Veterinary Ophthalmology (2015) 18, 3, 191–197
DOI:10.1111/vop.12109
Lamellar keratoplasty in rabbits using human and rabbit amniotic membrane grafts: a comparative study Cintia A. L. Godoy-Esteves,* Jose Alvaro P. Gomes,† Karina Yazbek,‡ Jose L. Guerra§ and Paulo S. M. Barros* *Laboratory of Investigation on Comparative Ophthalmology, School of Veterinary Medicine, University of S~ao Paulo, S~ao Paulo, SP, CEP 02478-000, Brazil; †Sector of Corneal and External Diseases, School of Medicine, Federal University of S~ao Paulo, S~ao Paulo, Brazil; ‡Agener Sa ude Animal, S~ao Paulo, Brazil; and §Department of Pathology, School of Veterinary Medicine, University of S~ao Paulo, S~ao Paulo, Brazil
Address communications to: C. A. L. Godoy-Esteves Tel.: 55 11 3981-3229/ 55 11 99949-5949 e-mail: cintiagodoy@terra. com.br
Abstract Objective To compare the behavior of human and rabbit amniotic membrane (AM) grafts in surgically induced corneal thinning in rabbits. Animals studied Thirty two NZWR were randomly assigned to two groups of 16 animals each according to AM type (Human AM: group HAM and Rabbit AM: group RAM). Procedure All animals were submitted to right keratectomy at a depth of 0.1 mm using a 5 mm trephine. Animals from HAM group had a button of 5 mm of human AM sutured into the corneal bed with a continuous pattern and 10.0 nylon monofilament suture, while animals from the RAM group had a button of 6 mm diameter of rabbit AM. Four animals in each group were euthanized 2, 7, 15, and 30 days postoperatively. Their corneas were harvested, fixed in 2% glutaraldehyde solution, and stained with haematoxylin–eosin, picrosirius red, and alcian blue for evaluation under light optical microscopy. Microscope images were digitalized and inflammatory cells and stromal blood vessels were counted. Results There were no clinically significant differences between groups, and complete corneal epithelialization was observed in all animals in 30 days. Light optical microscopy revealed AM incorporation and resorption in both groups. However, the number of inflammatory cells and blood vessels was significantly higher in group HAM than in group RAM (P < 0.05, Mann–Whitney test). Clinical responses to human or rabbit AM were similar; however, human AM induced greater inflammatory reaction and stromal neovascularization in the rabbit cornea than in rabbit AM. Conclusion These differences may reflect a potential reaction to the xenograft. More studies are needed to further characterize these findings. Key Words: amniotic membrane, cornea, graft, human, rabbit
INTRODUCTION
Human amniotic membrane (HAM) has gained popularity as a successful adjunct in ocular surface reconstructive surgery over the last decade. Most of the effects of amniotic membrane (AM) transplantation on the ocular surface have been determined through research in rabbits.1–12 Published studies present conflict regarding the HAM potential to induce immune reactions.13 Evidence shows that HAM may behave as an immunologically privileged tissue despite the presence of viable cells expressing HLA-G (major histocompatibility © 2013 American College of Veterinary Ophthalmologists
antigen) in this cryopreserved tissue. However, the HAM may have the potential to induce reactions to the transplanted tissue that may interfere with the interpretation of results. In addition, the effects of rabbit AM (RAM) transplantation and the response of rabbits to xenografts and allografts have received little attention. RAM has been used in rare occasions.8,12 The lack of comparative data between HAM and RAM grafts has motivated this investigation. The aim of this study was to determine whether the use of a xenograft transplantation using HAM in rabbit corneas has the potential to cause adverse host tissue
192 godoy-esteves
ET AL.
reactions when compared to the use of an allograft using RAM. The authors compared the following parameters: clinical and histological (haematoxylin–eosin stain) characteristics, determination of proteoglycan distribution and characterization (alcian blue stain), collagen type (picrosirius red stain) and inflammatory cells, and newly formed stromal vessels. MATERIALS AND METHODS
Harvesting of the amniotic membrane The HAMs employed in this study were donated by the Laboratorio de Doencßas Externas Oculares (External Ocular Disease Laboratory), Sector of Corneal and External Diseases, Department of Ophthalmology – UNIFESP. Human placentas were harvested following cesarean sections and washed with 0.9% saline solution under aseptic conditions. The AM was bluntly dissected from the remaining fetal membranes and flattened onto sterile nitrocellulose filter paper (Micropore, Bedfort, MA, USA) with the epithelial surface up. The AM with the filter was washed with a phosphate-buffered saline solution containing penicillin (1000 U/mL), streptomycin (20 lg/mL) and amphotericin B (2.5 lg/mL), cut into approximately 5 9 5 cm fragments and placed in a sterile container with 1:1 glycerol and corneal preservation medium (Baxter Healthcare corporation, Stone Mountain, GA, USA, and Ophthalmos, S~ao Paulo, SP, Brazil, respectively) and frozen at 80 °C.14 Rabbit placentas were harvested following ovariosalpingohysterectomy performed on 28th day of pregnancy. Following separation of the AM from the remaining fetal membranes, the same protocol was employed for AM preparation, except for the size of the fragments that corresponded to 3 9 3 cm. The low thickness and elasticity of the rabbit placenta made the manipulation difficult. Extreme care was required to avoid rupture and curling up of the membrane that would prevent its use. Animals Thirty-two young adult New Zealand rabbits weighing between 1.5 and 2.0 kg obtained from privately owned rearing facilities were used, conforming to Association for Research in Vision and Ophthalmology (ARVO) guidelines for research involving animal models. Animals were randomly assigned to two groups of 16 individuals for evaluation of each membrane type (group HAM and group RAM, respectively). Animals in each group were again randomly assigned to four subgroups with different follow-up periods (2, 7, 15, and 30 days, respectively). Anesthetic procedure Animals were fasted for 2 h previous to sedation with a combination of acepromazin–meperidin (0.4 mg/kg and 10 mg/kg, respectively) given intramuscularly and instillation of anesthetic eye drops containing proparacaine
hydrochloride. Following a 15-min lag phase, anesthesia was induced with a combination of ketamine and midazolam (10 mg/kg and 0.3 mg/kg, respectively) given intravenously (IV) and maintained with IV ketamine (5–10 mg/ kg) administered as necessary. Venous infusion with lactated Ringer’s solution was maintained throughout the surgical procedure, and flunixin meglumine (1 mg/kg) was given IV. The marginal ear vein was used for all IV medications.
Surgical procedure Lamellar keratectomy was performed using an operating microscope (DF Vasconcelos, Brasil) with 109 magnification. A 5 mm Castroviejo trephine was used to penetrate the cornea to a depth of 0.1 mm in the dorsotemporal area, 1 mm from the limbus, creating a recipient bed for implantation of HAM or RAM grafts. HAM and RAM grafts were created using a trephine (5 and 6 mm, respectively). Grafts (the epithelium side up) were sutured into the recipient bed with a continuous pattern and 10-0 nylon monofilament suture on a spatulated needle. Animals were kept in individual cages following surgery and treated with tobramycin eye drops three times daily for the first 2 weeks postsurgery. As signs of pain were not observed, no analgesics or Elizabethan collar was used. Animals were euthanized 2, 7, 15, or 30 days postsurgery. Following deep anesthesia with sodium thiopental, cardiorespiratory arrest was induced with IV 19.1% potassium chloride. Clinical assessment Animals were evaluated using a portable slit-lamp biomicroscope (SL-14 Portable Slit Lamp, Kowa, Tokyo, Japan) at 24-h intervals during the first week postsurgery and every 72 h thereafter until euthanasia according to the respective subgroup follow-up period. Clinical assessment was based on the presence of photophobia, blepharospasm, ocular discharge (occurrence and characteristics), conjunctival hyperemia, corneal opacity, new vessel formation and graft retention and viability. The changes observed were subjectively graded as (0) absent, +(1) mild, ++(2) moderate, and +++(3) severe. Microscopic evaluation A corneal button was obtained by full thickness trephination at the corneoscleral transition, fixed in 2% buffered glutaraldehyde solution, processed, and stained with HE, picrosirius red, and alcian blue for optical microscopy. Slides from the recipient area were used for evaluation of the morphological patterns related to the grafting procedure. Apart from morphological patterns related to grafting, HE-stained slides were also used for counting of inflammatory cells and vessels in the corneal stroma using a Nikon E-800 microscope (Nikon Metrology, Inc.,
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
keratoplasty in rabbits using amniotic membrane 193
Brighton, MI, USA) coupled with Image-Pro Plus software (Media Cybernetics, Inc., Rockville, MD, USA). The recipient area was divided randomly into smaller subareas of 35401.2 lm2 each, and five of these smaller subareas (177 006 lm in total) distributed along the recipient area were examined under 209 magnification. Values corresponding to the number of inflammatory cells and blood vessels counted in different animals were submitted to the Mann–Whitney nonparametric test. The level of significance was set at 5% (P < 0.05). RESULTS
Clinical assessment Human amniotic membrane and RAM grafts induced similar clinical responses: Blepharospasm: Three animals presented with mild signs in both groups; Conjunctival discharge: In group HAM, three animals presented with mild signs, eight with moderate signs, and two with severe signs. In group RAM, two animals presented with mild signs, one with moderate signs, and two with severe signs distributed through all time periods; and Conjunctival hyperemia: In group HAM, five animals presented with mild signs, five with moderate and two with severe signs. All were distributed through all periods. In group RAM, only five animals had hyperemia (four were mild in 2 days, and one was moderate in 30 days). Mild-to-moderate corneal opacity was noted in most animals in group HAM (32 with mild in 2-day period, and in the final period – 30 days, one with mild, two with moderate and one with severe). Mild corneal opacity was observed in all animals in group RAM throughout the follow-up period. Corneal vascularization was not observed clinically in the RAM group during the first 7 days postsurgery, while in HAM group, it was clinically obvious at day 2 in two animals (one with mild and one with moderate). Corneal vascularization progressed over time and at day 15 was observed in the HAM group as mild in one animal, as moderate in five, and as severe in one. At day 30, two animals had developed moderate vascularization and two had developed severe vascularization. Vascularization was also observed at day 15 in the RAM group as mild in five animals and as moderate in three. At day 30, one had developed mild, two moderate, and one severe signs. Graft retention and viability were observed throughout the follow-up period in both groups. Microscopy Two days postimplantation, the implanted tissue was visible under light microscopy as abundant acidophilic amorphous extracellular matrix in both groups. Inflammatory cell migration and invasion, consisting of polymorphonuclear cells, were observed immediately deep to the implanted tissue. A single or double layer of flat epithelium covered the implanted tissue.
At day 7, in both groups, the implanted tissue was integrated into the autochthonous tissue and remnants consisting of amorphous acidophilic material were visible. The corneal surface was completely covered with stratified flat epithelium. Many cells infiltrated the recipient area, and under greater magnification, cells with fibroblast-like morphology could be observed in the implanted tissue, indicating proliferation of these cells over the substrate represented by the implanted tissue. At day 15, areas of discontinuous amorphous acidophilic material corresponding to the integrated HAM were visible in group HAM (Fig. 1). The corneal surface was completely covered with stratified paved epithelium. A large number of inflammatory cells, consisting of polymorphonuclear cells, and cells with fibroblast-like morphology were present in the tissue underlying the graft. A negative image corresponding to the suture material surrounded by intense infiltrate of polymorphonuclear cells was noted at the periphery of the histological section. At day 15, in group RAM (Fig. 2), the AM was not as clearly defined as in group HAM. A thin band of amorphous acellular material possibly corresponding to remnants of the RAM graft was observed deep to the epithelial layer. The epithelium was classified as paved stratified, and numerous inflammatory cells, mostly polymorphonuclear cells, and cells with fibroblast-like morphology were noted in the stroma underlying the recipient area. Epithelial hyperplasia was noted over the operated area in some specimens. At day 30, the AM was integrated and well defined under paved stratified epithelium in both groups. Intense inflammatory cell infiltrate, characterized by polymorphonuclear cells in most animals, was observed in the stroma underlying the recipient area in group HAM (Fig. 3). In group RAM (Fig. 4), remnants of the RAM, cells with
Figure 1. Photomicrograph of the histological section of animal No. 12 corresponding to the area treated with human amniotic membrane graft 15 days post-implantation. Perfect integration of the implanted tissue (arrow) into the autochthonous tissue and epithelial regeneration are visible. The suture material surrounded by inflammatory infiltrate can be noted. HE staining; 49 magnification.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
194 godoy-esteves
ET AL.
Figure 2. Photomicrograph of the histological section of animal No. 7 corresponding to the area treated with rabbit amniotic membrane graft 15 days post-implantation. Epithelial regeneration, a negative image of the suture material surrounded by inflammatory infiltrate and the presence of inflammatory cells in the stroma can be noted. A button of rabbit amniotic membrane (arrow) is visible below the paved stratified epithelium. HE staining; 209 magnification.
Figure 3. Photomicrograph of the histological section corresponding to the area treated with human amniotic membrane graft 30 days post-implantation. The membrane is visible below the paved stratified epithelium. Intense inflammatory infiltrate can be noted in the surrounding stroma. HE staining; 109 magnification.
fibroblast-like morphology and moderate inflammatory infiltrate, also characterized by polymorphonuclear cells in most animals, were noted in the stroma above and below the implanted tissue. In the sections stained with picrosirius red and evaluated under polarized light, AMs of both human and rabbit were visible in orange-red color, suggesting the presence of Type I collagen. In the histological sections stained with alcian blue, the AM was visible as a pinkish material in both groups. Areas where collagen remodelled were likely stained blue, suggesting the presence of acid glycosaminoglycans. Blood vessel counting in the stroma adjacent to the grafted area was higher in group HAM than in group RAM (Fig. 5). The number of blood vessels was significantly (P < 0.05) higher in group HAM than in group
Figure 4. Photomicrograph of the histological section of animal No. 28 corresponding to the area treated with rabbit amniotic membrane graft 30 days post-implantation. Remnants of amorphous material corresponding to the grafted membrane are visible (arrow). Note the presence of cells with fibroblast-like morphology and inflammatory cells in the stroma. HE staining; 109 magnification.
RAM at 15 and 30 days. In group HAM, the number of corneal blood vessels was significantly (P < 0.05) higher in animals in the 30-day follow-up period subgroup than in the remaining subgroups. In group RAM, the number of corneal blood vessels did not differ significantly (P > 0.05) between subgroups. Under microscopic examination, blood vessel counting was significantly higher in group HAM. Throughout the follow-up, in most animals, the inflammatory response was characterized by polymorphonuclear cells. Four animals in each group had a mixed inflammatory response involving both mononuclear and polymorphonuclear cells. However, giant cells were not observed in any of the samples. In groups HAM and RAM, the number of inflammatory cells did not differ significantly (P > 0.05) between subgroups (2, 7, 15, and 30 days, respectively). However, when groups HAM and RAM were compared, a significant P = 0.043 ( 0.05) between subgroups.
Figure 6. Graphic representation of inflammatory cells counting in corneal samples of rabbits submitted to lamellar keratoplasty with human (HAM) and rabbit (RAM) amniotic membrane grafts. The difference between the groups HAM and RAM was significant in days 15 and 30. The counting was higher in group HAM than in group RAM.
animal model for ophthalmological studies. Also, the successful use of AM xenografts in veterinary medicine has been reported in lamellar and penetrating keratoplasty.15–17 There are very few scientific reports on the use of RAM8,12 as a substrate for limbal cell culture with posterior implantation of the cultured tissue in corneal lesions in rabbits. The paucity of studies using allograft AM in rabbits is likely due to the fragility of the AM and its frequent rupture during manipulation. Underutilization of the material has been reported in previous studies8 where the AM sampled from one female rabbit (4–8 fetuses) produced 3–4 culture inserts only (30 mm in diameter). This study has experienced the difficulties involved in rabbit AM harvesting. Extreme care was necessary to avoid rupture or curling up of the membrane. However, the AM harvested
from a single female rabbit was enough to produce 16 fragments of 3 9 3 cm each. The membrane was slightly thicker and more resistant to manipulation close to the umbilical cord.8 Under gross examination, one could observe that the thickness and elasticity differ between rabbit and human AM. Human AM is thicker, more elastic, and easier to fixate to the corneal bed. Due to these structural differences, a fragment larger than the corneal bed was required when rabbit AM was used to resist the tension exerted by the suture. Clinical signs were similar but milder than previously reported following the use of AM xenografts,17 AM allografts,18 and fetal membrane xenografts16 in dogs. Optical microscopy revealed epithelial regeneration and migration over the grafted tissue in the first days postsurgery. Complete epithelialization with paved stratified
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
196 godoy-esteves
ET AL.
epithelium was noted at 15 days, as reported by other researchers.16,18 Epithelial hyperplasia was noted over the operated area in some specimens and may represent a compensation mechanism in areas of stromal thinning19 or reflect the rapid regeneration of rabbit corneal epithelium that is interrupted only when 200% of the initial thickness of the epithelial layer is reached.20 Complete reepithelialization over the AM grafts supports the notion that AM acts as a substrate for restoration of corneal layers,15,21 mimics the properties of the corneal epithelium basal layer,22 promotes epithelial differentiation,23,24 and precludes epithelial apoptosis.25 Amniotic membrane was visible in most specimens under optic microscopy as abundant acidophilic amorphous extracellular matrix. Histological evaluation of the human cornea revealed remnants of amniotic tissue in some patients up to 5–8 months post-transplantation26 and absence in others 2 months post-transplantation,27 suggesting variable speeds of resorption of the implanted tissue. In this study, AM remnants were observed in either group regardless of the presence or absence of inflammatory cells or new vessels. This differed from histological data of studies using AM grafts in human corneas, where the absence of membrane remnants was correlated with the presence of inflammatory infiltrate and blood vessels.26,27 AM resorption was thus variable in both groups, with or without participation of inflammatory cells. When inflammatory cells do not participate in collagen degradation and AM resorption, active keratocytes manifest phagocytic properties against foreign materials (AM) and may contribute to the dissolution of stromal collagen by increasing the synthesis and secretion of collagen degrading enzymes.28 Alcian blue was used in this study to stain the implanted tissue (AM). This particular stain impregnates sulfated acid mucins that are abundant in the AM.29 However, blue staining was observed only in areas of collagen remodelling. The AM appeared in the histological section as pink-colored material. In other studies employing alcian blue29 to highlight the presence of AM remnants, the lack of impregnation led researchers to conclude that AM absorption had already occurred. The findings in this study contradict such assertions given the presence of AM remnants was evident despite the lack of impregnation with the blue stain. When picrosirius red was used, the AM and the stroma around the graft appeared in similar color in both groups, with thick red-orange-stained collagen fibers typical of Type I collagen.30,31 Type I collagen has been identified as a component of the HAM by several researchers,32–35 together with collagen Types III, IV, V, and VII. Image-Pro Plus (Media Cybernetics, Inc) proved to be an invaluable tool for inflammatory cell counting, enabling standardization of the area selected for evaluation and hence a consistent reading of all slides. Areas corresponding to the sutures were excluded due to the intense local
inflammatory reaction observed. The significantly (P < 0.05) higher number of inflammatory cells and new vessels observed in the stroma in group HAM than in group RAM suggests that HAM induces a more intense inflammatory reaction and neovascularization than RAM. Inflammatory cells and keratocytes are responsible for collagen degradation.28 Human AM is much thicker than rabbit AM and therefore has more collagen. The more intense inflammatory reaction observed in animals treated with human AM grafts may reflect the need to degrade and remodel a greater amount of collagen to achieve total incorporation of the membrane into the autochthonous stroma. In summary, HAM transplantation in rabbits may have host reaction potential, but longer follow-up period is required to better understand the inflammatory response induced by xenografts (HAM) in rabbits. Immunohistochemical identification of inflammatory cells is also important given the participation of lymphocytes in the inflammatory reaction may reflect an immune reaction against the graft, resulting from recognition of major (MHC) or minor (mH) histocompatibility antigens expressed in the grafted tissue by CD4 T (helper) cells.36 REFERENCES 1. Kim TC, Tseng SCG. Transplantation of preserved amniotic membrane for surface reconstruction in severely damaged rabbit corneas. Cornea 1995; 14: 473–484. 2. He Y, Alizadeh H, Kinoshita K et al. Experimental transplantation of cultured human limbal and amniotic epithelial cells onto the corneal surface. Cornea 1999; 18: 570–579. 3. Kim JS, Kim JC, Na BK et al. Amniotic membrane patching promotes healing and inhibits proteinase activity on wound healing following acute corneal alkali burn. Experimental Eye Research 2000; 70: 329–337. 4. Avila M, Espa~ na M, Moreno C et al. Reconstruction of ocular surface with heterologous limbal epithelium and amniotic membrane in a rabbit model. Cornea 2001; 20: 414–420. 5. Kim H, Sah W, Kim Y et al. Amniotic membrane, tear film, corneal, and aqueous levels of ofloxacin in rabbit eyes after amniotic membrane transplantation. Cornea 2001; 20: 628–634. 6. Wang MX, Gray TB, Park WC et al. Reduction in corneal haze and apoptosis by amniotic membrane matrix in excimer laser photoablation in rabbits. Journal of Cataract and Refrative Surgery 2001; 27: 310–319. 7. Woo HM, Kim MS, Kweon OK et al. Effects of amniotic membrane on epithelial wound healing and stromal remodelling after excimer laser keratectomy in rabbit cornea. The British Journal of Ophthalmology 2001; 85: 345–349. 8. Ti SE, Anderson D, Touhami A et al. Factors affecting outcome following transplantation of ex vivo expanded limbal epithelium on amniotic membrane for total limbal deficiency in rabbits. Investigative Ophthalmology & Visual Science 2002; 43: 2584–2592. 9. Marinho D, Hofling-Lima AL, Kwitko S et al. Does amniotic membrane transplantation improve the outcome of autologous limbal transplantation? Cornea 2003; 22: 338–342. 10. Nakamura T, Endo K, Cooper LJ et al. The successful culture and autologous transplantation of rabbit oral mucosal epithelial
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
keratoplasty in rabbits using amniotic membrane 197
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
cells on amniotic membrane. Investigative Ophthalmology & Visual Science 2003; 44: 106–116. Nakamura T, Yoshitani M, Rigby H et al. Sterilized, freeze-dried amniotic membrane: a useful substrate for ocular surface reconstruction. Investigative Ophthalmology & Visual Science 2004; 45: 93–99. Ti SE, Greterich M, Espana M et al. Correlation of long term phenotypic and clinical outcomes following limbal epithelial transplantation cultivated on amniotic membrane in rabbits. British Journal of Ophthalmology 2004; 88: 422–427. Kubo M, Sonoda Y, Muramatsu R et al. Immunogenicity of human amniotic membrane in experimental xenotransplantation. Investigative Ophthalmology & Visual Science 2001; 42: 1539–1546. Gomes JAP, Komagone CM, Santos N et al. Membrana amni otica nas cirurgias reconstrutivas da superfıcie ocular nas ceratoconjuntivites cicatriciais. Arquivos Brasileiros de Oftalmologia 1999; 62: 563–576. Barros PSM, Garcia JA, Laus JL et al. The use of xenologous amniotic membrane to repair canine corneal perforation created by penetrating keratectomy. Veterinary Ophthalmology 1998; 1: 119–123. Godoy CAL, Guerra JL, Barros PSM. Ceratoplastia lamelar em c~aes utilizando membrana fetal equina como enxerto. Estudo Experimental. Arquivos Brasileiros de Oftalmologia 2002; 65: 501–602. Barros PSM, Safatle AVM, Godoy CA et al. Amniotic membrane transplantation for the reconstruction of the ocular surface in three cases. Veterinary Ophthalmology 2005; 8: 189–192. Souza MSB. Estudo Comparativo Entre A Ceratoplastia Lamelar Em C~aes Utilizando Membrana Amniotica E Alantoamniotica Alogenas Preservadas Em Glicerina: Avaliacß~ao Clınica E Morfologica. 2003. 161 F. Dissertacß~ao (Doutorado Em Medicina Veterinaria). Faculdade De Medicina Veterinaria E Zootecnia, Universidade De S~ao Paulo, S~ao Paulo, 2003. Gris O, Wolley-Dod C, G€ uell JL et al. Histologic findings after amniotic membrane graft in the human cornea. Ophthalmology 2002; 109: 508–512. Prince JH. Cornea, trabecular region and sclera. In: The Rabbit in Eye Research. (ed. Prince JH) Charles C. Thomas, Springfield, IL, 1964; 652. Shimazaki J, Yang H, Tsubota K. Amniotic membrane transplantation for ocular surface reconstruction in patients with chemical and thermal burns. Ophthalmology 1997; 104: 2068–2076. Shimazaki J, Shinozaki N, Tsubota K. Transplantation of amniotic membrane and limbal autograft for patients with
23.
24.
25.
26.
27.
28.
29. 30.
31.
32.
33.
34.
35.
36.
recurrent pterygium associated with symblepharon. The British Journal of Ophthalmology 1998; 82: 235–240. Kurpakus MA, Stock EL, Jones JCR. The role of the basement membrane in differential expression of keratin proteins in epithelial cells. Developmental Biology 1992; 150: 243–255. Grueterich M, Espana EM, Touhami A et al. Phenotypic study of a case with successful transplantation of ex vivo expanded human limbal epithelium for unilateral total limbal stem cell deficiency. Ophthalmology 2002; 109: 1547–1552. Chen H, Pires RTF, Tseng SCG. Amniotic membrane transplantation for severe neurotrophic ulcers. The British Journal of Ophthalmology 2000; 94: 826–833. Stoiber J, Muss WH, Pohla-Gubo G et al. Histopathology of human corneas after amniotic membrane and limbal stem cell transplantation for severe chemical burn. Cornea 2002; 21: 482–489. Tosi GM, Traversi C, Schuerfeld K et al. Amniotic membrane graft: histopathological findings in five cases. Journal of Cellular Physiology 2005; 202: 852–857. Nishita T. Cornea. In: Cornea (Chapter 1). (eds Krachmer JH, Mannis MJ, Holland EJ, Palay DA) Mosby Yearbook, St. Louis, MO, 1998; 3–24. CD-ROM. Stevens A, Lowe J. Histologia Humana, 2nd edn. Editora Manole Ltda, S~ao Paulo, 2001; 408. Junqueira LCU, Bignolas G, Brentani RR. Picrosirius staining plus polarization microscopy: a specific method for collagen direction in tissue sections. Histochemical Journal 1979; 11: 447– 455. Massironi SMG, Giac oia MR, Maiorka PC et al. Skin morphology of the mutant hairless USP mouse. Brazilian Journal of Medical and Biological Research 2005; 38: 33–39. Aplin JD, Campbell S, Allen TD. The extracellular matrix of human amniotic epithelium: ultrastructure, composition and deposition. Journal of Cell Science 1985; 79: 119–136. Keene DR, Sakai LY, Lunstrum GP. Type VII collagen forms an extended network of anchoring fibrils. Journal of Cell Biology 1987; 104: 611–621. Yurchenco PD, Ruben GC. Basement membrane structure in situ: evidence for lateral associations in the Type IV collagen network. Journal of Cell Biology 1987; 105: 2559–2568. Modesti A, Scarpa S, Dorazi G. Localization of Type IV and V collagens in the stroma of human amnion. Progress in Clinical and Biological Research 1989; 296: 459–463. Forrester JV, Dick AD, McMenamin PG et al. The Eye: Basic Sciences in Practice, 2nd edn. W.B. Saunders, Philadelphia, PA, 2002; 447.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
DOI:10.1111/vop.12109
Lamellar keratoplasty in rabbits using human and rabbit amniotic membrane grafts: a comparative study Cintia A. L. Godoy-Esteves,* Jose Alvaro P. Gomes,† Karina Yazbek,‡ Jose L. Guerra§ and Paulo S. M. Barros* *Laboratory of Investigation on Comparative Ophthalmology, School of Veterinary Medicine, University of S~ao Paulo, S~ao Paulo, SP, CEP 02478-000, Brazil; †Sector of Corneal and External Diseases, School of Medicine, Federal University of S~ao Paulo, S~ao Paulo, Brazil; ‡Agener Sa ude Animal, S~ao Paulo, Brazil; and §Department of Pathology, School of Veterinary Medicine, University of S~ao Paulo, S~ao Paulo, Brazil
Address communications to: C. A. L. Godoy-Esteves Tel.: 55 11 3981-3229/ 55 11 99949-5949 e-mail: cintiagodoy@terra. com.br
Abstract Objective To compare the behavior of human and rabbit amniotic membrane (AM) grafts in surgically induced corneal thinning in rabbits. Animals studied Thirty two NZWR were randomly assigned to two groups of 16 animals each according to AM type (Human AM: group HAM and Rabbit AM: group RAM). Procedure All animals were submitted to right keratectomy at a depth of 0.1 mm using a 5 mm trephine. Animals from HAM group had a button of 5 mm of human AM sutured into the corneal bed with a continuous pattern and 10.0 nylon monofilament suture, while animals from the RAM group had a button of 6 mm diameter of rabbit AM. Four animals in each group were euthanized 2, 7, 15, and 30 days postoperatively. Their corneas were harvested, fixed in 2% glutaraldehyde solution, and stained with haematoxylin–eosin, picrosirius red, and alcian blue for evaluation under light optical microscopy. Microscope images were digitalized and inflammatory cells and stromal blood vessels were counted. Results There were no clinically significant differences between groups, and complete corneal epithelialization was observed in all animals in 30 days. Light optical microscopy revealed AM incorporation and resorption in both groups. However, the number of inflammatory cells and blood vessels was significantly higher in group HAM than in group RAM (P < 0.05, Mann–Whitney test). Clinical responses to human or rabbit AM were similar; however, human AM induced greater inflammatory reaction and stromal neovascularization in the rabbit cornea than in rabbit AM. Conclusion These differences may reflect a potential reaction to the xenograft. More studies are needed to further characterize these findings. Key Words: amniotic membrane, cornea, graft, human, rabbit
INTRODUCTION
Human amniotic membrane (HAM) has gained popularity as a successful adjunct in ocular surface reconstructive surgery over the last decade. Most of the effects of amniotic membrane (AM) transplantation on the ocular surface have been determined through research in rabbits.1–12 Published studies present conflict regarding the HAM potential to induce immune reactions.13 Evidence shows that HAM may behave as an immunologically privileged tissue despite the presence of viable cells expressing HLA-G (major histocompatibility © 2013 American College of Veterinary Ophthalmologists
antigen) in this cryopreserved tissue. However, the HAM may have the potential to induce reactions to the transplanted tissue that may interfere with the interpretation of results. In addition, the effects of rabbit AM (RAM) transplantation and the response of rabbits to xenografts and allografts have received little attention. RAM has been used in rare occasions.8,12 The lack of comparative data between HAM and RAM grafts has motivated this investigation. The aim of this study was to determine whether the use of a xenograft transplantation using HAM in rabbit corneas has the potential to cause adverse host tissue
192 godoy-esteves
ET AL.
reactions when compared to the use of an allograft using RAM. The authors compared the following parameters: clinical and histological (haematoxylin–eosin stain) characteristics, determination of proteoglycan distribution and characterization (alcian blue stain), collagen type (picrosirius red stain) and inflammatory cells, and newly formed stromal vessels. MATERIALS AND METHODS
Harvesting of the amniotic membrane The HAMs employed in this study were donated by the Laboratorio de Doencßas Externas Oculares (External Ocular Disease Laboratory), Sector of Corneal and External Diseases, Department of Ophthalmology – UNIFESP. Human placentas were harvested following cesarean sections and washed with 0.9% saline solution under aseptic conditions. The AM was bluntly dissected from the remaining fetal membranes and flattened onto sterile nitrocellulose filter paper (Micropore, Bedfort, MA, USA) with the epithelial surface up. The AM with the filter was washed with a phosphate-buffered saline solution containing penicillin (1000 U/mL), streptomycin (20 lg/mL) and amphotericin B (2.5 lg/mL), cut into approximately 5 9 5 cm fragments and placed in a sterile container with 1:1 glycerol and corneal preservation medium (Baxter Healthcare corporation, Stone Mountain, GA, USA, and Ophthalmos, S~ao Paulo, SP, Brazil, respectively) and frozen at 80 °C.14 Rabbit placentas were harvested following ovariosalpingohysterectomy performed on 28th day of pregnancy. Following separation of the AM from the remaining fetal membranes, the same protocol was employed for AM preparation, except for the size of the fragments that corresponded to 3 9 3 cm. The low thickness and elasticity of the rabbit placenta made the manipulation difficult. Extreme care was required to avoid rupture and curling up of the membrane that would prevent its use. Animals Thirty-two young adult New Zealand rabbits weighing between 1.5 and 2.0 kg obtained from privately owned rearing facilities were used, conforming to Association for Research in Vision and Ophthalmology (ARVO) guidelines for research involving animal models. Animals were randomly assigned to two groups of 16 individuals for evaluation of each membrane type (group HAM and group RAM, respectively). Animals in each group were again randomly assigned to four subgroups with different follow-up periods (2, 7, 15, and 30 days, respectively). Anesthetic procedure Animals were fasted for 2 h previous to sedation with a combination of acepromazin–meperidin (0.4 mg/kg and 10 mg/kg, respectively) given intramuscularly and instillation of anesthetic eye drops containing proparacaine
hydrochloride. Following a 15-min lag phase, anesthesia was induced with a combination of ketamine and midazolam (10 mg/kg and 0.3 mg/kg, respectively) given intravenously (IV) and maintained with IV ketamine (5–10 mg/ kg) administered as necessary. Venous infusion with lactated Ringer’s solution was maintained throughout the surgical procedure, and flunixin meglumine (1 mg/kg) was given IV. The marginal ear vein was used for all IV medications.
Surgical procedure Lamellar keratectomy was performed using an operating microscope (DF Vasconcelos, Brasil) with 109 magnification. A 5 mm Castroviejo trephine was used to penetrate the cornea to a depth of 0.1 mm in the dorsotemporal area, 1 mm from the limbus, creating a recipient bed for implantation of HAM or RAM grafts. HAM and RAM grafts were created using a trephine (5 and 6 mm, respectively). Grafts (the epithelium side up) were sutured into the recipient bed with a continuous pattern and 10-0 nylon monofilament suture on a spatulated needle. Animals were kept in individual cages following surgery and treated with tobramycin eye drops three times daily for the first 2 weeks postsurgery. As signs of pain were not observed, no analgesics or Elizabethan collar was used. Animals were euthanized 2, 7, 15, or 30 days postsurgery. Following deep anesthesia with sodium thiopental, cardiorespiratory arrest was induced with IV 19.1% potassium chloride. Clinical assessment Animals were evaluated using a portable slit-lamp biomicroscope (SL-14 Portable Slit Lamp, Kowa, Tokyo, Japan) at 24-h intervals during the first week postsurgery and every 72 h thereafter until euthanasia according to the respective subgroup follow-up period. Clinical assessment was based on the presence of photophobia, blepharospasm, ocular discharge (occurrence and characteristics), conjunctival hyperemia, corneal opacity, new vessel formation and graft retention and viability. The changes observed were subjectively graded as (0) absent, +(1) mild, ++(2) moderate, and +++(3) severe. Microscopic evaluation A corneal button was obtained by full thickness trephination at the corneoscleral transition, fixed in 2% buffered glutaraldehyde solution, processed, and stained with HE, picrosirius red, and alcian blue for optical microscopy. Slides from the recipient area were used for evaluation of the morphological patterns related to the grafting procedure. Apart from morphological patterns related to grafting, HE-stained slides were also used for counting of inflammatory cells and vessels in the corneal stroma using a Nikon E-800 microscope (Nikon Metrology, Inc.,
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
keratoplasty in rabbits using amniotic membrane 193
Brighton, MI, USA) coupled with Image-Pro Plus software (Media Cybernetics, Inc., Rockville, MD, USA). The recipient area was divided randomly into smaller subareas of 35401.2 lm2 each, and five of these smaller subareas (177 006 lm in total) distributed along the recipient area were examined under 209 magnification. Values corresponding to the number of inflammatory cells and blood vessels counted in different animals were submitted to the Mann–Whitney nonparametric test. The level of significance was set at 5% (P < 0.05). RESULTS
Clinical assessment Human amniotic membrane and RAM grafts induced similar clinical responses: Blepharospasm: Three animals presented with mild signs in both groups; Conjunctival discharge: In group HAM, three animals presented with mild signs, eight with moderate signs, and two with severe signs. In group RAM, two animals presented with mild signs, one with moderate signs, and two with severe signs distributed through all time periods; and Conjunctival hyperemia: In group HAM, five animals presented with mild signs, five with moderate and two with severe signs. All were distributed through all periods. In group RAM, only five animals had hyperemia (four were mild in 2 days, and one was moderate in 30 days). Mild-to-moderate corneal opacity was noted in most animals in group HAM (32 with mild in 2-day period, and in the final period – 30 days, one with mild, two with moderate and one with severe). Mild corneal opacity was observed in all animals in group RAM throughout the follow-up period. Corneal vascularization was not observed clinically in the RAM group during the first 7 days postsurgery, while in HAM group, it was clinically obvious at day 2 in two animals (one with mild and one with moderate). Corneal vascularization progressed over time and at day 15 was observed in the HAM group as mild in one animal, as moderate in five, and as severe in one. At day 30, two animals had developed moderate vascularization and two had developed severe vascularization. Vascularization was also observed at day 15 in the RAM group as mild in five animals and as moderate in three. At day 30, one had developed mild, two moderate, and one severe signs. Graft retention and viability were observed throughout the follow-up period in both groups. Microscopy Two days postimplantation, the implanted tissue was visible under light microscopy as abundant acidophilic amorphous extracellular matrix in both groups. Inflammatory cell migration and invasion, consisting of polymorphonuclear cells, were observed immediately deep to the implanted tissue. A single or double layer of flat epithelium covered the implanted tissue.
At day 7, in both groups, the implanted tissue was integrated into the autochthonous tissue and remnants consisting of amorphous acidophilic material were visible. The corneal surface was completely covered with stratified flat epithelium. Many cells infiltrated the recipient area, and under greater magnification, cells with fibroblast-like morphology could be observed in the implanted tissue, indicating proliferation of these cells over the substrate represented by the implanted tissue. At day 15, areas of discontinuous amorphous acidophilic material corresponding to the integrated HAM were visible in group HAM (Fig. 1). The corneal surface was completely covered with stratified paved epithelium. A large number of inflammatory cells, consisting of polymorphonuclear cells, and cells with fibroblast-like morphology were present in the tissue underlying the graft. A negative image corresponding to the suture material surrounded by intense infiltrate of polymorphonuclear cells was noted at the periphery of the histological section. At day 15, in group RAM (Fig. 2), the AM was not as clearly defined as in group HAM. A thin band of amorphous acellular material possibly corresponding to remnants of the RAM graft was observed deep to the epithelial layer. The epithelium was classified as paved stratified, and numerous inflammatory cells, mostly polymorphonuclear cells, and cells with fibroblast-like morphology were noted in the stroma underlying the recipient area. Epithelial hyperplasia was noted over the operated area in some specimens. At day 30, the AM was integrated and well defined under paved stratified epithelium in both groups. Intense inflammatory cell infiltrate, characterized by polymorphonuclear cells in most animals, was observed in the stroma underlying the recipient area in group HAM (Fig. 3). In group RAM (Fig. 4), remnants of the RAM, cells with
Figure 1. Photomicrograph of the histological section of animal No. 12 corresponding to the area treated with human amniotic membrane graft 15 days post-implantation. Perfect integration of the implanted tissue (arrow) into the autochthonous tissue and epithelial regeneration are visible. The suture material surrounded by inflammatory infiltrate can be noted. HE staining; 49 magnification.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
194 godoy-esteves
ET AL.
Figure 2. Photomicrograph of the histological section of animal No. 7 corresponding to the area treated with rabbit amniotic membrane graft 15 days post-implantation. Epithelial regeneration, a negative image of the suture material surrounded by inflammatory infiltrate and the presence of inflammatory cells in the stroma can be noted. A button of rabbit amniotic membrane (arrow) is visible below the paved stratified epithelium. HE staining; 209 magnification.
Figure 3. Photomicrograph of the histological section corresponding to the area treated with human amniotic membrane graft 30 days post-implantation. The membrane is visible below the paved stratified epithelium. Intense inflammatory infiltrate can be noted in the surrounding stroma. HE staining; 109 magnification.
fibroblast-like morphology and moderate inflammatory infiltrate, also characterized by polymorphonuclear cells in most animals, were noted in the stroma above and below the implanted tissue. In the sections stained with picrosirius red and evaluated under polarized light, AMs of both human and rabbit were visible in orange-red color, suggesting the presence of Type I collagen. In the histological sections stained with alcian blue, the AM was visible as a pinkish material in both groups. Areas where collagen remodelled were likely stained blue, suggesting the presence of acid glycosaminoglycans. Blood vessel counting in the stroma adjacent to the grafted area was higher in group HAM than in group RAM (Fig. 5). The number of blood vessels was significantly (P < 0.05) higher in group HAM than in group
Figure 4. Photomicrograph of the histological section of animal No. 28 corresponding to the area treated with rabbit amniotic membrane graft 30 days post-implantation. Remnants of amorphous material corresponding to the grafted membrane are visible (arrow). Note the presence of cells with fibroblast-like morphology and inflammatory cells in the stroma. HE staining; 109 magnification.
RAM at 15 and 30 days. In group HAM, the number of corneal blood vessels was significantly (P < 0.05) higher in animals in the 30-day follow-up period subgroup than in the remaining subgroups. In group RAM, the number of corneal blood vessels did not differ significantly (P > 0.05) between subgroups. Under microscopic examination, blood vessel counting was significantly higher in group HAM. Throughout the follow-up, in most animals, the inflammatory response was characterized by polymorphonuclear cells. Four animals in each group had a mixed inflammatory response involving both mononuclear and polymorphonuclear cells. However, giant cells were not observed in any of the samples. In groups HAM and RAM, the number of inflammatory cells did not differ significantly (P > 0.05) between subgroups (2, 7, 15, and 30 days, respectively). However, when groups HAM and RAM were compared, a significant P = 0.043 ( 0.05) between subgroups.
Figure 6. Graphic representation of inflammatory cells counting in corneal samples of rabbits submitted to lamellar keratoplasty with human (HAM) and rabbit (RAM) amniotic membrane grafts. The difference between the groups HAM and RAM was significant in days 15 and 30. The counting was higher in group HAM than in group RAM.
animal model for ophthalmological studies. Also, the successful use of AM xenografts in veterinary medicine has been reported in lamellar and penetrating keratoplasty.15–17 There are very few scientific reports on the use of RAM8,12 as a substrate for limbal cell culture with posterior implantation of the cultured tissue in corneal lesions in rabbits. The paucity of studies using allograft AM in rabbits is likely due to the fragility of the AM and its frequent rupture during manipulation. Underutilization of the material has been reported in previous studies8 where the AM sampled from one female rabbit (4–8 fetuses) produced 3–4 culture inserts only (30 mm in diameter). This study has experienced the difficulties involved in rabbit AM harvesting. Extreme care was necessary to avoid rupture or curling up of the membrane. However, the AM harvested
from a single female rabbit was enough to produce 16 fragments of 3 9 3 cm each. The membrane was slightly thicker and more resistant to manipulation close to the umbilical cord.8 Under gross examination, one could observe that the thickness and elasticity differ between rabbit and human AM. Human AM is thicker, more elastic, and easier to fixate to the corneal bed. Due to these structural differences, a fragment larger than the corneal bed was required when rabbit AM was used to resist the tension exerted by the suture. Clinical signs were similar but milder than previously reported following the use of AM xenografts,17 AM allografts,18 and fetal membrane xenografts16 in dogs. Optical microscopy revealed epithelial regeneration and migration over the grafted tissue in the first days postsurgery. Complete epithelialization with paved stratified
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
196 godoy-esteves
ET AL.
epithelium was noted at 15 days, as reported by other researchers.16,18 Epithelial hyperplasia was noted over the operated area in some specimens and may represent a compensation mechanism in areas of stromal thinning19 or reflect the rapid regeneration of rabbit corneal epithelium that is interrupted only when 200% of the initial thickness of the epithelial layer is reached.20 Complete reepithelialization over the AM grafts supports the notion that AM acts as a substrate for restoration of corneal layers,15,21 mimics the properties of the corneal epithelium basal layer,22 promotes epithelial differentiation,23,24 and precludes epithelial apoptosis.25 Amniotic membrane was visible in most specimens under optic microscopy as abundant acidophilic amorphous extracellular matrix. Histological evaluation of the human cornea revealed remnants of amniotic tissue in some patients up to 5–8 months post-transplantation26 and absence in others 2 months post-transplantation,27 suggesting variable speeds of resorption of the implanted tissue. In this study, AM remnants were observed in either group regardless of the presence or absence of inflammatory cells or new vessels. This differed from histological data of studies using AM grafts in human corneas, where the absence of membrane remnants was correlated with the presence of inflammatory infiltrate and blood vessels.26,27 AM resorption was thus variable in both groups, with or without participation of inflammatory cells. When inflammatory cells do not participate in collagen degradation and AM resorption, active keratocytes manifest phagocytic properties against foreign materials (AM) and may contribute to the dissolution of stromal collagen by increasing the synthesis and secretion of collagen degrading enzymes.28 Alcian blue was used in this study to stain the implanted tissue (AM). This particular stain impregnates sulfated acid mucins that are abundant in the AM.29 However, blue staining was observed only in areas of collagen remodelling. The AM appeared in the histological section as pink-colored material. In other studies employing alcian blue29 to highlight the presence of AM remnants, the lack of impregnation led researchers to conclude that AM absorption had already occurred. The findings in this study contradict such assertions given the presence of AM remnants was evident despite the lack of impregnation with the blue stain. When picrosirius red was used, the AM and the stroma around the graft appeared in similar color in both groups, with thick red-orange-stained collagen fibers typical of Type I collagen.30,31 Type I collagen has been identified as a component of the HAM by several researchers,32–35 together with collagen Types III, IV, V, and VII. Image-Pro Plus (Media Cybernetics, Inc) proved to be an invaluable tool for inflammatory cell counting, enabling standardization of the area selected for evaluation and hence a consistent reading of all slides. Areas corresponding to the sutures were excluded due to the intense local
inflammatory reaction observed. The significantly (P < 0.05) higher number of inflammatory cells and new vessels observed in the stroma in group HAM than in group RAM suggests that HAM induces a more intense inflammatory reaction and neovascularization than RAM. Inflammatory cells and keratocytes are responsible for collagen degradation.28 Human AM is much thicker than rabbit AM and therefore has more collagen. The more intense inflammatory reaction observed in animals treated with human AM grafts may reflect the need to degrade and remodel a greater amount of collagen to achieve total incorporation of the membrane into the autochthonous stroma. In summary, HAM transplantation in rabbits may have host reaction potential, but longer follow-up period is required to better understand the inflammatory response induced by xenografts (HAM) in rabbits. Immunohistochemical identification of inflammatory cells is also important given the participation of lymphocytes in the inflammatory reaction may reflect an immune reaction against the graft, resulting from recognition of major (MHC) or minor (mH) histocompatibility antigens expressed in the grafted tissue by CD4 T (helper) cells.36 REFERENCES 1. Kim TC, Tseng SCG. Transplantation of preserved amniotic membrane for surface reconstruction in severely damaged rabbit corneas. Cornea 1995; 14: 473–484. 2. He Y, Alizadeh H, Kinoshita K et al. Experimental transplantation of cultured human limbal and amniotic epithelial cells onto the corneal surface. Cornea 1999; 18: 570–579. 3. Kim JS, Kim JC, Na BK et al. Amniotic membrane patching promotes healing and inhibits proteinase activity on wound healing following acute corneal alkali burn. Experimental Eye Research 2000; 70: 329–337. 4. Avila M, Espa~ na M, Moreno C et al. Reconstruction of ocular surface with heterologous limbal epithelium and amniotic membrane in a rabbit model. Cornea 2001; 20: 414–420. 5. Kim H, Sah W, Kim Y et al. Amniotic membrane, tear film, corneal, and aqueous levels of ofloxacin in rabbit eyes after amniotic membrane transplantation. Cornea 2001; 20: 628–634. 6. Wang MX, Gray TB, Park WC et al. Reduction in corneal haze and apoptosis by amniotic membrane matrix in excimer laser photoablation in rabbits. Journal of Cataract and Refrative Surgery 2001; 27: 310–319. 7. Woo HM, Kim MS, Kweon OK et al. Effects of amniotic membrane on epithelial wound healing and stromal remodelling after excimer laser keratectomy in rabbit cornea. The British Journal of Ophthalmology 2001; 85: 345–349. 8. Ti SE, Anderson D, Touhami A et al. Factors affecting outcome following transplantation of ex vivo expanded limbal epithelium on amniotic membrane for total limbal deficiency in rabbits. Investigative Ophthalmology & Visual Science 2002; 43: 2584–2592. 9. Marinho D, Hofling-Lima AL, Kwitko S et al. Does amniotic membrane transplantation improve the outcome of autologous limbal transplantation? Cornea 2003; 22: 338–342. 10. Nakamura T, Endo K, Cooper LJ et al. The successful culture and autologous transplantation of rabbit oral mucosal epithelial
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
keratoplasty in rabbits using amniotic membrane 197
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
cells on amniotic membrane. Investigative Ophthalmology & Visual Science 2003; 44: 106–116. Nakamura T, Yoshitani M, Rigby H et al. Sterilized, freeze-dried amniotic membrane: a useful substrate for ocular surface reconstruction. Investigative Ophthalmology & Visual Science 2004; 45: 93–99. Ti SE, Greterich M, Espana M et al. Correlation of long term phenotypic and clinical outcomes following limbal epithelial transplantation cultivated on amniotic membrane in rabbits. British Journal of Ophthalmology 2004; 88: 422–427. Kubo M, Sonoda Y, Muramatsu R et al. Immunogenicity of human amniotic membrane in experimental xenotransplantation. Investigative Ophthalmology & Visual Science 2001; 42: 1539–1546. Gomes JAP, Komagone CM, Santos N et al. Membrana amni otica nas cirurgias reconstrutivas da superfıcie ocular nas ceratoconjuntivites cicatriciais. Arquivos Brasileiros de Oftalmologia 1999; 62: 563–576. Barros PSM, Garcia JA, Laus JL et al. The use of xenologous amniotic membrane to repair canine corneal perforation created by penetrating keratectomy. Veterinary Ophthalmology 1998; 1: 119–123. Godoy CAL, Guerra JL, Barros PSM. Ceratoplastia lamelar em c~aes utilizando membrana fetal equina como enxerto. Estudo Experimental. Arquivos Brasileiros de Oftalmologia 2002; 65: 501–602. Barros PSM, Safatle AVM, Godoy CA et al. Amniotic membrane transplantation for the reconstruction of the ocular surface in three cases. Veterinary Ophthalmology 2005; 8: 189–192. Souza MSB. Estudo Comparativo Entre A Ceratoplastia Lamelar Em C~aes Utilizando Membrana Amniotica E Alantoamniotica Alogenas Preservadas Em Glicerina: Avaliacß~ao Clınica E Morfologica. 2003. 161 F. Dissertacß~ao (Doutorado Em Medicina Veterinaria). Faculdade De Medicina Veterinaria E Zootecnia, Universidade De S~ao Paulo, S~ao Paulo, 2003. Gris O, Wolley-Dod C, G€ uell JL et al. Histologic findings after amniotic membrane graft in the human cornea. Ophthalmology 2002; 109: 508–512. Prince JH. Cornea, trabecular region and sclera. In: The Rabbit in Eye Research. (ed. Prince JH) Charles C. Thomas, Springfield, IL, 1964; 652. Shimazaki J, Yang H, Tsubota K. Amniotic membrane transplantation for ocular surface reconstruction in patients with chemical and thermal burns. Ophthalmology 1997; 104: 2068–2076. Shimazaki J, Shinozaki N, Tsubota K. Transplantation of amniotic membrane and limbal autograft for patients with
23.
24.
25.
26.
27.
28.
29. 30.
31.
32.
33.
34.
35.
36.
recurrent pterygium associated with symblepharon. The British Journal of Ophthalmology 1998; 82: 235–240. Kurpakus MA, Stock EL, Jones JCR. The role of the basement membrane in differential expression of keratin proteins in epithelial cells. Developmental Biology 1992; 150: 243–255. Grueterich M, Espana EM, Touhami A et al. Phenotypic study of a case with successful transplantation of ex vivo expanded human limbal epithelium for unilateral total limbal stem cell deficiency. Ophthalmology 2002; 109: 1547–1552. Chen H, Pires RTF, Tseng SCG. Amniotic membrane transplantation for severe neurotrophic ulcers. The British Journal of Ophthalmology 2000; 94: 826–833. Stoiber J, Muss WH, Pohla-Gubo G et al. Histopathology of human corneas after amniotic membrane and limbal stem cell transplantation for severe chemical burn. Cornea 2002; 21: 482–489. Tosi GM, Traversi C, Schuerfeld K et al. Amniotic membrane graft: histopathological findings in five cases. Journal of Cellular Physiology 2005; 202: 852–857. Nishita T. Cornea. In: Cornea (Chapter 1). (eds Krachmer JH, Mannis MJ, Holland EJ, Palay DA) Mosby Yearbook, St. Louis, MO, 1998; 3–24. CD-ROM. Stevens A, Lowe J. Histologia Humana, 2nd edn. Editora Manole Ltda, S~ao Paulo, 2001; 408. Junqueira LCU, Bignolas G, Brentani RR. Picrosirius staining plus polarization microscopy: a specific method for collagen direction in tissue sections. Histochemical Journal 1979; 11: 447– 455. Massironi SMG, Giac oia MR, Maiorka PC et al. Skin morphology of the mutant hairless USP mouse. Brazilian Journal of Medical and Biological Research 2005; 38: 33–39. Aplin JD, Campbell S, Allen TD. The extracellular matrix of human amniotic epithelium: ultrastructure, composition and deposition. Journal of Cell Science 1985; 79: 119–136. Keene DR, Sakai LY, Lunstrum GP. Type VII collagen forms an extended network of anchoring fibrils. Journal of Cell Biology 1987; 104: 611–621. Yurchenco PD, Ruben GC. Basement membrane structure in situ: evidence for lateral associations in the Type IV collagen network. Journal of Cell Biology 1987; 105: 2559–2568. Modesti A, Scarpa S, Dorazi G. Localization of Type IV and V collagens in the stroma of human amnion. Progress in Clinical and Biological Research 1989; 296: 459–463. Forrester JV, Dick AD, McMenamin PG et al. The Eye: Basic Sciences in Practice, 2nd edn. W.B. Saunders, Philadelphia, PA, 2002; 447.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 18, 191–197
