Accelerated collagen cross-linking was used to treat melting keratitis in 8 dogs. The following results were observed: 1) Pain improvement was seen in all cases within 3 days after treatment. 2) Epithelial healing was observed within 15 days for all cases. 3) Disappearance of cellular infiltration was observed in all cases by day 7. 4) Corneal vascularization disappeared in 4 cases and was reduced in 4 cases by 1 month. 5) At 1 month, all eyes had visual function despite variable corneal scarring. No recurrent infections occurred. The study suggests that accelerated collagen cross-linking is a valuable treatment for melting keratitis in dogs.

1. Evaluation of accelerated collagen cross-linking for the treatment
of melting keratitis in eight dogs
Frank Famose
Service d’Ophtalmologie, Clinique Veterinaire des Acacias, 42 avenue Lucien Servanty, 31700 Blagnac, France
Address communications to:
F. Famose
Tel.: +33 5 61 71 24 02
Fax: +33 5 61 71 65 52
e-mail: frankfamose@gmail.com
Abstract
Objective Melting keratitis is serious condition presenting a high risk of permanent
blindness and is caused by infectious or noninfectious factors. In humans, the clinical
efficacy of collagen cross-linking (CXL) has been described in the treatment of refrac-
tory infectious keratitis by arresting keratomalacia. The aim of this study was to evalu-
ate the efficacy of accelerated CXL for the treatment of melting keratitis in dogs.
Animal studied and Procedure Eight dogs were treated for unilateral melting keratitis by
accelerated CXL. Corneas were irradiated by UVA (370 nm) at 30 mW/cm² irradi-
ance for 3 min after soaking by 0.1% riboflavin in 20% dextran for 30 min. Follow-
up was conducted 3, 7, 14, and 30 days after treatment.
Results Pain improvement was observed for all cases within 3 days after treatment.
Epithelial healing was observed within 15 days for all cases.
Disappearance of cellular infiltration was observed for all cases at day 7. The corneal
vascularization disappeared for 4 of 8 dogs and was reduced for 4 of 8 dogs within
1 month. At 1 month, all cases presented a variable degree of corneal scarring, but all
eyes had visual function. No recurrent infection was observed.
Conclusions The main observation of this study is that all the cases have presented with
the same clinical result regardless of the presence and the sensitivity of the infectious
agents and regardless of the duration of the condition prior to CXL treatment. Accel-
erated CXL appears to be a valuable option for the treatment of melting keratitis.
Key Words: accelerated CXL, cross-linking, dogs, melting keratitis, optical coherence
tomography, riboflavin
INTRODUCTION
In dogs, melting keratitis is serious condition presenting
a high risk of permanent blindness.1–3
In melting kerati-
tis, stromal damage is initiated by various mechanisms
including bacterial proliferation, toxin secretion, micro-
bial or corneal proteases activation leading to keratomala-
cia or corneal melting. An imbalance between the
endogenous and exogenous matrix metalloproteinases
(MMP) and the proteinases present in the cornea and the
precorneal tear film leads to the destruction of the colla-
gen of the cornea.4–7
Microbial infection is usually sus-
pected to drive the inflammatory state responsible for
corneal melting but cannot always be demonstrated.1
There are few corneal pathogens that are associated with
primary corneal infections,8,9
and infectious corneal melt-
ing is typically due to secondary bacterial or fungal infec-
tions.1–3
Risk factors for the development of secondary
infectious keratitis include ocular trauma and corneal
exposition because of lagophthalmos, particularly in
brachycephalic breeds, with reduced corneal sensitivity,
and ocular surface conditions such as keratoconjunctivitis
sicca or other quantitative or qualitative tear film defi-
ciencies. Melting can also occur in the absence of infec-
tion, and the imbalance between proteolytic enzymes and
protease inhibitors is thought to be due to the produc-
tion of collagenolytic agents by resident and inflamma-
tory cells.4–6
Medical treatment is based on the administration of
topical antibiotics either commercially available prepara-
tions or reinforced compounded preparations10
and on the
topical use of protease inhibitors.7,11
Failure of the treat-
ment can lead to loss of vision because of the evolution of
keratomalacia and corneal perforation. These failures are
usually managed by tectonic surgery. Traditional surgical
procedures are conjunctival grafts, biomaterial grafts, or
amniotic membrane transplantation.12–15
These techniques
lead to variable degrees of corneal opacity.
© 2013 American College of Veterinary Ophthalmologists
Veterinary Ophthalmology (2013) 1–10 DOI:10.1111/vop.12085

2. Collagen cross-linking (CXL) is a technique which cre-
ates intrafibrillar covalent bonds in the collagen fibers of
the corneal stroma via the photo-activation of riboflavin
by ultraviolet-A (UVA) light. It has been used since 1998
in human ophthalmology for the treatment of ectatic
conditions of the cornea, such as progressive keratoconus,
pellucid marginal degeneration, and ectatic complications
of refractive surgeries.16–19
For these indications, safety
and efficacy of this procedure have been widely estab-
lished.16–25
The antimicrobial activity of CXL against
numerous bacteria and fungi26
has been demonstrated
under experimental conditions. Efficacy against fungal
species is very variable: no effect on Candida and Fusari-
um cultures;26,27
minimal/moderate effect on Fusarium in
vivo in rabbits;28
and an amphotericin B pretreatment–
dependent effect on Candida albicans, Fusarium spp, and
Aspergillus fumigatus.29
An increased collagen resistance
against enzymatic digestion30
has been demonstrated
under experimental conditions in vitro. More recently,
the clinical efficacy of CXL has been described in the
treatment of presumed infectious keratitis and corneal
melting.31–38
Because of the promising results obtained in human
ophthalmology and the recently described application of
CXL in dogs and cats in a pilot study,39
the aim of this
study was to evaluate the efficacy of accelerated CXL for
the treatment of melting keratitis in dogs.
MATERIALS AND METHODS
Inclusion criteria
For this prospective uncontrolled study, case recruitment
was based on a clinical diagnosis of a present or impend-
ing melting keratitis characterized by epithelial and ante-
rior stromal loss, cellular infiltration, corneal
vascularization, and loss of corneal integrity. The cases
were recruited after referral by the initial veterinarian after
worsening of ocular signs despite the initial medical treat-
ment. Cases with impending or confirmed corneal perfo-
ration were excluded from this study.
All animals were included after obtaining their owner’s
consent. All procedures were performed in accordance
with the French guidelines for animal care.
Ophthalmologic examination
Keratitis evaluation was made under the following criteria:
The specific signs of keratitis were evaluated by slit-
lamp examination (HawkeyeTM
, Dioptrix, Toulouse,
France). A clinical score modified from Tajima et al.40
(0–3; 0 = absent, 1 = mild, 2 = moderate, 3 = severe) was
used to grade the severity of eye symptoms and signs,
which included mucopurulent discharge, corneal edema,
corneal vascularization, conjunctivitis, blepharitis, and uve-
itis. The highest possible total score was 18. Cellular infil-
tration was noted as ‘present’ or ‘absent’ and was a
component of the diagnosis of the corneal melting, includ-
ing corneal edema and stromal dissolution.1,39
A pain
score modified from the Hackett and McDonald scoring41
(0–1; 0 = absent, 1 = present) was used to grade pain signs
that included prostration, aggressive behavior, blepharo-
spasm, enophthalmos, photophobia, ocular pruritus, and
defense reaction to examination. The highest possible total
score was 7.
Tear production was evaluated by Schirmer test I with
the absorbent strip of paper (Test de Schirmer, Virbac,
Carros, France) placed in the temporal third of the infe-
rior conjunctival fornix and reading after 1 min (normal
values 10–15 mm). A sample for bacteriological diagnosis
(culture and antibiotic sensitivity) was performed on each
animal by corneal scraping with a Kimura spatula (Moria
Surgical, Antony, France) and submitted for bacteriologic
analysis at a local laboratory dedicated to veterinary bacte-
riology (Laboratoire Meynaud, Toulouse, France).
Measurements of the corneal thickness and the ulceration
size
Dogs were anesthetized with 300 lg/m² medetomidine
(DomitorTM
, Pfizer, New York, USA) and 5 mg/kg keta-
mine (Imalgene 1000TM
, Merial, Lyon, France) delivered
intravenously followed by 0.5–2% isoflurane (Isoflurane
BelamontTM
, Nicholas Piramal Ltd., London, UK) in
oxygen after endotracheal intubation.
Dogs were maintained in dorsal recumbency with a vac-
uum cushion. Pachymetry was performed with an Optical
Coherence Tomography (OCT) device (iVueTM
, Optovue,
Fremont, CA, USA) as described elsewhere42
. Measure-
ments were performed with the focus at the center of the
corneal lesion. Minimal and maximal corneal thicknesses
were evaluated with the caliper tool on the OCT pictures
obtained (Fig. 1).
Ulceration size was measured with a manual caliper. As
the corneal ulcers have sometimes an elliptic or irregular
Figure 1. OCT pachymetry. Corneal thickness is measured at its
maximum and its minimum. Case no 3.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
2 f a m o s e

3. shape, the greater axis length was noted as the ulcer size
(in mm). Because it can interfere with UVA absorption,43
fluorescein staining was not used for initial measurements
of epithelial defects.
Treatment by accelerated cross-linking
The eyelids were kept open with the use of a lid specu-
lum. After the instillation of a topical anesthetic (oxybu-
procaine 0.4%, Cebesine 0.4%TM
, Bausch Lomb-
Chauvin, Montpellier, France), the ulcer margins were
cleaned and the debris removed from the corneal surface
with a microsurgery sponge. A solution of isotonic ribofla-
vin (riboflavin 0,1%, dextran 20%, VibexTM
, Avedro, Wal-
tham MA, USA) was instilled on the cornea for 30 min
(one drop every 2 minutes), and the corneal surface was
rinsed with balanced salt solution (BSS pour irrigation oc-
ulaireTM
, Alcon, Fort Worth, TX, USA) at the end of
riboflavin instillation. Penetration of riboflavin through
the cornea was confirmed by visualizing the fluorescence
of the riboflavin in the anterior chamber with slit-lamp bi-
omicroscopy using the cobalt blue light.
Corneas were irradiated with UVA (wave length =
370 nm) for 3 min with an irradiance of 30 mW/cm²
(5.4 J/cm² per dose) using the KXLTM
System (Avedro,
Waltham MA, USA). The beam of 9 mm in diameter
was centered and focused on the center of the corneal
lesion. Care was taken not to irradiate the corneal lim-
bus. One drop of riboflavin was instilled after 2 min of
irradiation. All animals have received a single CXL treat-
ment.
Postoperative treatment
With patient’s safety in mind, each eye was treated twice
daily with one drop of a tobramycin solution (Tobrex
0.3%TM
, Alcon, Rueil-Malmaison, France) after the CXL
treatment until complete epithelial healing. All anticolla-
genase treatments have been stopped.
Follow-up
The aim of the study was to evaluate corneal healing and
symptom reduction over a 30-day follow-up period. Eyes
were examined 3, 7, 14, and 30 days after treatment using
the same pretreatment protocol minus the Schirmer test.
Epithelial integrity was evaluated after instillation of a
drop of fluorescein solution (fluorescein 0.5% collyre uni-
dose TVM, Laboratoires TVM, Lempdes, France) and
rinsing (Ocryl nettoyant oculaire, Laboratoires TVM,
Lempdes, France) before cobalt-blue-light illumination
(Hawkeye, Dioptrix, France). Fluorescein dye staining of
the cornea was interpreted as a positive result. In cases of
epithelial healing, topical treatment was stopped. In cases
of epithelial defect, topical antibiotic treatment (Tobrex
0.3%TM
two drops/day) was maintained until the next
examination.
Photographs were taken for each case at each follow-up
appointment.
RESULTS
Preoperative features
Eight dogs were treated between April and December
2012 at a single site (Clinique veterinaire des Acacias,
Toulouse-Blagnac, France) by the same clinician (Frank
Famose). Table 1 presents individual preoperative fea-
tures. Five of eight dogs were from brachycephalic breeds,
two were Cavalier King Charles Spaniels with macro-
palpebral fissures, and the last one was a Great Dane.
Keratitis duration before CXL treatment ranged from
7 days to 1 month. None of the cases had a bacteriologic
analysis before initiation of the first treatment, and 5 of 8
received a surgical treatment (one epithelial debridement
and four nictitant membrane flaps).
Among eight samples submitted for bacteriologic analy-
sis, three were positive. Pseudomonas aeruginosa was isolated
from case no. 2 and case no. °4 and was resistant to am-
inosides (gentamicin and tobramycin) and sensitive to qui-
nolones (marbofloxacin and enrofloxacin). In case no. 7, a
Pseudomonas aeruginosa was isolated and was sensitive to
aminosides (gentamicin, tobramycin, kanamycin, framyce-
tin) and to quinolones (marbofloxacin, enrofloxacin).
Maximal corneal thickness ranged from 543 to
1370 lm, and minimal corneal thickness ranged from 292
to 810 lm. The relative depth of the ulceration ranged
from 29 to 56% of the corneal thickness. The size of
epithelial defect ranged from 2 to 8 mm in maximal
diameter.
Postoperative features
Individual scores are summarized on the Table 2.
Epithelial healing was observed at the 7th day post-
treatment for 6 of 8 cases and at 15 days for the two last
cases (Fig. 2).
Mean clinical and pain score showed a marked decrease
from the 3rd to the 7th day after treatment (Fig. 2). Cel-
lular infiltration of the cornea was present for all cases at
the time of treatment and had disappeared 7 days after
CXL treatment (Fig. 3).
Corneal vascularization was present in all cases from
day 8, even in the cases in which it was not observed at
D1. At 30 days post-treatment, it had disappeared in 4
of 8 cases and was mild in 4 of 8 cases (Fig. 4). At
1 month, all cases presented a variable degree of corneal
fibrosis ranging from mild (Fig. 5) to marked (Figs 6 and
7) and melanosis (Figs 5 and 7), but all eyes had visual
function.
No recurrent infection was observed during the study.
DISCUSSION
We have used accelerated collagen cross-linking for the
treatment of melting keratitis in eight cases. All treated
animals showed reduced pain 3 days after treatment. This
observation is similar to results in humans in which
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
t r e a t m e n t o f m e l t i n g k e r a t i t i s 3

4. ocular pain improvement was achieved over the same
time duration.31–33
Complete epithelial healing was achieved in 7 days for
6 of 8 dogs and within 14 days after treatment for 2 of 8.
In the description of Spiess et al., epithelial healing was
achieved in 13 days for one dog and 30 and 40 days for
the two others. The difference in healing rates between
these two studies could be explained by the difference in
the epithelial removal before riboflavin application. Epi-
thelium was removed on a surface of 9–11 mm in diame-
ter in Spiess et al. study and was limited to the ulcer
margins in the present study as described by Rosetta
et al.34
Resolution of corneal melting was observed in all trea-
ted eyes within 7 days after treatment. This effect was
observed by biomicroscopic examination as a reduction in
corneal thickness and cellular infiltration with recovery of
corneal transparency. These results are similar to those in
the literature31–38,44
and could be attributable to direct
effects of CXL. Clinical observations and experimental
data have shown that CXL may have three distinct effects
on cornea:
(1) A bactericidal effect because of free radical libera-
tion by photo-activation of riboflavin. This effect is
manifested by bacterial DNA and membrane altera-
tions.45
Although this effect has been well estab-
lished in experimental conditions, immediate
reduction in the bacterial load has not been demon-
strated in the different clinical reports.31–39
(2) An increase in the corneal resistance because of
mechanical and biochemical modifications of cor-
neal structure46,47
and the creation of intralamellar
covalent collagen bonds.48
This mechanism is
thought to be limited to the first anterior 200 lm
of the treated cornea49
and to contribute to
increased resistance to proteases.30
Experimentally,
an increase in mechanical rigidity and in resistance
to proteolytic enzymes has been shown for human
and swine corneas.30,50–52
Similar data are not avail-
able for dogs, but as long as architecture and colla-
gen structure of canine and human corneas are very
similar, it can be expected that the corneal effects of
CXL should be similar.
(3) A reduction in the corneal inflammatory response by
induction of apoptosis of the keratocytes located in
the anterior part of the stroma.53,54
This may con-
tribute to modification of the local immune response
mediated by Langerhans and dendritic cells and may
reduce corneal melting and vascularization as suggested
by Makdoumi et al.44
In our cases, we observed a dra-
matic reduction in cellular infiltration and in corneal
melting. Corneal vascularization increased in the first
7 days and regressed over the next 3 weeks but
persisted in a reduced state in 4 of 8 eyes.
As far as we know, limited data have been published about
the corneal effects of CXL in dogs under clinical conditions.
The first publication is very recent39
and presents an adap-
Table 1. Individual preoperative features
Case number Case no. 1 Case no. 2 Case no. 3 Case no. 4 Case no. 5 Case no. 6 Case no. 7 Case no. 8
Breed Cavalier
King Charles
Carlin Shih Tzu Shih Tzu Pekingese Cavalier King
Charles
French
Bulldog
Great Dane
Age 1 year 2 year 8 year 3 year 10 year 6 year 8 year 1 year
Affected eye OS OS OD OS OD OS OD OD
Duration prior to
CXL (from first
symptoms)
3 weeks 1 week 1 month 2 weeks 2 weeks 2 weeks 2 weeks 1 week
Initial bacteriology NP* NP NP NP NP NP NP NP
Previous medical
treatment (local)
Framycetin
Ciprofloxacin
NAC†
Gentamicin
14 mg/mL
NAC
Framycetin
Na-
Hyaluronate
Ciprofloxacin
Atropine
Gentamicin
14 mg/mL
NAC
Framycetin
Tobramycin
NAC
Tobramycin
NAC
Ciprofloxacin
Previous surgical
treatment
Epithelial
scraping
None 3rd eyelid
flap
3rd eyelid flap None 3rd eyelid flap 3rd eyelid
flap
None
Schirmer test
(mm/min)
12 14 12 8 17 12 12 13
Presurgical
bacteriology
Sterile Pseudomonas
aeruginosa
Sterile Pseudomonas
aeruginosa
Sterile Sterile Pseudomonas
aeruginosa
Sterile
Min. CT‡
(lm) 557 580 475 570 292 606 340 810
Max. CT (lm) 1002 1300 703 800 660 980 543 1370
Ulcer depth (%
corneal
Thickness)
44% 55% 32% 29% 56% 38% 37% 41%
Ulcer size
(mm)
3 8 8 3 2 4 8 2
*NP: Not performed.
†
NAC: N-acetylcysteine.
‡
CT: Corneal thickness.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
4 f a m o s e

5. tation of the treatment protocol used for human keratoc-
onus described by Wollensack16
with the use of a com-
pounded riboflavin solution. Accelerated cross-linking (with
the KXLTM
) is a recent adaptation of this traditional tech-
nique based on the use of a higher irradiance UV light
source. The energy delivered by both protocols (30 mW/
cm² for 3 min for the accelerated protocol or 3 mW/cm²
for 30 min for the traditional one) is the same (5.4 J/cm²)
achieving the same biological effects.55,56
However, in
experimental conditions, the biochemical stiffness of the
cornea seems to decrease in higher influences due to rapid
oxygen depletion, because CXL is an oxygen-dependant
process.57
The influence on the treatment of melting kerati-
tis is unknown. The use of accelerated cross-linking reduces
operating time and thus the duration of anesthesia. In
human patients, no additional adverse effects have been
observed with irradiances 3 mW/cm².55,56
The isotonic
riboflavin solution used in the present study (VibexTM
) is
commercially available in Europe. Its concentration is the
same as the compounded solution used by Spiess et al. and
is ready to use. The absorption spectra of fluorescein and
riboflavin for UV-A are very close. The presence of fluores-
cein in the anterior stroma limits UV-A absorption and
could explain some differences observed in healing rates in
human patients43
by reducing the photochemical reaction
of UV-A and riboflavin. In the present study, to avoid the
risk of interference, fluorescein staining was used only in
the follow-up period.
Case inclusion was based on clinical diagnosis of melt-
ing keratitis by the observation of epithelial loss, corneal
edema, cellular infiltration, and stromal dissolution. As
stated earlier, microbial infection is usually suspected to
drive the inflammatory state responsible for corneal melt-
ing by the contamination of a previous epithelial defect.3
However, a microbial infection cannot always be identi-
fied. Although bacterial contamination was suspected in all
cases, only 3 of 8 cases had a positive culture in our study,
Table 2. Summary of individual scores
Case
number Score D1 D4 D8 D15 D31
Case
no. 1
Clinical score 10 6 4 1 0
Pain score 5 2 0 0 0
Epithelial defect + + À À À
Cellular infiltration + + À À À
Case
no. 2
Clinical score 15 8 4 2 1
Pain score 6 3 0 0 0
Epithelial defect + + À À À
Cellular infiltration + + À À À
Case
no. 3
Clinical score 10 7 3 1 0
Pain score 4 3 0 0 0
Epithelial defect + + + À À
Cellular infiltration + À À À À
Case
no. 4
Clinical score 7 5 2 1 0
Pain score 4 3 1 0 0
Epithelial defect + + À À À
Cellular infiltration + À À À À
Case
no. 5
Clinical score 9 5 2 1 1
Pain score 6 5 1 0 0
Epithelial defect + + À À À
Cellular infiltration + + À À À
Case
no. 6
Clinical score 11 7 2 1 1
Pain score 4 3 0 0 0
Epithelial defect + + À À À
Cellular infiltration + + À À À
Case
no. 7
Clinical score 11 8 3 1 1
Pain score 6 5 1 0 0
Epithelial defect + + + À À
Cellular infiltration + À À À À
Case
no. 8
Clinical score 10 6 1 1 0
Pain score 6 2 0 0 0
Epithelial defect + + À À À
Cellular infiltration + + À À À
Mean clinical score 10.375 6.5 2.625 1.125 0.5
Mean pain score 5.125 3.25 0.375 0 0
With a cellular infiltration
(nb/8)
8 5 0 0 0
With an epithelial defect
(nb/8)
8 8 2 0 0
Figure 2. Evolution of the mean clinical score
(blue), of the mean pain score (red), and of the
number of dogs with epithelial defect (green)
during the 30-day follow-up period.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
t r e a t m e n t o f m e l t i n g k e r a t i t i s 5

6. which correlates with literature data.58
This fact can be
explained by the inhibition of bacteria by the medical
treatments prescribed before the presentation and also by
the hypothesis of a sterile keratitis due to the activation of
the proteases caused by mechanisms other than bacterial
proliferation.1,3,4
This observation joins the results pre-
sented by Spiess et al.39
in which none of the six cases
(three dogs and three cats) had a positive bacteriologic
culture. In the present study, the failure of the antibiotic
treatments performed before referral could be explained
by the presence of resistant bacteria, but bacteriologic
analysis isolated them only in one case. Other causes, such
as the advanced stage of the disease when presented to the
veterinarian, poor patient, and/or owner compliance could
influence the course of the disease.
The use of tobramycin in the postoperative phase may
be a bias for case no. 7 for which the germ identified was
sensitive. In this case, the postoperative evolution was
exactly the same as the other cases, and the use of tobra-
mycin did not show any added value in comparison.
The main observation of this study is that all the cases
have presented with the same outcome regardless of the
presence and the sensitivity of the isolated bacteria and
regardless of the duration of the condition prior to CXL
treatment. Similar observations have been reported by
Makdoumi in a series of 25 human cases.44
His conclusion
is that CXL should be considered as the initial treatment
for keratitis without the use of antibiotics, arguing that
this could be a way of reducing the risk of antibiotic resis-
tance. In the present study, topical antibiotic therapy was
maintained until complete epithelial healing, for the pre-
vention of a secondary bacterial contamination, and
according to the lack of data about efficacy of CXL treat-
ment in dogs. The next step of the definitive evaluation of
the CXL treatment would be the follow-up without any
postoperative treatment.
Although justified on a scientific basis to compare CXL
with a traditional treatment, no control group was
included in our series. Animals were presented after a pre-
vious medical treatment (topical antibiotics and, in some
cases, antiproteases) with no improvement in or worsening
of the clinical signs at the time of referral. Treatment with
antiproteases could have yielded similar results as
described in this study, and such therapy would have been
a possible alternative. Because no control group was
included, a comparison of treatment effects was not possi-
ble. However, all antiprotease treatments were stopped at
the time of CXL, and the arrest of corneal melting is
therefore probably attributable to the CXL procedure.
In human patients, three kinds of adverse effects of
CXL have been described: postoperative infection, herpes-
virus exacerbation, and corneal endothelial lesions. In the
treatment of human keratoconus by CXL, three cases of
postoperative infections have been described in the days
or weeks following the procedure.59–61
In all cases, stromal
infection was present 3 to 5 days after the procedure and
was attributed to the surgical removal of corneal epithe-
(a)
(b)
Figure 4. Case no. 4 Evolution of vascularization from pretreatment
presentation (a) to 1 month post-treatment presentation (b). Residual
vascularization is scored ‘mild’.
(a)
(b)
Figure 3. Case no. 1 Initial presentation (a) and 7 days after the
treatment (b). Cellular infiltration and corneal edema are reduced,
and transparency of the corneal peripheral to the lesion has greatly
increased. Epithelial healing was complete.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
6 f a m o s e

7. lium. These observations are not transposable to this
study in which epithelial loss was an inclusion criterium,
and previous bacterial infection was suspected. In the pres-
ent study, no epithelial debridement was performed, but
only a careful cleaning of cellular debris before the instil-
lation of riboflavin, as described by Rosetta et al.34
No
postoperative infection was observed during the follow-up
period, as described in the human studies.23–25,34
A case of
herpetic keratitis has been described in a human patient
5 days after his treatment for keratoconus by CXL.62
Viral
reactivation was described as a superficial keratitis and
materialized by dendritic ulceration and as an anterior
uveitis. In dogs, canine herpesvirus has a worldwide distri-
bution and a high-exposure prevalence,63
and ocular
symptoms in adult dogs have been occasionally
described.64,65
The risk of viral activation by CXL treat-
ment is unknown and should be investigated further.
Endothelial lesions may be directly attributable to the
CXL treatment by the cytolytic effect of riboflavin photo-
activation on endothelial cells.66,67
Experimentally, the
maximal absorption depth for UVA in a riboflavin-satu-
rated cornea is thought to be approximately 300 lm for
this protocol. UVA absorption does not stop at 300 lm,
but absorption levels drop below the toxic threshold in
the deeper parts of the cornea and eye as a whole. Stiffen-
ing effects of CXL seem to be limited to the more superfi-
cial 200 lm of the stroma,47,49
and apoptotic effects of the
procedure appear to be rare in the deeper parts of the cor-
nea (deeper than 300 lm).49
This concept, often referred
to as ‘riboflavin shielding’, correlates directly with the
minimal corneal thickness necessary for safe CXL treat-
ment. In this study, corneal thickness was significantly
300 lm in 7 of 8 cases and was 292 lm in case no. 5.
With the owner’s consent, we took the risk to treat such a
thin cornea. No endothelial effects were observed in any
treated cases. Before the treatment, many eyes presented a
thick cellular infiltration, which can interfere with UV
penetration and produce a heterogeneous photo-activation
of riboflavin. No difference in the results was observed
between these different cases regardless of the initial cor-
neal thickness or infiltration. In cases of melting keratitis,
corneal loss is sometimes much more pronounced than
the cases we have treated, and the residual stroma can be
less than 300 lm. In human patients, infected corneas
with a thickness 300 lm have been successfully treated
by CXL after soaking with hypotonic riboflavin34
. Very
thin corneas or corneas with impending perforation have
not been included in this study. An extension of our study
to a larger group should allow us to optimize treatment
parameters according to the size and the depth of the cor-
neal loss.
In the study of Spiess et al.39
two brachycephalic dogs
were treated, and one of these demonstrated the develop-
ment of corneal melanosis during the second month follow-
ing the CXL treatment. In the present study, at 30 days
after the treatment, corneal melanosis was observed in 3 of
5 brachycephalic dogs, but not in the two Cavalier King
Charles Spaniels that presented with macropalpebral fis-
(a)
(b)
Figure 5. Case no. 2. Pretreatment presentation (a) and 1 month
after treatment (b) with persistence of central superficial corneal
vascularization scored as ‘mild’ and central corneal melanosis.
(a)
(b)
Figure 6. Case no. 3. Pretreatment presentation (a) and 1 month
after CXL, with central corneal fibrosis and ‘mild’ vascularization.
No corneal melanosis is visible (b).
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
t r e a t m e n t o f m e l t i n g k e r a t i t i s 7

8. sures nor in the Great Dane with euryblepharon. As stated
by Spiess et al., it is not clear whether the keratitis or the
CXL treatment or both were significant factors in the devel-
opment of melanosis in these three patients. In this series, a
variable degree of corneal fibrosis has been observed in all
cases. Corneal haze has been described in human patients
after CXL treatment for keratoconus.18–22
In this study, it is
not clear whether the corneal fibrosis can be attributable to
the initial keratitis, the treatment procedure, or both. Fol-
low-up for a longer period could be useful to evaluate the
long-term effects of the treatment.
All the dogs treated in this study have presented a com-
plete healing or at least limited sequelae of their keratitis,
regardless of the presence of bacterial agents, the exten-
sion of the initial corneal lesions, the duration of the dis-
ease before treatment, and the previous treatments. The
excellent results achieved in this small series with no
observable adverse effects in the short-term suggest that
CXL could be a valuable therapeutic option for melting
keratitis. Previously published observations were con-
firmed. The CXL procedure needs a precise focalization
of the UV beam and thus needs general anesthesia.
Because the duration of the anesthesia is reduced in com-
parison with the traditional CXL protocol, accelerated
cross-linking could present a practical advantage while
providing the same biological effects. However, acceler-
ated CXL is performed with a commercially available
riboflavin (VibexTM
) whose price is significantly higher
than compounded riboflavin. These disadvantages (anes-
thesia, price) have to be taken into account in the treat-
ment decision. Based on the results of this study, CXL
could also be considered as primary treatment for keratitis
with or without the use of antibiotics. Further studies in
this direction will likely contribute to a better knowledge
of melting keratitis treatment and may also be useful in
comparative ophthalmology.
REFERENCES
1. Gilger BC, Ollivier FJ, Bentley E. Diseases and surgery of the
canine cornea and sclera. In: Veterinary Ophthalmology, 4thedn.
(ed. Gelatt KN) Blackwell Publishing, Ames, 2007; 690–752.
2. Ollivier FJ. Bacterial corneal diseases in dogs and cats. Clinical
Techniques in Small Animal Practice 2003; 18: 193–198.
3. Miller WW. Evaluation and management of corneal ulcerations:
a systematic approach. Clinical Techniques in Small Animal Practice
2001; 16: 51–57.
4. Ollivier FJ, Gilger BC, Barrie KP et al. Proteinases of the cornea
and preocular tear film. Veterinary Ophthalmology 2007; 10: 199–
206.
5. Wang L, Pan Q, Xue Q et al. Evaluation of matrix
metalloproteinase concentration in precorneal tear film from
dogs with Pseudomonas aeruginosa-associated keratitis. American
Journal of Veterinary Research 2008; 69: 1341–1345.
6. Brejchova K, Liskova P, Cejkova J et al. Role of matrix
metalloproteinase in recurrent corneal melting. Experimental Eye
Research 2010; 90: 583–590.
7. Ollivier FJ, Brooks DE, Kallberg ME et al. Evaluation of various
compounds to inhibit activity of matrix metalloproteinases in the
tear film of horses with ulcerative keratitis. American Journal of
Veterinary Research 2003; 63: 1081–1087.
8. Murphy JM, Lavach JD, Severin GA. Survey of conjunctival flora
in dogs with clinical signs of external eye disease. Journal of the
American Veterinary Medical Association 1978; 172: 66–68.
9. Ledbetter EC, Dubovi EJ, Kim SG et al. Experimental primary
ocular canine herpesvirus-1 infection in adult dogs. American
Journal of Veterinary Research 2009; 70: 513–521.
10. Regnier A. Clinical pharmacology and therapeutics part 2:
antimicrobial, antiinflammatory agents and antiglaucoma drugs.
In: Veterinary Ophthalmology. 4th edn. (ed. Gelatt KN) Blackwell
Publishing, Ames, 2007, 288–331.
11. Brooks DE, Ollivier FJ. Matrix metalloproteinase inhibition in
corneal ulceration. Veterinary Clinics of North America: Small
Animal Practice 2004; 34: 611–622.
12. Hollingsworth SR. Corneal surgical techniques. Clinical
Techniques in Small Animal Practice 2003; 16: 161–167.
13. Bussieres M, Krohne SG, Stiles J et al. The use of porcine small
intestinal submucosa for the repair of full-thickness corneal
defects in dogs, cats and horses. Veterinary Ophthalmology 2004; 7:
352–359.
14. Goulle F. Use of porcine small intestinal submucosa for corneal
reconstruction in dogs and cats: 106 cases. Journal of Small
Animal Practice 2012; 53: 34–43.
15. Barachetti L, Giudice C, Mortellaro C. Amniotic membrane
transplantation for the treatment of feline corneal sequestrum: a
pilot study. Veterinary Ophthalmology 2010; 13: 326–330.
16. Wollensack G, Spoerl E, Seiler T. Riboflavin/
Ultraviolet-A-induced collagen crosslinking for the treatment of
Keratoconus. American Journal of Ophthalmology 2003; 135: 620–
627.
(a)
(b)
Figure 7. Case no. 7. Pretreatment presentation (a) and 1 month
after CXL, with central corneal fibrosis, mild vascularization, and
nasal corneal melanosis (b).
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
8 f a m o s e

9. 17. Hafezi F, Kanellopoulous J, Wiltfrang R et al. Corneal collagen
cross-linking with riboflavin and ultraviolet A to treat induced
keratectasia after laser in situ keratomileusis ectasia and
keratoconus. Journal of Cataract and Refractive Surgery 2007; 33:
2035–2040.
18. Kymionis GD, Diakonis VF, Kalyvianaki M et al. One-year
follow-up of corneal confocal microscopy after corneal
cross-linking in patients with post laser in situ keratomileusis and
keratoconus. American Journal of Ophthalmology 2009; 147: 774–778.
19. Kymionis GD, Karavitaki AE, Kounis GA et al. Management of
pellucid marginal corneal degeneration with simultaneous
customized photorefractive keratectomy and collagen
cross-linking. Journal of Cataract and Refractive Surgery 2009; 35:
1298–1301.
20. Caporossi A, Baiocchi S, Mazzotta C et al. Parasurgical therapy
for keratoconus by riboflavin-ultraviolet type A rays induced
cross-linking of corneal collagen: preliminary refractive results in
an Italian study. Journal of Cataract and Refractive Surgery 2006;
32: 837–845.
21. Raiskup-Wolf F, Hoyer A, Spoerl E et al. Collagen crosslinking
with riboflavin and ultraviolet-A light in keratoconus: long-term
results. Journal of Cataract and Refractive Surgery 2008; 34: 796–801.
22. Spoerl E, Mrochen M, Sliney D et al. Safety of UVA-Riboflavin
cross-linking of the cornea. Cornea 2007; 26: 385–389.
23. Dhawan S, Rao K, Natrajan S. Complications of Corneal
Collagen Cross-Linking. Journal of Ophthalmology 2011. DOI:
10.1155/2011/899015. [Epub ahead of print].
24. Koller T, Mrochen M, Seiler T. Complication and failure rates
after corneal crosslinking. Journal of Cataract and Refractive
Surgery 2009; 35: 1358–1362.
25. Spoerl E, Hoyer A, Pillunat LE et al. Corneal cross-linking and
safety issues. The Open Ophthalmology Journal 2011; 5: 14–16.
26. Martins SA, Combs JC, Noguera G, et al. Antimicrobial efficacy
of riboflavin/UVA combination (365 nm) in vitro for bacterial
and fungal isolates: a potential new treatment for infectious
keratitis. Investigative Ophthalmology Visual Science 2008; 49:
3402–3408.
27. Kashiwabushi RT, Carvalho FRS, Khan YA et al. Assessment of
fungal viability after long-wave ultraviolet light irradiation
combined with riboflavin administration. Graefe’s Archives for
Clinical and Experimental Ophthalmology 2013; 251: 521–527.
28. Galperin G, Berra M, Tau J et al. Treatment of fungal keratitis
from Fusarium infection by corneal cross-linking. Cornea 2012;
31: 176–180.
29. Sauer A, Letscher-Bru V, Speeg-Schatz C et al. In vitro efficacy
of antifungal treatment using riboflavin/UV-A (365 nm)
combination and amphotericin B. Investigative Ophthalmology and
Visual Science 2010; 51: 3950–3953.
30. Spoerl E, Wollensack G, Seiler T. Increased resistance of
cross-linked cornea against enzymatic digestion. Current Eye
Research 2004; 29: 35–40.
31. Makdoumi K, Mortensen J, Crafoord S. Infectious keratitis
treated with corneal crosslinking. Cornea 2010; 19: 1353–1358.
32. Moren H, Malmsj€o M, Mortensen J et al. Riboflavin and
ultraviolet a collagen crosslinking for the treatment of keratitis.
Cornea 2010; 29: 102–104.
33. Panda A, Krishna SN, Kumar S. Photo-activated riboflavin
therapy of refractory corneal ulcers. Cornea 2012; 31: 1210–1213.
34. Rosetta P, Vinciguerra R, Romano MR et al. corneal collagen
cross-linking window absorption. Cornea 2013; 32: 550–554.
35. Al-Sabai N, Koppen C, Tassignon MJ. UVA/Riboflavin
crosslinking as treatment for corneal melting. Bulletin de la Societe
Belge d’Ophtalmologie 2010; 315: 13–17.
36. Anwar HM, El-Danasoury AM, Hashem AN. Corneal collagen
crosslinking in the treatment of infectious keratitis. Clinical
Ophthalmology 2011; 5: 1277–1280.
37. Iseli HP, Thiel MA, Hafezi F et al. Ultraviolet A/riboflavin
corneal cross-linking for infectious keratitis associated with
corneal melts. Cornea 2008; 27: 590–594.
38. Price MO, Tenkman LR, Schrier M et al. Photoactivated riboflavin
treatment of infectious keratitis using collagen cross-linking
technology. Journal of Refractive Surgery 2012; 28: 706–713.
39. Spiess BM, Pot SA, Florin M et al. Corneal collagen
cross-linking (CXL) for the treatment of melting keratitis in cats
and dogs: a pilot study. Veterinary Ophthalmology 2013; DOI:
10.1111/vop.12027. [Epub ahead of print].
40. Tajima K, Sinjyo A, Ito T et al. Methicillin-resistant Staphylococcus
aureus keratitis in a dog. Veterinary Ophthalmology 2013; 16: 240–
243.
41. Munger RJ. Veterinary ophthalmology in laboratory animal
studies. Veterinary Ophthalmology 2002; 5: 167–175.
42. Famose F. Assessment of the use of spectral domain optical
coherence tomography (SD-OCT) for evaluation of the healthy
and pathological cornea in dogs and cats. Veterinary
Ophthalmology 2013; DOI: 10.1111/vop.12028. [Epub ahead of
print].
43. Richoz O. UV-A absorption rate of fluorescein and its impact on
the efficacy of collagen cross-linking (CXL) for infectious corneal
melting. Abstract from the 8th International Congress of Corneal
Cross-Linking December 7-8, 2012, Geneva, Switzerland.
44. Makdoumi K, Mortensen J, Sorkhabi O et al. UVA-Riboflavin
photochemical therapy of bacterial keratitis: a pilot study.
Graefe’s Archives of Clinical and Experimental Ophthalmology 2012;
250: 95–102.
45. Ruane PH, Edrich H, Gampp D et al. Photochemical inactivation
of selected viruses and bacteria in platelet concentrates using
riboflavin and light. Transfusion 2004; 44: 877–885.
46. Messmer EM, Meyer P, Herwig MC et al. Morphological and
immunohistochemical changes after corneal cross-linking. Cornea
2013; 32: 111–117.
47. Wollensack G, Spoerl E, Seiler T. Stress-strain measurements of
human and porcine corneas after riboflavin-ultraviolet-A-induced
cross-linking. Journal of Cataract and Refractive Surgery 2003; 29:
1780–1785.
48. McCall AS, Kraft S, Edelhauser HF et al. Mechanisms of corneal
tissue cross-linking in response to treatment with topical
riboflavin and long-wavelength ultraviolet radiation (UVA).
Investigative Ophthalmology Visual Science 2010; 51: 129–138.
49. Kohlhaas M, Spoerl E, Schilde T et al. Biochemical evidence of
the distribution of cross-links in corneas treated with riboflavin
and ultraviolet A light. Journal of Cataract and Refractive Surgery
2006; 32: 279–283.
50. Spoerl E, Huhle M, Kasper M et al. Artificial stiffening of the
cornea by induction of intrastromal cross-links. Ophthalmologe
1997; 94: 902–906.
51. Spoerl E, Huhle M, Seiler T. Induction of cross-links in corneal
tissue. Experimental Eye Research 1998; 66: 97–103.
52. Kontadakis GA, Ginis H, Karyotakis N et al. In vitro effect of
corneal collagen cross-linking on corneal hydration properties
and stiffness. Graefe’s Archives of Clinical and Experimental
Ophthalmology 2013; 251: 543–547.
53. Wollensack G, Iomdina E, Didert DD et al. Wound healing in
the rabbit cornea after corneal collagen cross-linking with
riboflavin and UVA. Cornea 2007; 26: 600–605.
54. Wang F. UVA/Riboflavin-induced apoptosis in mouse cornea.
Ophthalmologica 2008; 222: 369–372.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
t r e a t m e n t o f m e l t i n g k e r a t i t i s 9

10. 55. Kanellopoulos AJ. Long term results of a prospective
randomized bilateral eye comparison trial of higher fluence,
shorter duration ultraviolet A radiation, and riboflavin collagen
cross linking for progressive keratoconus. Clinical Ophthalmology
2012; 6: 97–101.
56. Celik HU, Alag€oz N, Yildirim Y et al. Accelerated corneal
crosslinking concurrent with laser in situ keratomileusis. Journal
of Cataract and Refractive Surgery 2012; 38: 1424–1431.
57. Richoz O. Corneal Biomedical properties at different Corneal
Cross-Linking fluences. Abstract from the 8th International Congress
of Corneal Cross-Linking. December 7-8, 2012, Geneva,
Switzerland.
58. Massa KL, Murphy CJ, Hartmann FA et al. Usefulness of
aerobic microbial culture and cytologic evaluation of corneal
specimens in the diagnosis of infectious ulcerative keratitis in
animals. Journal of the American Veterinary Medical Association
1999; 215: 1671–1674.
59. Sharma N, Maharana P, Singh G et al. Pseudomonas keratitis
after collagen crosslinking for keratoconus: case report and
review of literature. Journal of Cataract and Refractive Surgery
2010; 36: 517–520.
60. Pollhammer M, Cursiefen C. Bacterial keratitis early after
corneal crosslinking with riboflavin and ultraviolet-A. Journal of
Cataract and Refractive Surgery 2009; 35: 588–589.
61. Perez-Santonja JJ, Artola A, Javaloy J et al. Microbial keratitis
after corneal collagen crosslinking. Journal of Cataract and
Refractive Surgery 2009; 35: 1138–1140.
62. Kymionis GD, Portaliou DM, Bouzoukis DI et al. Herpetic
keratitis with iritis after corneal crosslinking with riboflavin and
ultraviolet A for keratoconus. Journal of Cataract and Refractive
Surgery 2007; 33: 1282–1284.
63. Ledbetter EC, Kim SG, Dubovi EJ. Outbreak of ocular disease
associated with naturally-acquired canine herpesvirus-1 infection
in a closed domestic dog colony. Veterinary Ophthalmology 2009;
12: 242–247.
64. Gervais KJ, Pirie CG, Ledbetter EC et al. Acute primary canine
herpesvirus-1 dendritic ulcerative keratitis in an adult dog.
Veterinary Ophthalmology 2012; 15: 133–138.
65. Ledbetter EC, Marfurt CF, Dubielzig RR. Metaherpetic corneal
disease in a dog associated with partial limbic stem cell deficiency
and neurotrophic keratitis. Veterinary Ophthalmology. 2012; DOI:
10.1111/j.1463-5224.2012.01064.x. [Epub ahead of print].
66. Gokhale N. Corneal endothelial damage after collagen cross-
linking treatment. Cornea 2011; 30: 1495–1498.
67. Bagga B, Pahuja S, Murthy S et al. Endothelial failure after
collagen cross-linking with riboflavin an UV-A: case report with
literature review. Cornea 2012; 31: 1197–1200.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–10
10 f a m o s e