Cornea 20(4): 414–420, 2001.
, Philadelphia
Reconstruction of Ocular Surface With Heterologous
Limbal Epithelium and Amniotic Membrane in a
Rabbit Model
Marcel Avila, M.D., Mauricio España, M.D., Crisanto Moreno, M.D., and
Claudia Peña, D.D.S.
Purpose. To report in vivo reconstruction of the ocular surface
using amniotic membrane and heterologous transplants of epithe-
lial limbal cells in rabbits with chemical burns. Methods. After
severe damage to the ocular surface with n-heptanol and keratec-
tomy, 15 rabbits developed total limbal deficiency with conjunc-
tival epithelialization, vascularization, and chronic inflammation.
One month later, a complete keratectomy was performed in all
eyes: 12 received additional transplantation of human amniotic
membrane and heterologous limbal epithelial cells in a double
amniotic membrane layer, 2 received amniotic membrane only,
and 1 control eye received no procedure. Results. After 1 month of
follow-up, corneas in eight of the operated eyes presented minimal
vascularization, without signs of rejection. Corneal surface recon-
struction was demonstrated with the growth of new corneal-like
epithelial phenotype and integration of amniotic membrane to the
basal corneal surface. A superficial amniotic membrane (with the
amnion side up as a dressing) peeled off after 7 to 10 days. The
epithelialization with heterologous limbal epithelial cells was evi-
dent underneath. The other four operated eyes were followed for 6
months; the ocular surface was also stable with a corneal-like
epithelial phenotype. Conclusion. Simultaneous transplantation of
amniotic membrane and heterologous limbal epithelial cells in
severe ocular surface disorders could restore ocular surface and
may be useful in patients with severe bilateral limbal epithelial
loss, giving new perspectives for the treatment of severe ocular
surface disorders.
Key Words: Surface reconstruction—Amniotic membrane—
Epithelial transplantation procedure—Ocular surface disorders—
Chemical burns.
The ocular surface is covered by corneal, conjunctival, and lim-
bal epithelia, which are formed by distinct differentiated pheno-
type cell lines that maintain the ocular surface integrity.1–3 When
the limbus is destroyed, severe ocular surface disease occurs. This
disorder includes epithelial defects, recurrent erosions, chronic
stromal inflammation, corneal vascularization, and conjunctival
Submitted September 21, 2000. Revision received January 17, 2001.
Accepted January 18, 2001.
From Departamento de Oftalmología, Facultad de Medicina (M.A.,
M.E., C.M.), and Facultad de Odontología (C.P.), Hospital San Juan de
Dios, Universidad Nacional de Colombia, Colombia.
Address correspondence and reprint requests to Dr. M. Avila, Depart-
ment of Physiology, University of Pennsylvania, School of Medicine, 3700
Hamilton Walk, Philadelphia, PA 19104-6085, U.S.A. E-mail:
mavila@mail.med.upenn.edu
414
© 2001 Lippincott Williams & Wilkins, Inc.
epithelial ingrowth, leading to corneal blindness.4,5 The manage-
ment of this condition relies on the transplantation of limbal au-
tografts containing limbal stem cells obtained from the contralat-
eral eye.6 In the case of bilateral damage, there is no source for
limbal transplantation, so epithelial allografts of limbus are pro-
posed. Postoperative management requires systemic immunosup-
pression and the use of cyclosporin A.7,8 Tissue rejection is ob-
served in some patients; the risk is enhanced when vessels are
present in the receptor.9
Amniotic membrane has been used for ocular surface recon-
struction, initially by De Rötth10 as conjunctival graft in sym-
blepharon repair. Kim and Tseng11 reported the reconstruction of
rabbit ocular surface in a chemical burn model, with the epitheli-
alization of corneal surface with cells expressing a corneal type
keratin. The combination of this approach and allograft limbal
transplantation has been used for reconstruction of the ocular sur-
face in bilateral damage. However, allograft limbal transplantation
requires modulation of the immune response.12,13 These difficul-
ties lead us to investigate whether transplantation of human am-
niotic membrane and heterologous sheets of limbal epithelium
obtained by enzymatic fragmentation could be useful to restore the
ocular surface, with the advantage that no systemic immunosup-
pression is required.
MATERIALS AND METHODS
Rabbit Model of Ocular Surface Injury
All animals were handled according to the ARVO (Association
for Research in Vision and Ophthalmology) statements for the use
of animals in ophthalmic and vision research guides.
Fifteen New Zealand albino rabbits (weighing 2.5–3.0 kg) were
anesthetized with intravenous pentobarbital (20–30 mg/kg), xyla-
zine hydrochloride (10 mg/kg), and ketamine hydrochloride (15
mg/kg), complemented by topical proparacaine. The ocular surface
damage was performed in the right eye of each rabbit. Each eye
was treated with n-heptanol to remove the entire corneal epithe-
lium and limbus, this was followed by resection of limbus in 360°.
To ensure corneal and limbal epithelial destruction, this procedure
was repeated 3 weeks later. One month after the injury, all eyes
showed moderate to severe corneal vascularization with surface
irregularities and haze. Twelve rabbits received amniotic mem-
brane transplantation with heterologous epithelial cells, two eyes
received no treatment (only keratectomy), and one received am-
niotic membrane alone.
 RECONSTRUCTION OF OCULAR SURFACE
Preparation of Human Amniotic Membrane
Amniotic membrane and human placenta were obtained at the
time of elective cesarean section when human immunodeficiency
virus, human hepatitis B and C, and syphilis had been excluded by
serologic tests; blood from the umbilical cord was also seronega-
tive. Under a laminar flow hood, the placenta was cleaned of blood
clots with saline solution containing penicillin (50 ␮g/mL; ICN,
Costa Mesa, CA, U.S.A.), streptomycin (50 ␮g/mL; ICN), ampho-
tericin B (2.5 ␮g/mL; Bristol Myers Squibb Co., Princeton, NJ,
U.S.A.), and clindamycin (60 ␮g/mL; Pharmacia & Upjohn, Va-
lencia, Venezuela). The amnion was separated from the rest of the
chorion by blunt dissection. Approximately 3 × 3 cm of the am-
niotic membrane was placed in a plastic sterile container with
Dulbecco’s modified Eagle’s medium (ICN) and glycerol (ICN) at
a 1:1 ratio and preserved at −80°C, until use.
Amniotic Membrane Transplantation and
Heterologous Limbal Grafts
One month after the ocular surface injury, our procedure was
performed. Under general anesthesia as previously described, a
total lamellar keratectomy with Westcott scissors and blunt spatula
was performed, approximately 3 to 4 mm posterior to the limbus.
Amniotic membrane was placed over the exposed sclera and cor-
nea with the epithelial side up. Then the membrane was secured
with 9-0 sutures, with stitches to episclera. Heterologous limbal
epithelial cells were obtained from 2-month-old New Zealand al-
bino rabbits from a different colony. Subsequently, animals were
killed in accordance with ARVO guidelines.
After enucleation of the globes, the limbal area was preserved
carefully by leaving a rim of 4 mm undisturbed, excising the
central corneal area, and using the corneoscleral rim. Limbal ep-
ithelial cells were released from the corneoscleral rim as a coherent
sheet after incubation with neutral protease (Dispase II, 2.4 IU/mL;
Boehringer Mannheim Biochemicals, Indianapolis, IN, U.S.A.) at
37°C for 90 minutes. The sheets of cells were placed over the
amniotic membrane with the basal cells down. Then, a second
amniotic membrane was placed over the graft with the epithelial
face down and was secured with six 8-0 sutures (Fig. 1). After
surgery, lomefloxacin (Okacin; Ciba Vision, Basle, Switzerland)
and dexamethasone ointment (Ultracortenol; Ciba Vision) were
applied twice daily for 10 days.
Ocular Surface Evaluation
Rabbits were examined using a hand-light daily and a slit-lamp
weekly. No signs of complication were evident. The corneal sur-
FIG. 1. Scheme of surgical procedure. First, amniotic membrane is
fixed after removal of conjunctival epithelium with mesenchymal face
down. Limbal epithelial heterologous cells are taken from donors
with Dispase II and a second amniotic membrane is placed over the
graft; this membrane is peeled off.
415
face was also examined for reepithelialization (using fluorescein
drops) and for smoothness, clarity, and vascularization.
Immunofluorescent Studies of Epithelial Phenotypes
The following antibodies were used: AE5, mouse monoclonal
antibody specific for keratin K3 corneal-type differentiation (ICN)
and keratin 13, positive for marking conjunctiva and negative for
cornea (ICN).
Frozen tissue sections (5–8 ␮m) were incubated for 60 minutes
at room temperature with the primary antibody. The tissue sections
were rinsed three times with phosphate-buffered saline and then
incubated with fluorescein isocyanate-conjugated anti–mouse im-
munoglobulin G secondary antibodies. After rinsing, the sections
were examined. Parts of the samples were fixed on 10% formalin
in phosphate-buffered saline for periodic acid-Schiff staining.
Cytology was also performed with the group of 6-month follow-
up rabbits, again using the antibodies AE5 and anticytokeratin 13
(K13). Briefly, the cells from the central cornea (2 × 2 mm) were
removed carefully with a blunt spatula and fixed with acetone, and
then incubated in the same manner as the frozen sections previ-
ously described.
RESULTS
One month after chemical damage, the rabbit corneas showed
vascularization and conjunctivalization of the ocular surface, as
previously demonstrated.11
Amniotic Membrane Transplantation and
Heterologous Limbal Cell Grafts
One week after transplantation, the second amniotic membrane
was peeled off and/or retired, revealing complete epithelization
underneath. This finding was observed in treated eyes. These eyes
showed a smooth surface, minimal vessels, and a clear stroma.
Two eyes displayed moderate scar formation. Controls presented
diffuse vascularization with stromal opacity. The eyes that re-
ceived amniotic membrane alone did not present complete healing
with nonuniform fluorescein staining at the 6-month follow-up
(Figs. 2 and 3).
Ocular Surface Evaluation
Fluorescein staining of the experimental group showed com-
plete epithelialization. Healing occurred approximately 2 weeks
after surgery, displaying an avascular surface. Controls exhibited
delayed healing, with vascularization and severe stromal scar for-
mation (Fig. 2). The 6-month group also exhibited complete heal-
ing with minimal vascularization (Fig. 3).
Epithelial Phenotypical Studies
Epithelial phenotype was evaluated by immunofluorescence or
cytology using AE5 antibodies and anticytokeratin 13 (K13). A
corneal epithelial phenotype was identified by a positive AE5 and
a negative K13. A conjunctival phenotype was established by a
negative AE5 and a positive K13. Treated, control, and normal
corneas were all stained. In the cytology studios, this parameter
was also considered. AE5 was positive in all treated eyes, except
for one eye in which damage in the preparation precluded defini-
tive identification (Table 1). Cytology of the 6-month group was
also positive for AE5 antibodies (Table 2). Anticytokeratin 13 was
Cornea, Vol. 20, No. 4, 2001
416 M. AVILA ET AL.

FIG. 2. External examination. A: Experimental success case with smoother surface with minimal
vascularization and clear stroma. B: The same eye showing complete epithelialization with fluorescein
staining. C: Control rabbit eye with diffuse vascularization, granuloma, and stromal opacity. D: Control
with fluorescein staining showing delayed healing.

positive in normal conjunctiva and in controls eyes (Fig. 4). In one
treated eye, focal staining was observed (Table 1). In the 6-month
group, one eye also showed partial staining with this antibody
(Table 2).
Histologic Findings
On histologic section with periodic acid-Schiff, goblets cells
were found in the control eyes and in normal conjunctiva. In
contrast, normal corneas and transplanted eyes presented only
stratified corneal epithelial cells. No goblet cells and no signs of
rejection were observed on histologic analysis (Fig. 5).
DISCUSSION
Limbal transplantation has proven effective in restoring the cor-
neal surface after permanent chemical damage. This procedure is
often performed with autologous grafts. However, when limbal
cells are not available from the contralateral eye, heterologous
limbal transplantation is required. This procedure involves the use
of systemic immunosuppression, and final results are often disap-
pointing.14 Amniotic membrane has been previously used as a
graft for skin burns, skin ulcers, and the prevention of tissue ad-
hesions in surgery.15
De Rötth10 in 1940 reported the use of amniotic membrane for
conjunctival repair and symblepharon. Tseng has used amniotic
membrane for ocular surface reconstruction in a rabbit model with
success. It has also been used for reconstruction of ocular surface
in patients with chemical and thermal burns, by using autologous
limbus grafts.16
In the present work, we obtained a successful corneal epithelial
reconstruction with the use of heterologous limbal cell transplan-
tation and amniotic membrane, in contrast to controls that showed
features of conjunctivalization, vascularization, chronic inflamma-
tion, and conjunctival epithelialization.
The epithelial phenotype in rabbits with amniotic membrane and
heterologous limbal cell transplantation stained similarly to the
normal corneal epithelial phenotype (positive AE5 and negative
K13), in contrast to the controls’ staining (negative AE5) and to
the amniotic membrane alone. These findings suggest that recon-
structed rabbit corneal epithelium stain similarly to normal corneal
epithelium.17
Dispase II is useful in the separation of intact epithelial limbal
cells in culture from the substrate, which are required for posterior
transplantation. Dispase cleaves the basement membrane while
maintaining the viability of epithelial cells.18
Based on these findings, we conclude that sheets of limbal ep-
ithelial cells could be obtained and used as grafts over a new
basement membrane provided by the amniotic membrane.19,20
Basal membrane is essential for growth and attachment of epithe-
lial cells. In ocular surface reconstruction, this element is crucial
for successful establishment of the new epithelium.
The amniotic membrane is a thick basement membrane and
secretes various cytokines that could produce an ideal microenvi-

Cornea, Vol. 20, No. 4, 2001

 RECONSTRUCTION OF OCULAR SURFACE 417

FIG. 3. External examination of 6-month follow-up group. A: Preoperative eye. B: Six-month postop-
erative eye with smooth surface and clear cornea. C: Membrane amniotic alone with fluorescein
staining. D: Other postoperative eye with smooth surface, complete healing, and clear cornea.

ronment to promote the growth of epithelial cells.21,22 Other se- membrane may act as a semipermeable membrane, allowing the
creted factors, such as basic fibroblast growth factor and trans- diffusion of nutrients and avoiding the passage of antibodies,
forming growth factor beta, may modulate subconjunctival fibrosis complement proteins, and antigen-presenting cells. Limbal trans-
of the grafts.16 plantation procedures require a carrier tissue, either conjunctival
Amniotic membrane does not express human leukocyte antigens grafts, peripheral cornea, or both.23 However, conjunctival grafts
under normal circumstances, thus immunologic rejection is not a contain blood vessels and lymphatics, epithelium that includes
concern.22 Based on this concept, we speculate that the trans- Langerhans cells, and stroma including macrophages and dendritic
planted cells, “hidden” within our double amniotic membrane sys- cells. In experimental studies, conjunctiva elicit a vigorous allo-
tem, diminish the risk of rejection. We also conclude that amniotic immunity and severe immunologic rejection.24 Because vessels
and stroma are not transplanted in this model, rejection could
thereby be diminished.
TABLE 1. Immunostaining of differentiation-related protein
expression in reconstituted epithelium formed from heterologous Cultured corneal epithelia have been used as grafts in ocular
limbal epithelial cells grafted and amniotic membrane
transplantation (1-month follow-up)
TABLE 2. Immunostaining of differentiation-related protein
AE5 (Keratin 3) expression in reconstituted epithelium formed from heterologous
Corneal-type limbal epithelial cells grafted and amniotic membrane
Rabbit number differentiation Keratin 13 transplantation in cytology studies at 6 months of follow-up
Normal cornea + − AE5 (Keratin 3)
Normal limbus +/suprabasal Weak staining corneal-type
32 (treated) + − Rabbit number differentiation Keratin 13
33 (treated) + −
34 (treated) + − Normal cornea + −
35 (treated) ND − Normal conjunctiva − +
36 (treated) + − 1A (treated) + +
37 (treated) + − 1B (ma alone) − +
38 (keratectomy) − + 2A (treated) + −
39 (treated) +/focal staining +/focal staining 2B (treated) + −
40 (treated) + − 3A (keratectomy) − +
3B (treated) + −
+ indicates positive staining; −, negative staining; ND, not deter-
mined. + indicates positive staining; −, negative staining.

Cornea, Vol. 20, No. 4, 2001

418 M. AVILA ET AL.

FIG. 4. Immunofluorescent micrograph staining of the treated eye of rabbit 32 (A) with monoclonal
antibody AE5 and the treated eye of rabbit 36 (B) with AE5. These cases were negative for K13,
showing that epithelial phenotype is corneal. C: Control case with monoclonal antibody K13 (conjunc-
tival marker). This case was negative for AE5. D: Normal conjunctiva with K13. E: Normal cornea with
AE5. F: Normal limbus with AE5.

surface reconstruction. However, this procedure requires coculti-
vation with 3T3 irradiated cells and then transfer to a carrier.25 The
use of amniotic membrane with the epithelial grafts avoids cocul-
tivation, and the second amniotic membrane as a dressing patch
(epithelial face down) allows the reepithelialization under this
membrane, as shown in previous reports.26,27
The rapid epithelialization seen in our model may depend on the
source of the epithelial donors that we used (2-month-old rabbits).
According to Holland, the youngest tissue is preferred (including
neonatal) because the limbal stem cells would be more active than
those from human adults.28,29 Limbal stem cells are present in the
limbus, but it is difficult to demonstrate stem cells with a specific
marker. These cells can differentiate to other cells with high re-
generative capacity. However, if these cells transform to a more
mature epithelial form, this capacity will decrease and the epithe-
lium will be exhausted and, with time, will be changed by con-
junctival host cells. To address this problem, we followed a group
of rabbits for 6 months that were transplanted with our purposed
method.
The stability of ocular surface, characterized by a smooth sur-
face and clear cornea, and the presence of phenotypical corneal
epithelium persuade us that stem cells are present as a continuous
source of corneal epithelial cells. This concept is also supported by
the findings of Meller et al. They found that amniotic membrane
prolongs the lifespan of the epithelial progenitor cells and pro-
motes the expansion of limbal epithelial progenitor cells.30 The
second amniotic membrane is also beneficial in acting as a bio-
logic dressing patch. We did not find allograft rejection in the 1st
month or in the 6-month group. This suggested that the amniotic
membrane protected these cells from rejection.

Cornea, Vol. 20, No. 4, 2001

 RECONSTRUCTION OF OCULAR SURFACE
FIG. 5. A and B: Periodic acid-Schiff staining of reconstructed corneal epithelium of treated eye. Note
the absence of goblet cells and the well-polarized basal cells. C: Bulbar conjunctiva with some vas-
cularization and some goblet cells (arrows). D: Control cornea with many goblet cells (arrows) and
some infiltrate.
Xu et al.31 found that limbal allografts were 85% rejected at 28
days with no treatment and that they were stable for 45 ± 8 days
when cyclosporine was used for 28 days. Extrapolating these re-
sults to our work, amniotic membrane appears to have immuno-
protective properties for the transplanted cells. First, we concluded
that it acts as a barrier to immune cells. This hypothesis is sup-
ported by the findings of Heiligenhaus et al., who described the
ability of amniotic membrane to increase the apoptosis of neutro-
phils.32 Second, the amniotic membrane antiinflammatory proper-
ties by inhibiting IL1-␤ and IL 8 expression decrease the immune
response. Finally, the amniotic membrane provides antiangiogenic
proteins.33 All of these properties of the amniotic membrane and
the fact that only the epithelium is transplanted suggest that both
elements are important in maintaining the transplanted epithelial
cells without immunologic rejection. The transplantation of amni-
otic membrane with heterologous limbal cells seems to offer a new
perspective for ocular surface reconstruction, especially in bilateral
ocular surface lesions.
Acknowledgments: The authors thank Dr. Mortimer Civan for his edi-
torial advice in this manuscript and German Santamaria for technical as-
sistance.
REFERENCES
1. Thoft RA, Friend J. Biochemical transformation of regenerating ocular
surface epithelium. Invest Ophthalmol Vis Sci 1997;16:14–20.
419
2. Wei ZG, Sun T-T, Lavker R. Rabbit conjunctival and corneal epithe-
lial cell belong to two separate lineages. Invest Ophthalmol Vis Sci
1996;37:523–33.
3. Schermer S, Galvin S, Sun T-T. Differentiation-related expression of
a major 64K corneal keratin in vivo and in culture suggests limbal
location of corneal epithelial stem cells. J Cell Biol 1986;103:49–62.
4. Tseng SCG. Concept and application of limbal stem cells. Eye 1989;
3:141–57.
5. Tseng SCG. Regulation and clinical implications of corneal epithelial
stem cells. Mol Biol Rep 1996;23:47–58.
6. Kenyon KR, Tseng SCG. Limbal autograft transplantation for ocular
surface disorders. Ophthalmology 1989;96:709–20.
7. Holland E, Schwartz G. The evolution of epithelial transplantation for
severe ocular disease and a proposed classification system. Cornea
1996;15:549–56.
8. Tsai RJF, Tseng SCG. Human allograft limbal transplantation for cor-
neal surface reconstruction. Cornea 1994;13:389–400.
9. Thoft RA, Sugar J. Graft failure in keratoepithelioplasty. Cornea 1993;
12:362–5.
10. De Rötth A. Plastic repair of conjunctival defects with fetal mem-
branes. Arch Ophthalmol 1940;23:522–5.
11. Kim J, Tseng SCG. Transplantation of preserved human amniotic
membrane for surface reconstruction in severely damaged rabbit cor-
neas. Cornea 1995;14:473–84.
12. Tsubota K, Satake Y, Ohyama M, et al. Surgical reconstruction of the
ocular surface in advanced ocular cicatricial pemphigoid and Stevens-
Johnson syndrome. Am J Ophthalmol 1996;122:38–52.
13. Tseng SCG, Prasbhasawat P, Barton K, et al. Amniotic membrane
transplantation with or without limbal allografts for corneal surface
reconstruction in patients with limbal stem cell deficiency. Arch Oph-
thalmol 1998;116:431–41.
Cornea, Vol. 20, No. 4, 2001
420 M. AVILA ET AL.
14. Tan DTH, Ficker L, Buckley R. Limbal transplantation. Ophthalmol-
ogy 1996;103:29–36.
15. Trelford JD, Trelford-Sauder M. The amnion in surgery, past and
present. Am J Obstet Gynecol 1979;134:833–45.
16. Shimazaki J, Yang HY, Tsubota K. Amniotic membrane transplanta-
tion for ocular surface reconstruction in patients with chemical and
thermal burns. Ophthalmology 1997;104:2068–75.
17. Lindberg K, Brown M, Chavez H, et al. In vitro propagation of human
ocular surface epithelial cells for transplantation. Invest Opthalmol Vis
Sci 1993;34:2672–9.
18. Spurr SJ, Gipson IK. Isolation of corneal epithelium with Dispase II or
EDTA. Effects on the basement membrane zone. Invest Ophthalmol
Vis Sci 1985;26:818–27.
19. Fukuda K, Chikama T, Nakamura M, et al. Differential distribution of
subchains of the basement membrane components type IV, collagen
and laminin among the amniotic membrane, cornea and conjunctiva.
Cornea 1999;18:73–9.
20. Lwebuga-Mukasa JS, Thulin G, Madri JA, et al. An acellular human
amniotic membrane model for in vitro culture of type II pneumocytes:
the role of the basement membrane in cell morphology and function.
J Cell Physiol 1984;121:215–25.
21. Li DQ, Tseng SCG. Three patterns of cytokine expression potentially
involved in epithelial-fibroblast interactions of human ocular surface.
J Cell Physiol 1995;163:61–9.
22. Akle CA, Adinolfi M, Welsh KI, et al. Immunogenicity of human
amniotic epithelial cells after transplantation into volunteers. Lancet
1981;2:1003–5.
23. Dua HS, Azuara-Blanco A. Limbal stem cells of the corneal epithe-
lium. Surv Ophthalmol 2000;44:415–25.
Cornea, Vol. 20, No. 4, 2001
24. Hori J, Joyce N, Streilein JW. Epithelium-Deficient corneal allografts
display immune privilege beneath the kidney capsule. Invest Ophthal-
mol Vis Sci 2000;41:443–52.
25. Pellegrini G, Traverso C, Franzi A, et al. Long-term restoration of
damaged corneal surfaces with autologous cultivated corneal epithe-
lium. Lancet 1997;349:990–3.
26. Trelford JD, Anderson DG, Hanson F, et al. Amnion autografts and
allografts as a cover for skin defects in sheep. J Med 1972;3:81–7.
27. Trelford JE, Hanson F, Anderson D. Amniotic membrane as a living
surgical dressing in human patients. Oncology 1973;28:358–64.
28. Voelker R. Stem cells hold vision for the future. JAMA 1997;278:
1477–8.
29. Croasdale C, Schwartz G, Malling J, et al. Keratolimbal allograft:
recommendations for tissue procurement and preparation by eye
banks, a standard surgical technique. Cornea 1999;18:52–8.
30. Meller D, Pires R, Tseng SCG. Ex-vivo preservation and expansion of
human limbal epithelial progenitor cells by amniotic membrane [ab-
stract]. Invest Ophthalmol Vis Sci 1999;40(suppl 329):1744.
31. Xu KP, Wu Y, Zho J, et al. Survival of rabbit limbal stem cells
allografts after administration of cyclosporine. Cornea 1999;18:
459–65.
32. Heiligenhaus A, Bauer D, Mrzyk S, et al. Amniotic membrane trans-
plantation improves murine herpes stromal keratitis by induction of
neutrophil apoptosis [abstract]. Invest Ophthalmol Vis Sci 1999;40
(suppl 257):1359.
33. Hao YX, Ma DHK, Kim WS, et al. Identification of anti-
neovascularization proteins in human amniotic membrane [abstract].
Invest Ophthalmol Vis Sci 1999;40(suppl 579):3045.