were for epikeratophakia); prepared from the Araldite embedded samples, Improvedpreservation ofhuman corneal basement membranefollowingfreezing ofdonor tissueforepikeratophakia 865 Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from

tf

Vk~~~~~~~~

$~~~~~~~~~~~~~~~~~

.4~~~~~~~~~~~~4

i..!#A,~~ ~ ~ ~ ~ #

A.~~~~~~~~~~~~~~~~~"if http://bjo.bmj.com/ JW,~.#-rf~ ~ ~ ~ ~ -

*+4i, '.9" 7 ...~~;~ 4*W~~~~~~~~~~~~~~~~~~W

'i*rW.. ~~ 4%r~~~~~~~~~~~~~~~~~~~4 on September 26, 2021 by guest. Protected copyright.

Figure 1 Immunolocalisation ofhuman corneal basement membrane zone components. Examples ofhuman cornea exposed to different procedures in relation to cryoprotection and rates ofcooling and warming. Comparative phase (left) andfluorescence (right) microscopy images showv (a) cornea, without cryoprotection (sample 4), exposed to rapid cool and rapid warmn, stainedfor type IV collagen; (b) cryoprotected cornea, exposed to rapid cool and rapid warmn (sample 8), stained for type VII collagen; (c) cryoprotected cornea, exposed to rapid cool and slowv warm (sample 7), stained with LH39for anchoringfilaments; (d) control preparation: cryoprotected cornea, exposed to rapid cool and rapid warmn (sample 8), incubated without primary antibody. 866 Young, Armnitage, Bowerman, Cook, Easty

contrasted in uranyl acetate and lead citrate for TRANSMISSION ELECTRON MICROSCOPY examination in a Philips 400 transmission

electron microscope at 80 kV. Without cryoprotection Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from The epithelial basement membrane in the untreated corneal sample conformed to earlier Observations descriptions of this zone.'7 Cell membranes of the basal epithelial cells were separated from an PHASE AND IMMUNOFLUORESCENCE MICROSCOPY amorphous, sheet-like basal lamina by a clear Corneal sections from all experimental categories lamina lucida and were punctuated by mem- revealed strongly positive fluorescent signals brane condensations, hemidesmosomes, from towards all three monoclonal antibodies which numerous fine anchoring filaments employed. Localised staining was visible as a extended proximally to attach to the basal continuous linear signal corresponding to the site lamina. This structure in turn was associated of the epithelial basement membrane, irrespec- with Bowman's layer via many anchoring fibrils tive of whether or not the epithelium was pre- (Fig 2a), within the lamina fibroreticularis. The served intact. Epithelial detachment was more BMZ was often denuded of epithelium in non- common in samples not incubated with the cryoprotected cornea after cooling and re- cryoprotectant (Fig la). The antibody directed warming. In addition, cornea exposed to rapid against type IV collagen was detected as a fine cooling and slow warming (Figs 2b and c), line of fluorescence, while that against type VII showed extensive disorganisation of the BMZ, collagen and LH39 appeared as a broader, including tears and breaks in the basal lamina, intense signal coincident with the distal face of condensation ofanchoring fibrils, and deposition Bowman's layer (for example, Figs la-c). of cellular and membranous debris, presumably Control sections where the primary antibody was the residue of damaged epithelial cells over the omitted, in each case were completely negative distal surface. In contrast, non-cryoprotected (Fig Id). There was no clear difference either in cornea cooled rapidly and subsequently warmed the distribution of the signal or in its intensity rapidly (Fig 2d) displayed markedly improved between cryoprotected or non-cryoprotected BMZ preservation in terms ofthe organisation of corneal samples. In addition, no individual the basal lamina and anchoring fibrils. Epithelial differences could be assigned to corneal quad- retention, in common with other samples denied rants receiving different treatments within these cryoprotection, was none the less poor in this two groups. specimen. http://bjo.bmj.com/ on September 26, 2021 by guest. Protected copyright.

C d

'igure 2 Electron microscopy of'basement membrane zone (BMZ) in human cornea without cryoprotection, treated as follows: (a) no cooling (sample 1); (b) rapid cool, cryolathed, and slow warm (sample 2); and (c) rapid cool and slow warm (sample 3), showing loss ofepithelium, breaks in basal lamina (arrows) and condensation ofanchoringfibrils. (d) Rapid cool and rapid warm (sample 4), showing improved BMZ preservation. E, epithelium; Bl basal lamina; H, hemidesmosomes; L, lamina lucida; a, anchoringfilaments; af, anchoringfibrils; B, Bowman's layer. Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from

Figuren e .e..c.ano.i.new r p c n r d lo ( o o

<] X g3 t...... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Figure~~~~~~~~~~~~~~~~~~

?;',t''' < < , ".' StiL S,3 X',', ;'~~~~~~~~~~~~~~~~~~~~~zlIX'' cassno structural alteration to the BMZ; (b) slow c?yoprotectionappearance ofbasal lamina;obvious(c) rapid cool and slow warm (sample 7); (d) rapidcontrolledcool andcoolrapidand rapidwarmwarmin mediumin medium(sample(sample8): optimal6): diffusepreservationand interruptedofBMZ. E, epithelium; Bl, basal lamina; H hemdesmosomes; L, lamina lucida; a, anchoringfilaments; af, anchoringfibrils; B, Bowman'slayer http://bjo.bmj.com/ on September 26, 2021 by guest. Protected copyright.

Figure 4 Electron microscopy ofhuman corneal keratocytes and stroma. Stromal keratocytes (a) in non-cryoprotected cornea, after rapid cool, cryolathing and slow warm (sample 2); and (b) in cryoprotected cornea after rapid cool and rapid warmn (sample 8), showing nuclear crenation, vacuolation and rupture ofcellular organelles. Stromal lamellae are disorganised with increased collagen interfibrillar distance and intralamellar lakes in (c) non-crvoprotected cornea after rapid cool, cryolathing, and slow warm (sample 2), while in cryoprotected cornea (d), after rapid cool and rapid warm (sample 8), stromal integrity is well Bar= 1 5 p.m (a) and (b); 0 5 p.m, (c) and (d). preserved. 868 Young, Armitage, Bowerman, Cook, Easty

With cryoprotection poor re-epithelialisation remain obscure; but

Addition and subsequent removal of 10% (v/v) epithelial complications are not common with Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from DMSO without freezing appeared to cause no penetrating or lamellar keratoplasty. Problems structural alteration to the BMZ (Fig 3a). The with both host and donor tissue, such as organisation ofanchoring fibrils and filaments in manipulation of the recipient cornea or differ- association with the basal lamina and hemi- ences in viability of the graft tissue, have been desmosomes was comparable with that seen in put forward in explanation for epithelial insuf- cornea not exposed to the cryoprotectant (Fig ficiency.8 The steepened at 2a). Epithelial retention in cryoprotected corneal the host-graft interface, which may impede cell samples was found to be generally better than in migration and expose loosely attached cells to the corneas frozen without cryoprotectant, although action of the has also been considered. some sites were located where loss or separation Steps in the preparation of lenticules, including of epithelium was evident (Figs 3b and 3d). In freezing, lathing, lyophilisation, and the use of cryoprotected cornea, cooled rapidly and the corneal press have all been implicated in warmed slowly (Fig 3c), distortion of the basal causing structural damage to the lenticule which lamina was observed, reminiscent of that seen in might inhibit the establishment of a normal BMZ in samples frozen in the same way but functional epithelium.8 without cryoprotectant (Figs 2b and c). Epithelial healing is closely associated with the Slow cooling in the presence of cryopro- renewal of the BMZ, a complex structural inter- tectant, followed by rapid warming (Fig 3b), face between the basal cells and underlying reduced the gross distortion of the BMZ and Bowman's layer, including components mediat- avoided the changes in the anchoring fibril ing attachment of the cells to their substratum. component that were seen in cornea cooled Lenticules removed postoperatively owing to rapidly and warmed slowly (Figs 2b, 2c, 3c). persistent epithelial defects exhibited a marked However, in this case the basal lamina appeared reduction in numbers of cell attachment struc- diffuse and interrupted, presumably in response tures, hemidesmosomes, and basal lamina. 19 to the deposition of large ice crystals during slow Factors influencing the deposition ofa functional cooling. The most favourable preservation of basal lamina and the reformation of hemi- BMZ structure was identified in cryoprotected desmosomes in healing epithelia have been little cornea subjected to a regimen of rapid cool- studied. However, the adhesion, migration, and ing and rapid warming (Fig 3d). Here, all mitosis of epithelial cells are known to be components of the BMZ appeared intact enhanced by basement membrane components, although at sites of epithelial detachment, a fine including type IV collagen. 12 The elegant studies deposit of basal cell membrane, including hemi- of Gipson et al20 on BMZ renewal in recombined desmosomes and anchoring filaments, could be rabbit corneal epithelial sheets and denuded observed attached to the basal lamina (Fig 3d). stroma in vitro showed that the sites of residual anchoring fibril insertions into the basal lamina may represent nucleation sites for the formation Keratocytes and stroma new of hemidesmosomal attachments in the http://bjo.bmj.com/ A wide range of morphological changes were healing epithelial cells. However, subsequent displayed by keratocytes in response to different studies on the sequence of reassembly of the cooling and warming procedures, both with or BMZ under different experimental conditions without cryoprotection. Cytoplasmic vacuola- have not found universal agreement."') Hemi- tion, disruption ofcell membrane and organelles, desmosomes may precede anchoring fibril together with nuclear crenation and aggregation formation in situations where the basal lamina

of nuclear chromatin could be located in cells of and superficial stroma are absent or damaged,2' on September 26, 2021 by guest. Protected copyright. all categories (for example, Figs 4a and b). and the presence of type VII collagen may not be Isolated cells with comparatively normal ultra- essential for hemidesmosome formation per se structural features were also present. because sheep with the disease dystrophic Observations on collagen fibrils in stromal epidermolysis bullosa, where type VII collagen is lamellae were made in all eight experimental not expressed, have intact hemidesmosomes. categories. In general, rapid cooling and rapid The potential for epithelial cells to attach and warming ofcryoprotected stroma gave rise to less restore a surface layer, even on suboptimal disruption of lamellae and of the component substrates is clearly demonstrated by these pre- fibrillar organisation, than that produced by vious studies. It seems reasonable to expect the rapid cooling and slow warming with cryoprotec- process to be greatly facilitated by the provision tion (for example, Figs 4c and d). of an intact undamaged basal lamina for new epithelial cell growth. Yet the effects of freezing on structural integrity of epikeratophakia Discussion lenticules have not previously been assessed in The major clinical complications associated with detail. We found that current methods of epikeratophakia have been related to the failure cryolathing involving rapid cooling followed by of host epithelial cells to migrate over and slow warming, without cryoprotection, gave rise establish a normal epithelium on the graft to considerable distortion ofthe BMZ, with tears surface.8-'0 Even when the epithelium is re- in the basal lamina and condensation of anchor- established, specular microscopy has shown that ing fibrils. Fine structural damage to the surface there may still be abnormalities in cell shape and of the lenticule can therefore be attributed organisation which may persist up to 16 months specifically to the freezing and warming pro- after surgery.8 Those aspects of the epikerato- cedures rather than the lathing process itself. phakia procedure that are responsible for These modifications to BMZ fine structure may Improvedpreservation ofhuman corneal basement membranefollozvingfreezingofdonor tissueforepikeratophakia 869

be important contributory factors in the inhibi- tection on keratocyte viability have yet to be tion of epithelial attachment and re-surfacing. determined. Previous authors have reported Our studies show that the integrity ofthe graft improved keratocyte viability following slow Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from surface can be markedly improved by rapid cooling in the absence of cryoprotection,29 yet rather than slow warming of frozen cornea. This our studies showed a slow cooling rate to be was achieved by freezing the cornea in situ on detrimental to the preservation of the basal the cryolathe and then quickly transferring the lamina. Used in combination with rapid cooling frozen cornea to tissue culture medium at 22°C. and warming rates, which we found favour BMZ Rapid warming would have allowed less time for preservation, cryoprotective agents may permit structural disruption through recrystallisation of the retention both of surface morphology and ice, the process whereby larger ice crystals grow keratocyte viability in the prospective graft. at the expense of smaller, and thus less thermo- Further work is in progress to assess the efficacy dynamically stable, crystals. In addition, the of these protocols for the improvement of graft presence ofantibody binding to type IV and type tissue for epikeratophakia in order to overcome VII collagen and to anchoring filaments demon- problems of re-epithelialisation. strates the compositional integrity of basal and fila- This work was supported by a project grant from the Guide Dogs lamina, anchoring fibrils, anchoring for the Blind Association of the United Kingdom. The authors are ments following this treatment. Type IV and indebted to Professor I M Leigh for generously donating mono- VII in the basal lamina and clonal antibodies against BMZ components and to Mrs A C type collagens Phillips for skilled assistance with cryostat sectioning and fluor- anchoring fibrils, together with laminin, escence microscopy. nidogen, and heparan sulphate proteoglycan in the lamina lucida of the BMZ have all been 1 Kaufman HE. The correction of aphakia. Am J Ophthalmol identified as potential ligands for the attachment 1980; 89:1-10. 2 Kaufman HE, Werblin TP. Epikeratophakia for the treatment of basal epithelial cells." Immunogold electron of . AmJ Ophthalmol 1982; 93: 342-7. demonstrated that the 3 McDonald MB, Klyce SD, Suarez H, Kandarakis A, microscopy anchoring Friedlander MH, Kaufman HE. Epikeratophakia for filament antibody used in this study showed myopia correction. Ophthalmology 1985; 92: 1417-22. towards sites in the lamina lucida 4 McDonald MB, Kaufman HE, Aquavella JV, Durrie DS, specificity Hites DA, Hunkeler ID, et al. The nationwide study of below hemidesmosomes, corresponding to the epikeratophakia for aphakia in adults. Am J Ophthalmol filaments.23 the 1987; 103:358-65. anchoring Although precise 5 McDonald MB, Kaufman HE, Aquavella JV, Durrie DS, epitope for this antibody is as yet unknown, it Hites DA, Hunkeler ID, et al. The nationwide study of may be related to a discovered epikeratophakia for myopia. Am J Ophthalmol 1987; 103: epiligrin, recently 375-83. glycoprotein which appears to be an important 6 McDonald MB, Kaufman HE, Durrie DS. Epikeratophakia cell adhesion for and other for keratoconus, the nationwide study. Arch Ophthalmol ligand epidermal 1986; 104: 1294-300. epithelial cells via the integrin a3,B1 expressed on 7 Halliday BL. Epikeratophakia for aphakia, keratoconus and their cell membranes.24 myopia. BrJ Ophthalmol 1990; 74: 67-72. 8 Binder PS, Zavala EY. Why do some epikeratoplasties fail? The improvement in structure of the BMZ Arch Oihthalmol 1987; 105: 63-9. 9 Binder PS, Baumgartner SD, Fogle JA. Histopathology of a brought about by rapid warming was observed case of epikeratophakia (aphakic epikeratoplasty). Arch irrespective of the presence of cryoprotective Ophthalmol 1985; 103: 1357-63. agent. On the other hand, retention of epithe- 10 Rodrigues M, Nirankari V, Rajagopalan S, Jones K, Funderburgh J. Clinical and histopathologic changes in the http://bjo.bmj.com/ lium and integrity of stromal lamellae did appear host cornea after epikeratoplasty for keratoconus. Am J Ophthalmol 1992; 114: 161-70. to be better when corneal tissue was frozen in the 11 Yurchenco PD, Schittny JC. Molecular architecture of base- presence of DMSO. Such a result would not be ment membranes. FASEBJ 1990; 4:1577-90. not 12 Olivero DK, Furcht LT. Type IV collagen, laminin and unexpected since only would the DMSO fibronectin promote adhesion and migration of rabbit lens protect the epithelial cells against freezing epithelial cells in vitro. Invest Ophthalmol Vis Sci 1993; 34: 2825-34. injury, but less ice would have formed in the 13 Rich LF, Friedlander MH, Kaufman HE, Granet N. Kerato- stroma owing to the colligative effect of the cyte survival in keratophakia lenticules. Arch Ophthalmol of the These 1981; 99: 677-80. on September 26, 2021 by guest. Protected copyright. presence cryoprotectant. aspects 14 Lee TJ, Wan WL, Kash RL, Kratz KL, Schanzlin DJ. warrant further investigation because cryopro- Keratocyte survival following a controlled-rate freeze. Invest tection of Ophthalmol VisSci 1985; 26: 1210-5. tissue for refractive keratoplasty has 15 Mazur P. Freezing of living cells: mechanisms and implica- previously been discouraged, mainly as a result tions. AmJ Physiol 1984; 247: c125-42. of the deleterious effects of on 16 Leibo SP, Mazur P. Methods for the preservation of mam- putative glycerol malian embryos by freezing. In: Daniel JC, ed. Methods of endothelial cells in the recipient cornea.25 The mammalian reproduction. New York: Academic Press, 1978: of to facilitate 179-201. original purpose cryoprotection, 17 Gipson IK, Spurr-Michaud SJ, Tisdale SJ. Anchoring fibrils the survival of keratocytes during freezing, was form a complex network in human and rabbit cornea. Invest Ophthalmol VisSci 1987; 28: 212-20. questioned because viability of donor kerato- 18 Rao GN, Grant S, Aquavella JV. Specular microscopy of the cytes was considered not to be essential for the corneal epithelium after epikeratophakia. AmJ Ophthalmol survival of the could be 1987; 103: 392-6. graft, although they 19 Azar DT, Spurr-Michaud SJ, Tisdale AS, Moore MB, Gipson important in relation to stromal integrity. More- IK. Reassembly of the corneal epithelial adhesion structures following human epikeratoplasty. Arch Ophthalmol 1991; over, keratocyte epithelial cell interactions are 109: 1279-84. now thought to be important in the homeostasis 20 Gipson IK, Grill SM, Spurr SJ, Brennan SJ. Hemidesmosome of the BMZ.26 The death of keratocytes from formation in vitro. J Cell Biol 1983; 97: 849-57. 21 Stock EL, Kurpakus MA, Sabol B, Jones JC. Adhesion freeze injury during the preparation oflenticules complex formation after small keratectomy wounds in the may result in the absence of viable cells from the cornea. Invest Ophthalmol Vis Sci 1992; 33: 304-13. 22 Gipson IK, Spurr-Michaud SJ, Tisdale SJ, Keough M. graft stroma for months after surgery, before Reassembly of the anchoring structures of the corneal slow takes cell epithelium during wound repair in the rabbit. Invest repopulation place by migration Ophthalmol Vis Sci 1989; 30: 425-34. from the recipient cornea.'327 Cryoprotective 23 Almeida BM, Challacombe SJ, Eveson JW, Smith CG, Leigh would be to facilitate IM. A novel lamina lucida component of epithelial and agents expected keratocyte endothelial basement membranes detected by LH39 mono- survival as demonstrated recently.2" In our clonal antibody. J Pathol 1992; 166: 243-53. structure was hetero- 24 Carter WG, Ryan MC, Gahr MI. Epiligrin, a new cell adhesion study, however, keratocyte ligand for integrin t3rdl in epithelial basement membrane. geneous in all groups and the effects of cryopro- Cell 1991;65: 599-610. 870 Young, Armitage, Bowerman, Cook, Easty

25 Binder PS, Zavala EY, Deg J, Akers PH. Refractive kerato- of keratophakia. Arch Ophthalmol 1982; 100: 101-8. plasty. Tissue dyes and cryoprotective solutions. Arch 28 Kratz-Owens KL, Huff JW, Schanzlin DJ. New cryopro-

Ophthalmol 1983; 101: 1591-6. tectant for cryorefractive surgery. Cataract Refract Surg Br J Ophthalmol: first published as 10.1136/bjo.78.11.863 on 1 November 1994. Downloaded from 26 Marinkovich MP, Keene DR, Rimberg CS, Burgeson RE. 1991; 17: 608-12. Cellular origin of the dermal-epidermal basement 29 Lee TJ, Wan WL, Kash RL, Kratz KL, Schanzlin DJ. membrane. Developmental Dynamics 1993; 197: 255-67. Keratocyte survival following a controlled-rate freeze. Invest 27 Binder PS, Beale JP, Zavala EY. The histopathology of a case Ophthalmol Vis Sci 1985; 26: 1210-5. http://bjo.bmj.com/ on September 26, 2021 by guest. Protected copyright.