Veterinary Ophthalmology (2016) 1–10 DOI:10.1111/vop.

12392

Corneal grafting for the treatment of full-thickness corneal defects

in dogs: a review of 50 cases
Rodrigo P. Lacerda,* Maria T. Pe~
na Gimenez,† Fernando Laguna,* Daniel Costa,† Jose Rıos‡ and
Marta Leiva†
*Servei d’Oftalmologia, Fundaci
o Hospital Cl
ınic Veterinari, Carrer del Hospital s/n, Bellaterra, Barcelona 08193, Spain; †Departament de Medicina i
Cirurgia Animal, Facultat de Veterinaria, Universitat Autonoma de Barcelona, Edific V, Campus UAB, Bellaterra, Barcelona 08193, Spain; and ‡Medical
Statistics Core Facility, DIBAPS (Hospital Cl
ınic), Barcelona 08036, Spain

Address communications to: Abstract

M. Leiva
Tel.: +34-93-581-1387
Fax: +34-93-581-2006
e-mail: marta.leiva@uab.cat
Objective To describe corneal grafting for the treatment of full-thickness corneal
defects in dogs and to determine its effectiveness in preserving vision.
Methods A review of the medical records of dogs that underwent corneal grafting
following corneal perforations (≥3 mm) at the VTH-UAB from 2002 to 2012 was
carried out.
Results Fifty dogs of different breed, age and gender were included. Brachycephalic
breeds were overrepresented (37/50;74%). All cases were unilateral, with euryble-
pharon being the most common concurrent ocular abnormality (20/50;40%). Full-
thickness penetrating keratoplasties (FTPK) were performed in 21/50 eyes (42%) and
lamellar keratoplasties (LK) in 29/50 eyes (58%). Frozen grafts (FroG) were used in
43/50 eyes (86%) and fresh homologous grafts (FreHoG) in 7/50 (14%). Of the for-
mer group, 26 were homologous (FroHoG:60%) and 17 heterologous (FroHeG:40%).
A combination of topical medication (antibiotics, corticosteroids, cycloplegics, and
0.2% cyclosporine A) and systemic mycophenolate mofetil was administered. Median
follow-up time was 200 days. Postsurgical complications included wound dehiscence
(6/50;12%) and glaucoma (4/50;8%). Clinical signs of graft rejection were diagnosed
as follows: FroHoG (13/26;50%), FroHeG (11/17;65%), FreHoG (4/7;57%), FTPK
(12/21;57%), and LK (16/29;55%). Medical treatment successfully controlled graft
rejection in 11/28 eyes (39%). Good anatomical outcome was achieved in 86% (43/
50), of which 95% (41/43) were visual at last examination, with moderate opacification
to complete transparency of the graft present in 48.2%.
Conclusions Corneal grafting is an effective surgical treatment for full-thickness corneal
defects in dogs. If graft rejection is present, additional medical or surgical therapy may
be necessary, achieving a highly satisfactory visual outcome.

Key Words: allograft, corneal ulcer, lamellar keratoplasty, penetrating keratoplasty,

rejection, xenograft

ophthalmology, the mainstay of treatment for this presen-

INTRODUCTION
Corneal perforation is a frequent clinical presentation in
human and veterinary ophthalmology. The causes vary
mildly between species but, in people, corneal infection is
the most common etiology, followed by inflammation,
trauma, xerosis, exposure, neurotrophic, degeneration/
ectasia, and surgery.1 In veterinary patients, infection,
inflammation, trauma, surgery, and feline corneal seques-
trum have been described as frequent causes of corneal
perforation.2–5 In both human and veterinary
tation is surgery. Several techniques have been used in
humans for this purpose such as lamellar/penetrating ker-
atoplasty,6–8 keratoprosthesis,9 tissue adhesives,10 synthetic
grafts11, and biomaterials.12 In domestic species, apart
from these techniques, others have been developed: cor-
neoscleral transposition,13 conjunctival grafts,14 amniotic
and renal membrane,15,16 peritoneum grafts,17 and carti-
lage.18
Dr. Eduard Zirm, a Czech ophthalmologist, performed
the first successful human corneal transplant in 1905. Dr.

© 2016 American College of Veterinary Ophthalmologists

2 lacerda ET AL.
Zirm transplanted both corneas of the same patient, but
despite the rejection of one of the grafts, the other
remained transparent, without the use of immunosuppres-
sive medication.19 In veterinary medicine, it was Dr. Big-
ger who first described corneal grafting in a pet gazelle in
1837.20 It was not until 1957 that Dr. Holt performed the
first documented homograft in a dog.21 Corneal grafting
surgery and postoperative care have evolved dramatically
over the following decades. Nowadays, according to the
World Health Organization, more than 120 000 corneal
transplants are performed each year worldwide, and cor-
neas are one of the most common solid tissues trans-
planted.22 The modern indications for corneal
transplantation in human and veterinary ophthalmology
can be categorized into four groups: optical, therapeutic,
tectonic, and cosmetic.23,24 Their use in canine, feline and
equine species has had promising results3,4,25 but, to the
authors’ knowledge, no study has determined the success
rates of tectonic grafts in dogs.
The purpose of this study was to describe corneal graft-
ing for the treatment of full-thickness corneal defects in
dogs and to determine its effectiveness in preserving
vision.
MATERIALS AND METHODS
Animals
Medical records of canine patients that underwent corneal
grafting surgery for the treatment of corneal perforations
(≥3 mm) at the Veterinary Teaching Hospital of the
Autonomous University of Barcelona (VTH-UAB) from
2002 to 2012 were reviewed.
Data analysis consisted of breed, age, gender, concur-
rent ocular diseases, type of keratoplasty performed
(lamellar/penetrating), corneal tissue used (homologous/
heterologous, fresh/frozen), graft size (mm), postoperative
treatment, complications, follow-up time (days), graft
healing time (days), and assessment of vision and graft
opacification at last visit.
Ophthalmic examination
All dogs in this study underwent a complete and bilateral
ophthalmic examination that consisted of Schirmer tear test
I (Schering-Plough Animal Health Corporation, Union,
NJ), neuro-ophthalmic examination (menace response, daz-
zle and pupillary light reflexes), biomicroscopy (Kowa SL-
15â; Kowa company, Tokyo, Japan), rebound tonometry
(Tonovetâ; Tiolat Oy, Helsinki, Finland), indirect fundus-
copic examination (Heine Omega 180â; Heine, Herrsching,
Germany), and fluorescein test. Only, patients with positive
consensual pupillary light and dazzle reflexes in the affected
eye were included in the study.
Donor cornea preservation
Donor grafts were obtained from the VTH-UAB Eye
Bank. This bank retrieves and stores eyes for corneal
© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
transplantation and research purposes. Donor corneal tis-
sue was harvested from dogs, cats and horses euthanized
for reasons other than systemic infection or neoplasia, and
pigs’ eyes obtained from a local slaughterhouse. The eyes
were aseptically enucleated by subconjunctival technique.
All eyes underwent routine ocular surface disinfection with
povidone iodine solution (Braunolâ; B. Braun vetcare S.
A., Barcelona, Spain), diluted in 1:50 saline. Each globe
was then placed in a sterile container and a commercially
available antibiotic solution (Oftalmotrimâ; trimethoprim–
sulfamethoxazole, Alcon cusı laboratoris, El Masnou, Bar-
celona, Spain) was applied over it (Fig. 1). The containers
were frozen at 20 °C and used as needed for a period
from 1 month to 1 year. When corneas were harvested
from fresh globes, the same preparation method was uti-
lized and the eyes were preserved refrigerated (4 °C) for a
maximum of 48 h. At the time of surgery, preference was
given to homologous fresh tissue, if possible, and if none
was available, homologous and then heterologous frozen
grafts were used.
General surgical considerations
Preoperative topical medication varied, as it depended on
the treatment regime used by the referring veterinarian
and the clinical phase of the disease prior to perforation,
but it was discontinued on admission. Before surgery, all
patients were administered systemic antibiotics (cephazolin
25 mg/kg intravenously [IV], Kurganâ; Laboratorios Nor-
mon S. A., Madrid, Spain), and systemic nonsteroidal anti-
inflammatory drugs (flunixin meglumine 0.5 mg/kg IV,
Flunixin Ynyectable Norbrookâ; Laboratorios Karizoo, S.
A., Barcelona, Spain). All dogs were routinely premedi-
cated and anesthetized; induction consisted of diazepam
Figure 1. Position of a canine globe before storage. Two non-
absorbable sutures are placed on the bulbar conjunctiva, holding the
eye inside the sterile container, avoiding eye movement and corneal
contact with the walls of the container. A pack of sterile gauzes are
used as an antibiotic reservoir.

0.5 mg/kg IV (Valiumâ; Roche Farma, S. A., Madrid,
Spain) and propofol 2–4 mg/kg IV (Vetofolâ; Laborato-
rios Dr. Esteve, S. A., Barcelona, Spain) and maintained
under general anesthesia with isoflurane 1.5–2% (Isofloâ;
Laboratorios Dr. Esteve, S. A., Barcelona, Spain) and
100% oxygen. The periocular skin was clipped and asepti-
cally cleaned with povidone iodine solution (Braunolâ; B.
Braun vetcare S. A., Barcelona, Spain), diluted in 1:10 sal-
ine and the corneal surface cleansed with saline. The
patient was positioned in dorsal recumbency with the
muzzle tilted vertically so that the surface of the iris would
be parallel to the floor. Neuromuscular blocking agent
(atracurium besylate 0.2 mg/kg IV, Tacriumâ;
GlaxoSmithKline, S. A., Madrid, Spain) was used. Two
board certified surgeons performed the procedures (MTP
and ML). The recipient cornea was surgically prepared by
excising all necrotic and collagenolytic tissue with a No.
64 beaver blade (SM64â; Swann-Morfon limited, Shef-
field, UK) whilst preserving as much healthy tissue as pos-
sible, and excessive corneal bleeding was controlled with
epinephrine sterile vials (1:1000) (Adrenalinaâ; B. Braun
vetcare, S. A., Barcelona, Spain). Sodium hyaluronate vis-
coelastic solution (Acrivet Biovisc 2%â; S&V Technolo-
gies AG, Hennigsdorf, Germany) was used whenever
needed at the beginning of the surgery.
If frozen cornea was used, it was partially thawed at
room temperature and the devitalized corneal epithelium
was removed by gentle scraping with a cellulose sponge.
Thereafter, the corneal defect was measured and the
donor button was collected, 0.5–1 mm larger in diameter.
When needed, the graft was trimmed to fit the recipient’s
corneal bed with corneal scissors. The graft was stabilized
by initially placing four cardinal sutures at 12, 6, 9, and
3-O’clock position, using 9-0 suture materials: polyglactin
910 (Vicrylâ; Ethicon, Johnson & Johnson Intl, St-Ste-
vens-Woluwe, Belgium) or nylon (Ethilonâ; Ethicon,
Johnson & Johnson Intl, St-Stevens-Woluwe, Belgium).
Simple interrupted sutures were placed in each quadrant
and spaced between each other, every 1–2 mm.
Lamellar graft preparation
A corneal trephine (F. C. I. Instruments Co., Paris, France)
was placed over the central area of the donor cornea, and
gentle downward pressure was applied in a single direction
rotational movement, allowing for a clear-cut incision, with
vertical sides. The thickness of the graft was 70–80% that of
the recipient cornea, and the button was then obtained by
keratectomy with a crescent blade knife (Ophthalmic cres-
cent knifeâ; Alcon, TX) or by sharp lamellar dissection
using a 65 Beaver blade (SM65, Swann-Mortonâ; Sheffield,
UK) or a blade made from a sterile razorblade.
Full-thickness graft preparation
When using frozen grafts, a standard ab externo surgical
approach for penetrating keratoplasty was utilized. A full-
thickness button was trephined, using the previously
© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
full-thickness corneal grafting in dogs 3
described technique for lamellar keratoplasty, and the but-
ton placed on a corneal silicon concave stand (Teflon
block), with the Descemet’s membrane side up and cov-
ered with viscoelastic.
If fresh grafts were used, an ab interno technique was
performed. The donor globe was sectioned, through 360°,
2 mm distal to the limbus. The tissue was then placed on
a Teflon block—endothelium side up—and a penetrating
keratoplasty was performed using a trephine. This tech-
nique allows a precise incision of the endothelium – vital
for graft transparency.
Postoperative treatment
Postoperative treatment consisted of topical 1% cyclopen-
tolate (Colircusı Cicloplejicoâ; Alcon cusı laboratoris, Bar-
celona, Spain) or 1% atropine (Colircusı Atropinaâ 1%;
Alcon cusı laboratoris, El Masnou, Barcelona, Spain) two to
three times daily for 5–7 days, and tobramycin (Tobrexâ;
Alcon cusı laboratoris, El Masnou, Barcelona, Spain) six
times daily for 2–4 weeks. Dexamethasone phosphate (Max-
idexâ; Alcon cusı laboratoris, El Masnou, Barcelona, Spain)
and 0.2% cyclosporine A ointment (Optimmuneâ; Shering-
Plough SA, Madrid, Spain) were administered three times
and twice daily, respectively, both for a minimum of six
weeks. From 2009 onwards, systemic mycophenolate mofe-
til (MM) 15 mg/kg once daily (Cellceptâ; Roche Farma,
Madrid, Spain) was administered to every patient. If nylon
suture material was used, removal of the sutures was carried
out in two stages; half were removed 3–4 weeks after sur-
gery and the remaining sutures, 3 weeks later.
Graft healing time, graft rejection, vision and graft
opacification assessment
Graft healing time was determined by the time it took for
complete graft re-epithelization, assessed by means of a
negative fluorescein uptake by the grafted cornea. If fresh
donor graft was used, the corneal epithelium was not
removed; therefore, graft healing time was not applicable
in these eyes.
Corneal graft rejection was characterized by one of the
following: epithelial, stromal or endothelial rejection
bands, keratic precipitates, and increase in corneal thick-
ness (edema).
Vision was assessed at the last visit by a board certified
ophthalmologist by means of a positive menace response
and/or maze test. Graft opacification was classified in five
grades: transparent, mildly opaque, moderately opaque,
severely opaque, and completely opaque. Grading was
based on high-resolution digital photographic images
(Nikonâ D3100 and Sigmaâ EM-140 EG flash) through a
computerized analysis, when available, or on the clinical
description present on the patient’s record.
Statistical analysis
Results are described using the median and range (min,
max) for continuous variables and absolute frequencies
4 lacerda ET AL.

and percentage for categorical variables. Inferential analy- Graft size ranged from 3 to 15 mm, with a median of
ses for prognostic factor for outcome were performed esti- 7.5 mm in diameter.
mating hazard ratio (HR) and their 95% confidence Polyglactin 910 suture material was used in 31 cases
intervals (95% CI) by means Cox regression models. An (62%) and nylon in 19 (38%). At the time of surgery, the
approach for a multivariate model was made using prog- following procedures were performed concurrently: can-
nostic factors with a P-value <0.10 in univariate analyses. thoplasty in 19 cases (38%); and phacoemulsification;
Study of independent risk factors for glaucoma, graft Morgan–Moore technique; nasal fold resection; Hotz-Cel-
rejection, or suture dehiscence was evaluated by means of sus; and eyelid cryosurgery, a single case each (2%).
odds ratio (OR), and their 95% CI by means of logistic
regression models. In all analyses, a bilateral Type I error Follow-up, graft healing time, and complications
of 5% was applied without a formal correction for multi- Follow-up time ranged from 34 days to 6 years (me-
plicity. The SPSS 20 (IBM, Armonk, NY) program package dian = 0.55 years). Complete graft healing time ranged
was used for the statistical analyses. from 4 to 62 days with a median of 24 days. The three
main complications observed during the study are shown
in Table 2. As seen in Table 3, diagnosis of glaucoma
RESULTS
became statistically significant in larger sized grafts, OR:
Animals 1.96 (95% CI 1.08; 3.54).
A total of 50 patients were included in the study. There In cases where graft rejection was detected, medical
were 13 different breeds (Table 1), and brachycephalic therapy was used as the first therapeutic approach; and in
dogs were the most common (74%). There were 26 entire nonresponsive cases, when the eye could be salvaged, con-
females, 22 entire males, one spayed female, and one neu- junctival graft surgery was performed. In the former
tered male. Median age was 42 months, with a range of group, 11 of 28 cases (39%) were successfully managed
4–132 months. with additional use of topical dexamethasone phosphate,
hypertonic solution, 0.2% CsA, and systemic MM. Four
Ophthalmic findings at first examination eyes could not be salvaged, but the remaining 13 refrac-
In every case, the presentation was unilateral. The right tory patients underwent surgery to control rejection, and
eye was affected in 19/50 (38%) patients and the left eye 85% (11/13) responded positively.
in 31/50 (62%). Euryblepharon was the most common
concurrent ocular disease in 20/50 (40%), but entropion Outcome and graft opacification
in 7/50 (14%), trichiasis in 6/50 (12%), keratoconjunctivi- Anatomical outcome was classified as good in 86% of the
tis sicca in 6/50 (12%), distichiasis in 3/50 (6%), prolapsed cases (43/50). Of these, 41 eyes (95%) were visual at last
gland of the nictitating membrane in 1/50 (2%), and examination. Among the studied variables (Table 4) that
lagophthalmia in 1/50 (2%) were also diagnosed. could affect the anatomical and visual outcome, the graft’s
size, with HR: 1.41 (95% CI 1.07; 1.85) and origin, with
Type and size of the corneal grafts HR: 15.75 (95% CI 1.51; 163.54) for frozen heterologous,
Lamellar grafts were performed in 29 eyes (58%) and full- were the only two that showed statistical significant risk
thickness grafts in 21 (42%). In 7 (14%) keratoplasties, factors. A poor anatomical outcome was directly associated
fresh homologous tissue was used (Fig. 2), and in 43 with larger sized grafts, HR: 1.53 (95% CI 1.08; 2.16). As
(86%) frozen tissue, of which 26 were homologous (60%) for the origin of the tissue, the group that had the lowest
(Fig. 3) and 17 heterologous (40%) (Fig. 4). risk of poor outcome was frozen homologous, followed by
fresh homologous and lastly frozen heterologous, for more
details see Table 4. Furthermore, a synergistic effect
Table 1. Canine breeds included in the study
between the origin and size of the graft was observed in
Breed Number of cases the multivariable study. When frozen heterologous and
fresh homologous grafts were used, the larger they were,
Pug 15
French Bulldog 9 the higher the likelihood of a poor outcome (HR: 33.41
Shih Tzu 8 [95% CI 2.34; 477. 47] and HR: 13.67 [95% CI 1.00;
English Bulldog 4 187.14], respectively).
American Staffordshire Bull Terrier 2 Opacification of the graft could only be determined in
Boxer 2
29 cases, as the remaining patients either underwent con-
Cross-breed 2
Chinese Shar-Pei 2 junctival graft surgery (13/50;26%) or the eye was lost due
Yorkshire Terrier 2 to secondary complications (7/50;14%) and in one case
American Cocker Spaniel 1 the patient died during the postoperative period. The cor-
Beagle 1 neal graft remained transparent in 3/29 cases (10.3%)
Chihuahua 1
(Fig. 2c), mildly opaque in 5/29 (17.2%) (Fig. 5a), moder-
Miniature Poodle 1
ately opaque in 6/29 (20.7%) (Fig. 5b), severely opaque in

© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10

 full-thickness corneal grafting in dogs 5

(a) (b) (c)

Figure 2. Fresh homologous 12 mm full-thickness keratoplasty in a 2 year-old female Boxer with corneal perforation (a). Five days post-
operative results (b). Note the transparency of the corneal graft 3 years postoperatively (c). The remaining corneal opacification corresponds to
the receptor peripheral cornea.

(a) (b) (c)

Figure 3. Frozen homologous 11 mm full-thickness keratoplasty in a 6-month-old male Chinese Shar-Pei with corneal perforation and iris
prolapse (a). Temporal lateral tarsorrhaphy was performed simultaneously (b). Note the complete graft opacification, in an otherwise potentially
visual eye (c).

5/29 (17.2%) (Fig. 5c), and completely opaque in 10/29
(34.5%) (Fig. 3c).
DISCUSSION
Full-thickness corneal defects (≥3 mm) represent an oph-
thalmological emergency, threatening both vision and
preservation of the patient’s globe. Different surgical tech-
niques have been described but to the authors’ knowledge,
no study has determined the success rates of tectonic
grafts in dogs for the treatment of corneal perforations.
The present study shows that surgical management of
these cases using corneal grafting is very efficient, whilst
achieving good anatomical and visual outcomes (86% [43/
50] and 82% [41/50], respectively).
In veterinary medicine, other surgical techniques apart
from keratoplasties have been reported to successfully man-
age corneal perforations such as: grafts of equine amniotic
membrane,15 grafts of porcine small intestine submucosa
(SIS),26,27 autologous conjunctival grafts14,28, and cor-
neoscleral or corneoconjunctival transposition (CCT)
grafts.13 Equine amniotic membrane grafts have been
associated with marked granulation tissue formation and
opacification of the graft, and its use is only recommended
for peripheral corneal lesions.15 Furthermore, the prepara-
tion of amniotic membrane for transplantation is moder-
ately complex29 and availability is limited, thus not making
this method easily accessible to all clinicians. The use of SIS
for the treatment of corneal perforations has been described
as a single procedure27 or in combination with conjunctival
pedicle graft.26 Regarding the former description, it is
worth mentioning that there is no information available
about the size of the corneal lesions treated.27 Similarly, in
the study of SIS and conjunctival pedicle graft, the size of
the corneal lesions was not provided for every case of cor-
neal perforation but, for the ones it was, the largest was
3 mm. In these small lesions, the opacification secondary to
the conjunctival graft may not be significant and the patient
may retain vision, but in larger defects it might compromise
visual outcome. Therefore, based on these reports it is not
possible to establish how useful this method is in treating
large full-thickness defects.
Although, as previously described, conjunctival and
CCT grafts have also been used for the treatment of

© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10

6 lacerda ET AL.

(a) (b) (c)

Figure 4. Frozen heterologous 9 mm full-thickness keratoplasty in a 8 year-old male Pug with corneal perforation (a). Post-operative result
after 10 days (b). Note the graft opacification, corneal neovascularization and corneal pigmentation 2 months after surgery, in an otherwise
potentially visual eye (c).

Table 2. Complications after corneal grafting for the treatment of
large full-thickness corneal defects in dogs
Complications Number of cases (n = 50) (%)
Graft rejection 28 (56)
Rejection in lamellar grafts 16/29 (55)
Rejection in full-thickness grafts 12/21 (57)
Rejection in FroHeG 11/17 (65)
Rejection in FroHoG 13/26 (50)
Rejection in FreHoG 4/7 (57)
Wound dehiscence 6 (12)
Glaucoma 4 (8)
FroHeG = frozen heterologous graft; FroHoG = frozen homologous
graft; FreHoG = fresh homologous graft.
corneal perforations, their use in large corneal perforation
is controversial, mainly due to the lack of an efficient tec-
tonic support, the need for healthy cornea to serve as
autologous graft and the limited size of corneal lesions in
which they can be used (defects smaller than 25–30% of
the corneal diameter).30 These limitations are not present
in corneal grafting.
In human ophthalmology, the success rate of corneal
grafting surgery is dependent on the immunologic/clinical
state of the recipient cornea. Keratoplasties for the treat-
ment of noninflammatory corneal dystrophies have higher
success rates (from 88% [1 year postoperative] to 62%
[10 years postoperative])31 than keratoplasties for acquired
degenerative corneal diseases, which have higher rates of
rejection.31–35 In the latter group, corneal inflammation,
infection and/or vascularization are responsible for most
of the rejected or failed grafts.31–37 The rejection rates
range from values as low as 0–10% in avascular recipient
corneas,35 to values as high as 25–50% in severely vascu-
larized corneas.34 In fact, vascularization has been deter-
mined to be the most important risk factor for rejection.
The risk of graft failure is directly associated with the
number of blood vessels per corneal quadrant if they cross
at least 2 mm from the limbus.32,34
Although in humans the rejection rate of penetrating
keratoplasties is well known, ranging from 25 to 65%
(50–65% of which respond to medical treatment),32,35 in
veterinary medicine, there are no similar reports. The pre-
sent study is the first to determine the rejection rate of
penetrating keratoplasties in dogs (rejection rate of 56%),
which is in accordance with what has been previously
described in humans. Furthermore, it is worth mentioning
that all eyes in this study were considered to be at high
risk of failure or rejection as the grafts were placed in
inflamed and vascularized corneas, and in large corneal
defects. Therefore, it is possible to hypothesize that, as in
humans, corneal grafting for the treatment of nonvascular,
noninflamed pathologies, could be associated with a lower
rejection rate.
Despite the high rejection rates detected in the present
study, the anatomical and visual outcome was high (86%
and 82%, respectively). These favorable results are cer-
tainly associated with the medical management of rejec-
tion. In humans, reducing the donor antigen tissue load
and suppressing the host immune response have been
shown to prevent corneal graft rejection.32 For this pur-
pose, the recommended medical management for the pre-
vention of corneal rejection consists of topical and
systemic corticosteroids, both pre- and postopera-
tively,32,34 and topical and systemic cyclosporine A
(CsA).32 In the present study, all patients were prescribed
topical corticosteroids (three times daily) and 0.2% CsA
(twice daily) immediately after surgery.
Mycophenolate mofetil has shown similar38 to slightly
better39 effectiveness than CsA, although with fewer side
effects, in preventing acute rejection following high-risk
corneal transplantation.38,40 Mycophenolate mofetil is
commonly used in patients undergoing solid organ trans-
plantation or hematopoietic cell transplantation,41 and in
veterinary medicine, it has been used for acquired myas-
thenia gravis,42 control of renal allograft rejection,43
immune-mediated hemolytic anemia, and Pemphigus vul-
garis.44 The recommended dose range is 10–20 mg/kg per
os every 12 h, but gastrointestinal hemorrhage, anorexia,
and diarrhea have been reported in dogs.43 To reduce the
toxicity levels, a 15 mg/kg daily dose was used, as it has

© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10

 Table 3. Statistical analysis of glaucoma, suture dehiscence, and graft rejection

Glaucoma Suture dehiscence Graft rejection

No Yes OR (95%CI) P-value No Yes OR (95%CI) P-value No Yes OR (95%CI) P-value

Breed Brachicephalic 34 (73.9%) 3 (75%) 35 (77.8%) 2 (40%) 5.25 (0.77; 35.89) 0.091 18 (78.3%) 19 (70.4%) 0.66 (0.18; 2.4) 0.527
Other 12 (26.1%) 1 (25%) 0.94 (0.09; 9.97) 0.962 10 (22.2%) 3 (60%) 5 (21.7%) 8 (29.6%)
Graft Median 7 (3; 13) 11.3 (10; 15) 1.96 (1.08; 3.54) 0.026 7.8 (3; 15) 6 (5; 10) 0.82 (0.56; 1.2) 0.301 7.5 (4; 12) 7.8 (3; 15) 0.94 (0.77; 1.15) 0.568
size (mm) (min, max)
Graft type Lamellar 26 (56.5%) 3 (75%) 2.31 (0.22; 23.89) 0.483 27 (60%) 2 (40%) 0.44 (0.07; 2.93) 0.399 14 (60.9%) 15 (55.6%) 1.24 (0.4; 3.85) 0.705
Full-thickness 20 (43.5%) 1 (25%) 18 (40%) 3 (60%) 9 (39.1%) 12 (44.4%)
Graft origin
Fresh 6 (13%) 1 (25%) 4.17 (0.23; 76.6) 0.337 7 (15.6%) 0 (0%) NC 0.005* 3 (13%) 4 (14.8%) 0.75 (0.14; 4.04) 0.738
homologous
Frozen 25 (54.3%) 1 (25%) 1 23 (51.1%) 3 (60%) 13 (56.5%) 13 (48.1%) 1
homologous
Frozen 15 (32.6%) 2 (50%) 3.33 (0.28; 39.98) 0.342 15 (33.3%) 2 (40%) NC 0.821* 7 (30.4%) 10 (37%) 0.7 (0.2; 2.41) 0.571
heterologous
Suture Vicryl 9-0 30 (65.2%) 1 (25%) 27 (60%) 4 (80%) 16 (69.6%) 15 (55.6%)
material Ethylon 9-0 16 (34.8%) 3 (75%) 5.63 (0.54; 58.58) 0.149 18 (40%) 1 (20%) 0.38 (0.04; 3.63) 0.397 7 (30.4%) 12 (44.4%) 0.55 (0.17; 1.76) 0.311

NC = not calculable; CI = confidence intervals.

*From Bootstrap with n = 1000 samples.

Table 4. Statistical analysis of visual and anatomical outcome

Visual outcome Anatomical outcome

Total Good Poor HR (95%CI) P-value Good Poor HR (95%CI) P-value

© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10

Breed Brachicephalic 37 (74%) 30 (73.2%) 7 (77.8%) 32 (74.4%) 5 (71.4%)
Other 13 (26%) 11 (26.8%) 2 (22.2%) 1.6 (0.30; 8.13) 0.601 11 (25.6%) 2 (28.6%) 2.05 (0.40; 11.35) 0.413
Graft size (mm) Median (min, max) 7.5 (3; 15) 6.8 (3; 12) 10.5 (7; 15) 1.41 (1.07; 1.85) 0.015 7 (3; 12) 10.5 (7.5; 15) 1.53 (1.08; 2.16) 0.018
Graft type Lamellar 29 (58%) 24 (58.5%) 5 (55.6%) 25 (58.1%) 4 (57.1%)
Full-thickness 21 (42%) 17 (41.5%) 4 (44.4%) 0.53 (0.14; 2.02) 0.349 18 (41.9%) 3 (42.9%) 0.50 (0.11; 2.26) 0.364
Graft origin
Fresh homologous 7 (14%) 5 (12.2%) 2 (22.2%) 8.6 (0.78; 95.16) 0.079 5 (11.6%) 2 (28.6%) 8.72 (0.79; 96.52) 0.078
Frozen homologous 26 (52%) 23 (56.1%) 3 (33.3%) 1 24 (55.8%) 2 (28.6%) 1
Frozen heterologous 17 (34%) 13 (31.7%) 4 (44.4%) 15.75 (1.51; 163.54) 0.021 14 (32.6%) 3 (42.9%) 12.88 (1.16; 142.63) 0.037
Suture material Vicryl 9-0 31 (62%) 27 (65.9%) 4 (44.4%) 29 (67.4%) 2 (28.6%)
Ethylon 9-0 19 (38%) 14 (34.1%) 5 (55.6%) 1.55 (0.4; 5.99) 0.527 14 (32.6%) 5 (71.4%) 2.64 (0.50; 13.91) 0.253
Multivariable analysis Graft size 1.70 (1.12; 2.60) 0.013 2.31 (1.15; 4.62) 0.018
Graft origin
Fresh homologous 13.67 (1.00; 187.14) 0.050 29.77 (1.27; 696.71) 0.035
Frozen homologous 1 1
Frozen heterologous 33.41 (2.34; 477.47) 0.010 60.52 (2.18; 1681.25) 0.016

CI = confidence intervals; HR = hazard ratio.

full-thickness corneal grafting in dogs 7
8 lacerda ET AL.
(a)
Figure 5. Different degrees of corneal graft opacification after treatment of corneal perforation in dogs. Note the 3 mm mildly (a), 5 mm
moderately (b) and 5 mm severely opaque (c) corneal grafts.
shown to have a bioavailability of 65%.41 All the patients
medicated with MM were free of adverse reactions.
Apart from the standard medication for prevention of
corneal rejection, every patient in which rejection was
clinically diagnosed (n = 28;56%) received an increased
frequency of topical corticosteroids (six times daily) and
topical 0.2% CsA (four times daily). This medical strategy,
applied at the appropriate time, successfully managed 39%
(11/28) of rejections, preserving vision in the affected
eyes.
The present study agrees with previous reports in
human beings, where poor outcomes (anatomical and
visual) were associated with larger grafts,32,37,45 and their
proximity to the limbus; as the Langerhans cells that
express class II antigens are primarily located in the
peripheral cornea.32–34 Furthermore, and contrary to pre-
vious research, this study identifies a significant difference
between the origin of the graft donor tissue and the pres-
ence of postoperative complications (i.e., corneal vascular-
ization, corneal edema, or opacification of the graft), the
latter being more frequent in frozen heterologous grafts
and less frequent in fresh homologous tissue.46 However,
it is well known that the immunologic recognition in fresh
donor corneal grafts is directed mainly toward corneal
endothelium and epithelium,47 which means that fresh
full-thickness grafts have a greater risk of rejection.
Apart from clinical rejection, other identified complica-
tions included secondary glaucoma (4/50;8%) and wound
dehiscence (6/50;12%), which have been previously
reported in both human and dogs.3,34,48,49 Similar to what
has been previously described in people, the risk of glau-
coma statistically increased when larger grafts were used
associated with the possible formation of pre-iridal
fibrovascular membranes (PIFM) and/or secondary periph-
eral anterior synechiae at the graft host junction.32,48
In the present study, no statistically significant variables
were associated with suture dehiscence. The choice of
suture material was based on the balance between restor-
ing maximal visual acuity vs. tectonic support. In those
cases where size and location of the graft allowed for a
© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
(b) (c)
good recovery of vision, a nonabsorbable suture material
(nylon) was used. When the size of the graft was signifi-
cantly larger, and the main goal of surgery was globe
preservation (good anatomical outcome), an absorbable
suture (polyglactin 910) was used. The recommendation
in cases where the reason for grafting is optical is to use
nylon, a nonreactive suture material, and therefore less
likely to attract inflammatory cells.2,31
The decision to use lamellar3,4,50,51 or full-thickness
grafts49,52,53 was mostly an historical one, as statistically
there was no significant difference between groups. With
the recent updates54,55 in corneal microsurgical techniques
and donor and postoperative recipient tissue responses to
grafting (reduced degree of immediate postoperative
edema), lamellar keratoplasty has become the technique of
choice, regardless of location or size of the corneal lesion.
The success of corneal grafting for the treatment of
full-thickness defects is achieved when the cornea is
repaired and the patient retains vision after treatment.
Although complete transparency to moderate opacification
of the graft was present in 48.2% of the cases (14/29), the
overall results of this study are very promising (86% of
the eyes with good anatomical outcome, of which 95%
were visual at last examination). In recent publications2,25,
there has been mention of possible migration of endothe-
lial cells adjacent to the donor button, allowing for the
frozen graft to become transparent months after surgery.
The same event has been witnessed in this study in some
of the smaller sized grafts, supporting that this endothe-
lium behavior is an encouraging factor for vision recovery.
The use of fresh grafts would assure a more transparent
cornea, as endothelial cells are preserved. However, in
high-risk corneas, the immune response toward the
endothelium and epithelium is more severe; therefore, a
careful decision must be made before choosing fresh tis-
sue. A second procedure, using fresh cornea, was offered
when opacification and size of the graft were significant.
However, most owners were reluctant to proceed as they
were pleased with the normal visual behavior shown by
their dog.
 full-thickness corneal grafting in dogs 9

One of the potential limitations of the present study 18. Rohrbach JM, Wohlrab TM, Sadowski B et al. Biological corneal
was the reduced number of cases. However, they are a replacement-an alternative to keratoplasty and keratoprosthesis?
A pilot study with heterologous hyaline cartilage in the rabbit
representative sample of what is found in everyday prac-
ur Augenheilkunde 1995; 207: 191–
model. Klinische Monatsbl€atter f€
tice, providing a valid and strong interpretation of the
196.
results. The postoperative treatment regime varied 19. Zirm E. Eine erfolgreiche totale Keratoplastik [translated from
amongst patients, which may justify why similar cases the German by Chris Sandison, Technical Translations from
responded differently. This fact was due to the newer German, Bath].Graefes Archive. Ophthalmology 1906; 64: 580–593.
studies and drugs that become available and that helped 20. Bigger SL. An inquiry into the possibility of transplanting the
control graft rejection more efficiently. cornea, with the view of relieving blindness (hitherto deemed
incurable) caused by several diseases of that structure. The Dublin
In conclusion, corneal grafting is an effective surgical
Journal of Medical Science 1837; 3: 408–417.
treatment for full-thickness corneal defects in dogs. If
21. Holt JR. Corneal Graft in a Dog. Vet. Rec. 1957; 69: 454.
graft rejection is present, additional medical or surgical 22. World Health Organization. Human organ and tissue
therapy may be necessary, achieving a highly satisfactory transplantation. http://www.who.int/ethics/topics/en/human_trans
visual outcome. plant_report.pdf?ua=1 [Accessed on 2003 May].
23. Barraquer JI. Method for cutting lamellar grafts in frozen cornea:
New orientation for refractive surgery. Archivos de la Sociedad
REFERENCES Americana de Oftalmolog
ıa y Optometr
ıa 1958; 1: 271–286.
24. Barraquer JI. Lamellar keratoplasty (special techniques). Annals of
1. Doughman DJ. Treatment of corneal thinning and perforation.
Ophthalmology 1972; 4: 437–469.
Journal of Continuing Education in Ophthalmology 1978; 39: 15.
25. Brooks DE, Plummer CE, Kallberg ME et al. Corneal
2. Gelatt KN, Brooks DE. Surgery of the cornea and sclera. In:
transplantation for inflammatory keratopathies in the horse:
Veterinary Ophthalmic Surgery, 1st edn (ed Gelatt KN). Elsevier
visual outcome in 206 cases (1993-2007). Veterinary
Saunders, New York, 2011; 191–236.
Ophthalmology 2008; 2: 123–133.
3. Hansen PA, Guandalini A. A retrospective study of 30 cases of
26. Bussieres M, Krohne SG, Stiles J. The use of porcine small
frozen lamellar corneal graft in dogs and cats. Veterinary
intestinal submucosa for the repair of full-thickness corneal
Ophthalmology 1999; 4: 233–241.
defects in dogs, cats and horses. Veterinary Ophthalmology 2004; 7:
4. Pe~na Gimenez MT, Farina IM. Lamellar keratoplasty for the
352–359.
treatment of feline corneal sequestrum. Veterinary Ophthalmology
27. Goulle F. Use of porcine small intestinal submucosa for corneal
1998; 1: 163–166.
reconstruction in dogs and cats: 106 cases. Journal of Small
5. Laguna F, Leiva M, Costa D et al. Corneal grafting for the
Animal Practice 2012; 53: 34–43.
treatment of feline corneal sequestrum: a retrospective study of
28. Scagliotti RH. Tarsoconjunctival island graft for the treatment of
18 eyes (13 cats). Veterinary Ophthalmology 2015; 18: 291–296.
deep corneal ulcers, desmetoceles, and perforations in 35 dogs
6. von Hippel A. Eine neue Methode der Hornhauttransplantation.
and 6 cats. Seminars in Veterinary Medicine and Surgery (Small
ur Ophthalmologie 1888; 34: 108–130.
Graefes Archiv f€
Animal) 1988; 1: 69–76.
7. Vanathi M, Sharma N, Titiyal JS et al. Tectonic grafts for
29. Lassaline ME, Brooks DE, Ollivier FJ et al. Equine amniotic
corneal thinning and perforations. Cornea 2002; 8: 792–797.
membrane transplantation for corneal ulceration and
8. Elshing A, Vorisek EA. Keratoplasty. Archives of Ophthalmology
keratomalacia in three horses. Veterinary Ophthalmology 2005; 5:
1930; 4: 165–173.
311–317.
9. Nussbaum N. Cornea artificialis, ein substitu f€ ur die transplatatio
30. Wilkie DA, Whittaker CJG. Surgery of the cornea. The
cornea. Deutsche Klinik 1853; 34: 3667.
Veterinary Clinics of North America: Small Animal Practice 1997;
10. Refojo MF, Dohlman CH, Ahmad B et al. Evaluation of adhesives
27: 1067–1107.
for corneal surgery. Archives of Ophthalmology 1968; 80: 645–656.
31. Coster DJ. Evaluation of corneal transplantation. British Journal
11. Stoiber J, Fernandez V, Kaminski S et al. Biological response to
of Ophthalmology 1997; 81: 618–619.
a supraDescemetic synthetic cornea in rabbits. Archives of
32. Panda A, Vanathi M, Kumar A et al. Corneal graft rejection.
Ophthalmology 2004; 12: 1850–1855.
Survey of Ophthalmology 2007; 52: 375–396.
12. Li F, Griffith M, Li Z et al. Recruitment of multiple cell lines by
33. Boisjoly HM, Tourigny R, Bazin R et al. Risk factors of corneal
collagen-synthetic copolymer matrices in corneal regeneration.
graft failure. Ophthalmology 1993; 100: 1728–1735.
Biomaterials 2005; 16: 3093–3104.
34. Wilson SE, Kaufman HE. Graft failure after penetrating
13. Parshall CJ. Lamellar corneo-scleral transposition. Journal
keratoplasty. Survey of Ophthalmology 1990; 34: 325–356.
American Animal Hospital Association 1973; 9: 270.
35. V€olker-Dieben HJ, Kok-van Alphen CC, Lansbergen Q et al.
14. Hakanson NE, Meredith RE. Conjunctival pedicle grafting in
Different influences of corneal graft survival in 539 transplants.
the treatment of corneal ulcers in the dog and cat. Journal
Acta Ophthalmologica (Copenhagen) 1982; 60: 190–202.
American Animal Hospital Association 1987; 23: 641.
36. Niederkorn JY. The immune privilege of corneal allografts.
15. Barros PS, Garcia JA, Laus JL et al. The use of xenologous amniotic
Transplantation 1999; 67: 1503–1508.
membrane to repair canine corneal perforation created by
37. Vail A, Gore SM, Bradley BA et al. Conclusions of the corneal
penetrating keratectomy. Veterinary Ophthalmology 1998; 1: 119–123.
transplant follow up study. British Journal of Ophthalmology 1997;
16. Andrade A, Laus J, Figueiredo F. The use of preserved renal
81: 631–636.
capsule to repair lamellar corneal lesions in normal dogs.
38. Reis A, Reinhard T, Voiculescu A et al. Mycophenolate mofetil
Veterinary Ophthalmology 1999; 2: 79–82.
versus cyclosporin A in high risk keratoplasty patients: a
17. Barros PS, Safatle AM. Congenital scleral staphyloma in a dog
prospectively randomised clinical trial. British Journal of
repaired with preserved homologous peritoneum. Veterinary
Ophthalmology 1999; 11: 1268–1271.
Ophthalmology 2000; 3: 27–29.

© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10

10 lacerda ET AL.
39. Maier P, B€ ohringer D, Reinhard T. Clear graft survival and
immune reactions following emergency keratoplasty. Albrecht von
Graæes Archiv f€ur Ophthalmologie 2007; 3: 351–359.
40. Reis A, Reinhard T, Sundmacher R et al. Effect of
mycophenolate mofetil, cyclosporin A, and both in combination
in a murine corneal graft rejection model. British Journal of
Ophthalmology 1998; 6: 700–703.
41. Lupu M, McCune J, Kuhr C et al. Pharmacokinetics of oral
mycophenolate mofetil in dog: bioavailability studies and the
impact of antibiotic therapy. Biology of Blood and Marrow
Transplantation 2006; 12: 1352–1354.
42. Dewey CW, Cerda-Gonzalez S, Fletcher DJ et al.
Mycophenolate mofetil treatment in dogs with serologically
diagnosed acquired myasthenia gravis: 27 cases (1999–2008).
Journal of American Medical Association 2010; 6: 664–668.
43. Gregory CR. Immunosuppressive Agents. In: Kirk’s Current
Veterinary Therapy XIV, 14th edn (ed Bonagura JD) Saunders: St.
Louis, 2009; 254–259.
44. Whitley NT, Day MJ. Immunomodulatory drugs and their
application to the management of canine immune-mediated
disease. Journal of Small Animal Practice 2011; 2: 70–85.
45. Hara H, Cooper DK. The immunology of corneal
xenotransplantation: a review of the literature. Xenotransplantation
2010; 5: 338–349.
46. Hacker DV. Frozen corneal grafts in dogs and cats: a report on
19 cases. Journal of the American Animal Hospital Association 1991;
27: 387–398.
© 2016 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
47. Boruchoff SA. Penetrating keratoplasty. In: Principles and Practice
of Ophthalmology (eds Albert DM, Jakobiec KA) W B Saunders:
Philadelphia, 1994; 325–337.
48. Ziaei M, Sharif-Paghaleh E, Manzouri B. Pharmacotherapy of
corneal transplantation. Expert Opinion on Pharmacotherapy 2012;
6: 829–840.
49. Henriksen Mde L, Plummer CE, Mangan B et al. Visual
outcome after corneal transplantation for corneal perforation and
iris prolapse in 37 horses: 1998-2010. Equine Veterinary Journal.
Supplement 2012; 43: 115–119.
50. Paufique L. Indications for the therapeutic lamellar corneal graft.
American Journal of Ophthalmology 1950; 33: 24–25.
51. Paufique L, Philps S. Acute eye lesions treated by lamellar corneal
grafting. British Journal of Ophthalmology 1950; 34: 746–748.
52. Townsend WM, Rankin AJ, Stiles J. Heterologous penetrating
keratoplasty for treatment of a corneal sequestrum in a cat.
Veterinary Ophthalmology 2008; 4: 273–278.
53. Jonas J, Rainer MR, Budde W. Tectonic sclerokeratoplasty and
tectonic penetrating keratoplasty as treatment for perforated or
predescemetal corneal ulcers. American Journal of Ophthalmology
2001; 1: 14–18.
54. Brooks DE. Targeted lamellar keratoplasty in the horse: A
paradigm shift in equine corneal transplantation. Equine
Veterinary Journal. Supplement 2010; 37: 24–30.
55. Bessant D, Dart J. Lamellar keratoplasty in the management of
inflammatory corneal ulceration and perforation. Eye 1994; 8:
22–28.