Introduction

Many biological membranes have been used in ophthalmic reconstructive surgeries. Autogenic, allogeneic and xenologous implants have been evaluated in the reconstruction of ocular surfaces after traumas and surgical excisions showing good results. Among them, amniotic membrane has provided excellent results (Barros et al. 2005, Cremonini et al. 2007). Amniotic membranes can be used in two ways in ophthalmic surgery. As a graft, the membrane is used as a scaffold for cell migration, becoming epithelized and incorporated in the host tissue. For this, it should be place on the ocular surface with its epithelial surface side up. As a patch it is used to restrain inflammatory reaction while epithelization occurs under it, being applied with its epithelial surface against the ocular surface, in contact to the wound bed (Dua et al. 2004).

Clinical, histological and histomorphometric analysis were carried out in order to evaluate three aspects: 1) the effect of canine amniotic membranes, previously preserved in glycerin, on experimentally-made superficial corneal ulcers in rabbits; 2) compare the time needed for corneal epithelization between membrane treated defects and non treated ones; 3) time needed for complete corneal transparency after the repair of the ulcers in both groups.

Materials and Methods

The work using animals here described were firstly approved by the Ethic Committee on Animal Experimentation of the Viçosa Federal University Veterinary Department, protocol number 85/2006, following the Association for Research in Vision and Ophthalmology (ARVO) guidelines. Amniotic membranes prepare was carried out according to methods described by Kim & Tseng (1995). Twenty-eight healthy New Zealand white rabbits were needed, all adults weighting 3 to 4 kg randomly separated in two groups of 14 animals. The treated group (TG) received the xenologous amniotic membrane as a patch, previously preserved in glycerin 99%. The control group (CT) received no treatment. The animals were sedated using acepromazine intravenously (0.1 mg/kg) and, after 15 minutes, they were anesthetized using an association of tiletamine and zolazepam intramuscularly (30 mg/kg). The eye under surgery received anesthetic drops of lidocaine 4% during the whole procedure. Superficial keratectomy was made on the left eye of each animal using a Castroviejo trephine; this defect was made on a clock equivalent position at 2 hours, 3mm away from the corneal limbus. A 5 mm diameter fragment was excised, 0.15mm thick. A fragment of the amniotic membrane was hydrated for 10 minutes and then placed on the defect with the epithelial surface facing the cornea of the animals in the treated group. Its fixation to the cornea was made using simple separated 9-0 nylon suture. Postsurgery treatment included atropine and antibacterial drops (neomycin, polymyxin B and bacitracin) during 7 days (1 drop, 4 times a day). All animals were kept in individual cages using Elizabethan collars until the complete corneal epithelization. Clinical exams included itching observation and direct ophthalmoscopy to evaluate blepharospasm, ocular discharge, conjunctival vascular congestion and corneal neovascularization. These were classified in present or absent beginning the exams 24 hours after surgery with 48 hours-intervals in the first 7 days and then in 4 days-interval until the end of the observation period. Fluorescein test allowed to establish until what day the amniotic membrane remained in place, being also useful to clinically determine the amount of epithelization in the CG and after removing fragments of membrane that were still attached to the sutures in TG. This test began 24 hours after surgery in 48 hours intervals until a negative result was achieved. Corneal and implant opacity were classified as absent when at least one of them was completely transparent; discrete when colored white but allowing anterior chamber visualization; intense when colored white but not allowing anterior chamber visualization. The animals were euthanized on days 1, 2, 7, 15, 30, 60 and 180 after surgeries. The eyes where enucleated and then fixed in Bouin solution for histological and histomorphometric analysis after staining with hematoxylin-eosin (HE) and Gomori's trichrome (GT). Qualitative variables were submitted to Wilcoxon nonparametric test, considering a significance level of p < 0.05. Dichotomic qualitative variables were compared in contingence tables and analyzed using chi-squared test, also p < 0.05 (Sampaio 2002).

Results

Blepharospasm, mucous ocular discharge, itching and conjunctival vascular congestion were significantly greater in TG than CG. Membranes were found degrading seven days after surgery. Histology showed epithelization beginning on the 2nd day in the TG while in the CG it started only on day 7, however, this process was complete between days 7 and 15 after surgery in the CG but not in the TG. Clinical exams using fluorescein tests for epithelization corroborated these findings, seen by positive results extending further in TG than CG. The number of polymorphonucleated cells invading corneal stroma in CG was significantly greater than in the TG; in the latter, these cells were mostly seen on the membranes. Corneal neovascularization began on the 4th day after surgery and remained for variable times in TG, a process not seen in CG. Fibroblasts number was greater on the TG than on the CG animals. Corneal opacity was seen in all animals during this experiment, being significantly more intense on the TG than CG in the first 18 days. From this day forward, no difference was found between groups. Corneal edema was present only in the first day after surgery in all animals. Histomorphometric evaluation showed no significant difference in cornea thickness between normal and wounded stroma in TG, the same occurred with the cornea epithelium. In the CG, a significant difference was found between normal and wounded stroma thickness, but not between normal and wounded corneal epithelium.

Discussion and Conclusions

Blepharospasm was probably caused by friction of the suture with the palpebral conjunctiva (Barros et al. 1998) and its spontaneous resolution can be attributed to the removal of sutures and membranes; the last one was degrading at this time point. Results suggest that the inflammatory reaction occurred due to the amniotic membrane, which in turn led to mucous ocular discharge and persistent conjunctival congestion in the treated group animals. As reported by Azuara-Blanco et al. (1999), amniotic membrane shows bacteriostatic properties and does not show immunogenicity if it is allogeneic. Additionally, both the inflammatory inhibitors found in the epithelial cells of the non-preserved amniotic membrane (Hao et al. 2000) and the absence of leukocyte in the amnion confirm biocompatibility and so, allows halo-transplants (Trelford & Trelford-Sauder 1979). Ocular discharge in control animals was considered normal since ulcerative processes are related to corneal and conjunctival processes due to the stimulus of the caliciform cells (Kern 1990). Conjunctival vascular congestion in control animals was less important then in treated group and was probably due to the trauma caused by the sutures applied during surgery to stabilize the eyeball. Histology confirmed clinical results observed in fluorescein test, indicating that amniotic membrane accelerated the beginning of the corneal repair process as also reported by Woo et al. (2001). However, the membrane delayed such process, mostly during its latter period, like reported by Sampaio (2006). Wilson et al. (2001) reported that interleukin 1 (IL-1)--present in tear film and released by injured epithelial cells of the cornea and inflammatory cells--bind to receptor in keratinocytes and fibroblasts and modulate apoptosis. The amniotic membrane used on corneas of the treated group acted retaining inflammatory cells on its surface and played an important role protecting the lesion, as a patch. In TG, IL-1 probably did not bind to keratinocytes in the stroma in the same intensity that it has in the control corneas. As a result, keratinocytes in treated group corneas can present a less intense apoptosis process. Thus, the remaining keratinocytes in the treated corneas can start proliferating and migrating, beginning an earlier repair process in comparison to the control corneas. Corneal neovascularization results indicate that the amniotic membrane lost its antiangiogenic property and that it delayed corneal repair in the latter observation dates, resulting in vascular formation. This has contributed to perpetuate the aggression and indicated a chronic response in treated group animals. Corneal repair occurred more rapidly in control animals and was not associated to corneal vascularization. Corneal neovascularization can occur depending on the aggression type and duration; not-complicated lesions are repaired without vascularization (Slatter & Hakanson 2007). Dua et al. (2004) stated that angiogenic chemical inhibitors would be present in fresh amniotic membranes and, therefore, this would be more effective than preserved membranes. The presence of the preserved amniotic membrane can also explain corneal opacity by delaying the repair process and stimulating corneal vascularization. Neovascularization allowed granulation tissue deposition and cicatrix formation; the latter was seen more dense and opaque in treated than control corneas where an avascular repair process occurred (Slatter & Hakanson 2007). It was also seen that stromal edema was only present in animals which epithelial regeneration was absent or incomplete. Slatter & Hakanson (2007) have described that the hydrophilicity of stromal mucopolysaccharides and collagen influence water entrance in the cornea and the epithelium acts as a barrier against this mechanism. According to Spencer (1996), the stroma thickens when edema occurs or when it is invaded by inflammatory cells during acute inflammatory phase. An important number of inflammatory cells were seen in the injured stroma of the control animals, what can explain greater mean values for thickness in this area in comparison to normal ones. Treated group did not show significant values for inflammatory cells except around the sutures--explaining the non-existence of a difference on the thickness between normal and injured stroma. As the epithelization took longer periods of time in the group treated with the amniotic membrane used as a patch, neovascularization occurred during the repair process allowing more fibroblasts deposition. This resulted in a significant corneal opacity that persisted during the whole experiment in the treated group. In conclusion, xenologous amniotic membrane preserved in glycerin and used as a patch on superficial corneal ulcers shows great benefits during initial corneal repair, but need to be removed before seven days after its implantation.

Acknowledgments

Thanks to Ophthalmos Farmaceutic Industry Ltd. and to Petrovich Surgical Instruments for the material support in this work.

References

1.  Barros PSM, Safatle AMV, Godoy CA, Souza MSB, Barros LFM, Brooks DE. 2005. Amniotic membrane transplantation for the reconstruction of the ocular surface in three cases. Vet. Ophthalmol. 8:189-192.

2.  Cremonini DN, Ranzani JJT, Marques MEA, Rodrigues GN, Brandão CVS. 2007. Transplante de membrana amniótica canina criopreservada para cicatrização de córnea com deficiência de células límbicas em coelhos. Arq. Bras. Med. Vet. Zootec. 59 (6):14621467.

3.  Dua HS, Gomes JAP, King AJ, Maharajan VS. 2004. The amniotic membrane in ophthalmology. Surv. Ophthalmol. 49:51-77.

4.  Kim JC, Tseng SCG. 1995. Transplantation of preserved human amniotic membrane for surface reconstruction in severely damaged rabbit corneas. Cornea. 14 (5):473-484.

5.  Sampaio IBM. 2001. Estatística aplicada à experimentação animal. Fundação de Ensino e Pesquisa em Medicina Veterinária e Zootecnia, Belo Horizonte, p.184-265.

6.  Barros PSM, Garcia JA, Laus JL, Ferreira AL, Gomes TLS. 1998. The use of xenologous amniotic membrane to repair canine corneal perforation created by penetrating keratectomy. Vet. Ophthalmol. 1:119-123.

7.  Azuara-Blanco A, Pillai CT, Dua HS. 1999. Amniotic membrane transplantation for ocular surface reconstruction. Br. J. Ophthalmol. 83:399-402.

8.  Hao Y, Ma DH, Hwang DG. 2000. Identification of antiangiogenic and antiinflammatory proteins in human amniotic membrane. Cornea. 19:348-352.

9.  Trelford JD, Trelford-Sauder M. 1979. The amnion in surgery, past and present. Am. J. Obst. Ginecol. 134:833-845.

10. Kern TJ. 1990. Ulcerative keratitis. Vet. Clin. North Am. Small Anim. Pract. 20:643-666.

11. Woo H-M, Kim MS, Kweon O-K, Kim D-Y, Nam T-C, Kim JH. 2001. Effects of amniotic membrane on epithelial wound healing and stromal remodelling after excimer laser keratectomy in rabbit cornea. Br. J. Ophthalmol. 85:345-349.

12. Sampaio RL, Ranzani JJT, Júnior VR, Stacciarini MS, Braga EM, Borrezzi C. 2006. Aspectos clínicos e imunopatológicos da ceratoplastia com membrana amniótica xenógena fresca e conservada em glicerina. Estudo experimental em coelhos. Arq. Bras. Med. Vet. Zootec. 58 (6):1077-1085.

13. Wilson SE, Mohan RR, Mohan RR, Júnior RA, Hong JW, Lee JS. 2001. The corneal wound healing response: cytokine-mediated interation of the epithelium, stroma, and inflammatory cells. Progress in Retinal and Eye Research. 20 (5):625-637.

14. Slatter D, Dietrich U. 2007. Córnea e esclera, p.1368-1395. In: Slatter D. (Ed), Manual de cirurgia de pequenos animais. 3^rd ed. Manole, São Paulo.

15. Spencer WH. 1996. Cornea, p.133-157. In: Ophthalmic Pathology. 4^th ed. W. B. Saunders, Philadelphia.