Article décrivant l'aspect OCT des principales affections cornéennes du chien et du chat

1. Veterinary Ophthalmology (2013) 1–11 DOI:10.1111/vop.12028
Assessment of the use of spectral domain optical coherence
tomography (SD-OCT) for evaluation of the healthy and
pathological cornea in dogs and cats
Frank Famose
Service d’Ophtalmologie, Clinique vetrinaire des Acacias, Blagnac, France
´e
Address communications to: Abstract
F. Famose
Purpose Morphologic evaluation of the cornea is based on the slit-lamp examination.
Tel.: +33 5 61712402 In human ophthalmology, optical coherence tomography (OCT) has opened a new
Fax: +33 5 61716552
field in the clinical approach to anterior segment disorders and more specifically
e-mail: frankfamose@gmail.com
the cornea. The aim of our study is to describe spectral domain OCT examination
of the cornea in dogs and cats in clinical practice conditions.
Material and methods One hundred eyes were examined from 52 dogs and 41 cats pre-
sented to a private practice referral center with an Optovue iVue SD-OCT device.
Sixteen healthy animals were used as control group, and the others were examined
for various corneal conditions. All animals were examined after sedation or anesthesia.
Results Normal and pathological aspects of canine and feline cornea were described
for various conditions such as corneal ulcers, microbial keratitis, corneal sequestrum,
infiltrations, foreign bodies, corneal dystrophies, and surgical conditions.
Conclusion SD-OCT examination of normal and pathological corneal conditions in
dogs and cats gave an accurate evaluation of each component of the cornea. The
advantage of the technique is the in vivo, real-time evaluation of all corneal layers with
the absence of corneal contact. Constraints included the necessity of sedation for
precise focus and the low quality of images obtained with too pigmented or thickened
corneas.
Key Words: cats, cornea, dogs, keratitis, optical coherence tomography, spectral
domain
segment spectral domain OCT (SD-OCT) has been the
INTRODUCTION
subject of more than 500 scientific articles over the last
Several imaging methods exist for the evaluation of 5 years. The application of SD-OCT for medical and
corneal morphology. The most frequently used method is surgical purposes has been specifically described for the
slit-lamp biomicroscopy, but visualization of deeper cor- structural analysis of tear meniscus, normal and pathologi-
neal structures can be limited by light absorption and scat- cal cornea, and iridocorneal angle, as well as evaluation of
tering through the cornea, particularly in case of corneal the anterior chamber, iris, and lens.5–16 OCT has also
edema. The application of new technologies including been compared with slit-lamp examination and with Sche-
high-frequency ultrasound (ultrasound biomicroscopy impflug imaging for corneal endothelial evaluation.17
[UBM]), confocal microscopy, Scheimpflug imaging, and Spectral domain optical coherence tomography function
optical coherence tomography (OCT) has led to improved is similar to ultrasonography with a few major differences.
resolution and complementary evaluation of corneal Ultrasonography uses ultrasound waves emitted by a probe
conditions.1–3 OCT, developed in 1990, was initially dedi- in contact with the tissue to be studied, whereas SD-OCT
cated to evaluation of the retina in human ophthalmol- uses infrared (IR) light emitted at a distance from the cor-
ogy.4 The use of OCT for the examination of the anterior nea, making contactless image acquisition possible.1 While
segment of the eye, a practice started in the early 2000s, optical transparency is not required for ultrasound, the
has opened a new field in the clinical and experimental use of IR light in OCT requires transparent media. The
approaches to evaluating these structures.1–3 Anterior light passing through different ocular media experiences
© 2013 American College of Veterinary Ophthalmologists

2. 2 famose
interference, which is compared with light reflected on a (14 cases), conjunctival grafts (eight cases), corneal seques-
reference mirror at the same working distance. Scanning tra (five cases), corneal foreign body (three cases), corneal
the reference mirror through a range of distances allows dystrophy (three cases), chronic superficial keratitis (three
generation of an axial image (A-scan). A series of axial sec- cases), Florida keratopathy (two cases), stromal hemor-
tions are combined to produce a composite image, similar rhage (one case), glaucoma with marked corneal edema
to the standard two-dimension (B-scan) image produced (two cases), and bullous keratopathy (one case). All ani-
in ultrasonography. SD-OCT images have an axial resolu- mals were examined after consent was obtained from their
tion of 2–4 lm and a lateral resolution of 20–25 lm. As a owner. All procedures were performed in accordance with
comparison, the axial resolution of 50-MHz ultrasonogra- the French guidelines for animal care.
phy is 50 lm.12 Additional information about technical
features of SD-OCT can be found elsewhere.1,4 Anesthesia
Heavy and costly SD-OCT devices were initially lim- All animals examined were sedated or under general anes-
ited to specialized human ophthalmology centers or thesia. Cats were anesthetized with 0.01 mg/kg medetomi-
research centers. But now, lighter and more affordable dine (Domitorâ, Pfizer, NY, USA) and 5 mg/kg ketamine
models have become available. The use of SD-OCT for (Imalgene 1000â, Merial, Lyon, France) via intramuscular
the morphological analysis of the cornea of dogs and cats administration. Dogs were anesthetized with 300 lg/m²
in healthy and pathological conditions has yet to be medetomidine and 5 mg/kg ketamine delivered intrave-
described. Our study aims to evaluate spectral domain nously followed by 0.5–2% isoflurane (Isoflurane Bela-
OCT for corneal imaging in dogs and cats in clinical montâ, Nicholas Piramal Ltd, London, UK) in oxygen
practice conditions. after endotracheal intubation.
Ophthalmologic examination
MATERIAL AND METHODS
All animals were submitted to a thorough ophthalmologic
Animals examination including slit-lamp examination, Schirmer
One hundred eyes were examined in 93 animals. The test, and tonometry. In cases where bacterial infection was
group studied included 52 dogs and 41 cats examined suspected, samples were submitted for a bacteriologic
between July and November 2011 at the Acacias Veteri- analysis in a local laboratory dedicated to veterinary bacte-
nary Clinic Ophthalmology department. The healthy ani- riology (Laboratoire Meynaud, Toulouse, France).
mal control group included eight dogs and eight cats.
Control animals did not have any ocular lesions and were Imaging
anesthetized for minor operations with no ocular involve- Imaging was performed using an iVue SD-OCT system
ment. The remaining animals (44 dogs and 33 cats) were (Optovue EBC Medical, Paris, France) connected to a
anesthetized prior to an examination or ocular surgery for computer interface and a laptop (Fig. 1). The system func-
the following conditions: chronic superficial corneal ulcers tions at 840 nm, which is in the IR radiation range, and
(13 cases), deep corneal ulcers (10 cases), anterior uveitis can perform 26 000 A-scans per second. The imaging unit
(three cases), bacterial keratitis (10 cases), corneal wound has a working distance of 13 mm and can be positioned to
(a) (b)
Figure 1. The OCT device (a) and the animal in
examination position (b).
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

3. sd-oct evaluation of canine and feline cornea 3
capture A-scans horizontally on a table or vertically, held relatively low reflectivity compared with the pre-corneal
in place by an adjustable support for use in the operating tear film and anterior stroma. The stromal layer is thicker
room. For this study, the image capture unit was posi- and appears heterogeneous with an intermediate reflectiv-
tioned vertically (Fig. 1a). ity. In the deepest layer, Descemet’s membrane and the
The iVue device is designed to examine the retina and endothelium combine as a thin, dense line.
the anterior segment. There are two operating modes for The average corneal thickness was 535 lm in dogs
imaging the anterior segment, both of which require (interval of values 500–620 lm) and 600 lm in cats (inter-
installation of an additional corneal-anterior module (or val of values 540–660 lm). In cats, the heterogeneously
CAM) at the end of the image capture unit. The ‘pachy- organized anterior stroma was distinguishable from the
metry’ mode images eight, 6-mm wide radial sections of posterior stroma where the homogeneous organization of
the cornea, with the focus line positioned in the center of collagen lamellae was visible (Fig. 2b).
the cornea. The ‘angle’ mode images a section along a line In the healthy animals, the average measured epithelial
chosen to correspond with the area or interest. thickness was 55 lm in dogs (interval of values 50–60 lm)
Animals were placed in dorsal recumbency, with the and 60 lm in cats (interval of values 55–65 lm). The
head supported by a vacuum cushion (Fig. 1b). The image average measured stromal thickness was 480 lm in dogs
capture unit, supported by an articulated arm, was brought (450–560 lm) and 540 lm in cats (485–595 lm). The
close to the eye and then focused using an adjustment number of healthy animals imaged was not sufficient to
knob. Care was taken to humidify the corneal surface with perform a statistical analysis on these parameters, and the
artificial tears during imaging to avoid desiccation artifact. endothelio-Descemet layer was not thick enough to be
Images were captured using a foot control and converted measured in healthy animals.
to two-dimensional B-scans on which it was possible to Although care was taken to humidify the corneal surface
draw, write, and measure with calipers. Image analysis was during examination, artifact lesions due to superficial cor-
performed in the same manner as ultrasound image analy- neal desiccation were observed and appeared as a notch in
sis and covered three aspects: structure identification, the corneal thickness, with a reduction in epithelial thick-
description of lesions by analysis of light interference areas ness (Fig. 3a,b).
(increased or reduced reflection), and measurements. The
same operator (Frank Famose) performed image examina- Epithelial and subepithelial disorders
tion and processing. Thirteen chronic epithelial ulcers were examined. The
main features observed in both species were an increase in
the thickness and reflectivity of the epithelial layer, accom-
RESULTS
panied by epithelial detachment and a clearly visible
Normal cornea increase in anterior stromal reflectivity (Fig. 4a2). Epithe-
Examination in ‘pachymetry’ mode reveals the different lial thickness at the margins of the ulcer ranged from 80
layers in the cornea (Fig. 2a,b). The cornea appears as a to 150 lm.
composite of three layers of varying reflectivity and thick- Three cases of canine chronic superficial keratitis with
ness. The epithelial layer appears homogeneous with a corneal melanosis were examined. In all cases, we observed
(a)
Figure 2. Normal cornea of dog (a) and cat (b)
in pachymetric mode. The values in green
represent the total thickness of the cornea and
(b)
that of the epithelium.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

4. 4 famose
Figure 3. Corneal modification due to
desiccation. The red arrow represents the section
area on the macroscopic view (a). OCT (b)
revealed reduced epithelial thickness (white
(a) (b)
arrows).
(a1) (a2)
Figure 4. Epithelial and subepithelial conditions.
The red arrow represents the section area. (a1,a2)
Chronic epithelial erosion in a dog. Hyperplastic
epithelium is shown by green arrows. Yellow
(b1) (b2) arrows show epithelial detachment. Stromal
increased density is shown by the white arrows.
(b1,b2) Superficial corneal melanosis. The
presence of melanocytes is shown by increased
superficial reflectivity (yellow arrows). Stromal
image is homogeneously attenuated. (c1,c2)
Presumed lipid and/or calcic degeneration.
Presumed lipid and/or calcic deposits are
identified by high reflectivity in the anterior
stroma (green arrows). Posterior stroma is
unchanged. Epithelium thickness is
heterogeneous. Melanosis (yellow arrows) and a
(c1) (c2)
corneal blood vessel (red arrow) are visible.
a sharp increase in the reflectivity of the epithelial layer, thickness was 380 and 530 lm, respectively, presumed
with a homogeneous attenuation of the stromal image lipid and/or calcic deposit depth ranged from 80 to
(Fig. 4b1,b2). This increased anterior reflectivity is due to 250 lm. Melanosis, and corneal blood vessels were clini-
the presence of melanocytes, including melanic pigments, cally present and visible on OCT scans.
which absorb light in the IR range.
In the two cases of presumed lipid and/or calcium cor- Stromal disorders
neal degeneration, we observed increased anterior stromal Ten cases of suspected bacterial keratitis were evaluated.
reflectivity with no modification of the posterior stroma. Positive cultures confirmed the diagnosis in eight of the
Variations of the epithelial thickness were observed and ten cases. SD-OCT analysis revealed a reduction in
were associated with the density of the presumed lipid stromal thickness localized to the anterior stroma (Fig. 5).
and/or calcium deposits (Fig. 4c1,c2). Total corneal The center of the lesion appeared either as a homoge-
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

5. sd-oct evaluation of canine and feline cornea 5
(a1) (a2)
(b1) (b2)
(c1) (c2)
Figure 5. Infectious keratitis and deep ulcers.
The red arrow represents the section area. (a1,a2) (d1) (d2)
infection with loss of surface substance (grey
arrows) in a homogeneous stroma. (b1,b2) deep
keratitis with an area of collagenolysis (yellow
arrows) and increased peripheral stromal
reflectivity. (c1,c2) deep corneal ulceration. Area
of stromal densification (white arrows) and debris
(pink arrow) are visible. Residual stromal
thickness can be measured. (d1,d2) deep keratitis
with loss of stromal continuity and central
elevation of deep stroma (green arrows). (e1,e2)
Predescemetic lesion. The residual corneal
(e1) (e2)
thickness is 50 lm.
neous loss of stroma or as a heterogeneous association of Spectral domain optical coherence tomography evalua-
high- and low-reflectivity zones. High-reflectivity zones tion of deep corneal ulcers revealed a homogeneous
corresponding to cellular infiltration surrounded these increase in stromal reflectivity surrounding the ulcerated
lesions. The epithelial layer was partially absent, and in area and the presence of debris at the bottom of the ulcer.
some cases of deep stromal destruction, Descemet’s Analysis allowed measurement of the ulcer depth, infor-
membrane was pushed forward by intraocular pressure. mation that is critical to choosing the appropriate surgical
Residual stromal thickness was measurable. treatment.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

6. 6 famose
The three corneal foreign bodies (Fig. 6) were charac- Three cases of major increase of the stromal thickness
terized by the presence of a posterior cone-shaped shadow were observed. In two cases of glaucoma with an intense
related to their opacity. Depending on the penetration corneal edema, corneal structure was maintained, but the
depth, an endothelial reaction was visible. Localization of spacing of collagen lamellae was altered (Fig. 9a1,a2). In
the end of the foreign body was challenging, and serial the other case, a feline bullous keratopathy, the normal
scans were required to identify perforations. stromal architecture was severely disrupted, and pockets of
We also examined two cases of Florida keratopathy liquid among highly modified stromal structures were
(Fig. 7a1,a2). In SD-OCT images, these lesions were observed (Fig. 9b1,b2). Corneal thickness could not be
characterized by a clear increase in superficial and deep measured in either case due to the size of the cornea,
stroma reflectivity. This increased reflectivity was deeper which was not entirely visible on the screen.
than that observed in the case of stromal hemorrhage
(Fig. 7b1,b2). The lesions were not accompanied by epi- Endothelial disorders
thelial alteration, an increase in stromal thickness, or In a case of persistent pupillary membrane with localized
increased IR light reflexion behind the lesions. endothelial dystrophy (Fig. 10a1,a2), a localized endothe-
Spectral domain optical coherence tomography imaging lial hyper-reflectivity was observed supported by a sus-
of feline corneal sequestrum was challenging. In the five pended structure in the anterior chamber. Filaments were
cases examined, sequestrum appeared as an intense, local- difficult to highlight in two-dimensional scans, and serial
ized light reflection in the anterior stroma. IR light reflec- scans were required to monitor the filament trajectory.
tion varied according to the density and thickness of the In the three cases of anterior uveitis, precipitates behind
sequestrum. In two cases, the posterior stroma remained the cornea appeared in SD-OCT imaging as highly reflec-
visible (Fig. 8a1,a2), making it possible to measure the tive wide-based lesions on the endothelium (Fig. 10b1,b2).
thickness of the necrotic area before making a decision A diffuse inhomogeneous granular reflectivity of the ante-
concerning an operation. In the three remaining cases, this rior chamber was observed. Anterior synechia appeared as
pre-operative evaluation was not possible due to high dense lesions attached to the endothelium in continuity
lesion reflectivity (Fig. 8b1,b2). with the anterior face of the iris (Fig. 10c1,c2).
(a1) (a2)
(b1) (b2)
Figure 6. Corneal foreign body. The red arrow
represents the section area. Corneal foreign body
(FB) appears with a strong IR absorption with
posterior cone-shaped shadow (***). Endothelial
(c1) (c2) reaction (yellow arrows) is visible. Serial scans
(a1–c2) show total corneal perforation.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

7. sd-oct evaluation of canine and feline cornea 7
(a1) (a2)
Figure 7. Diffuse stromal lesions. The red arrow
represents the section area. (a1,a2) Florida spots
with stromal cellular infiltration (white arrows).
(b1) (b2)
(b1,b2) stromal hemorrage (green arrows).
(a1) (a2)
Figure 8. Feline corneal sequestrum. The red
arrow represents the section area. (a1,a2) corneal
sequestrum (grey arrow) with deterioration of
stromal signal. (b1,b2) corneal sequestrum with
(b1) (b2)
total signal absorption (yellow arrows).
Surgical disorders DISCUSSION
Corneal structure was evaluated after conjunctival grafts in
eight animals using SD-OCT (Fig. 11a1,a2). The images The goal of this study was to test the feasibility of using
obtained showed a conjunctival structure covered with an SD-OCT to image the cornea in cats and dogs in com-
epithelium and embedded in the stroma. The conjunctival mon practice situations. Our results confirm that this
structure was recognizable due to higher density and the technique is applicable to the evaluation of the cornea in
presence of blood vessels. healthy and pathological conditions. For most cases stud-
In the 14 cases of corneal wounds, the images obtained ied, SD-OCT evaluation provides reliable and accurate
did not identify the precise area of perforation. However, additional information to the standard examination using
images did show the presence of fibrin and blood in the a slit lamp for corneal disorders. Through the study of
anterior chamber (Fig. 11b1,b2). Evaluation of corneal many different conditions, we were able to underline diag-
incisions (Fig. 11c1,c2) to check the wound edge junction nosis- and image-related advantages and limitations of
was possible, although assessment of deep sections of the SD-OCT.
stroma was difficult because of the IR absorption related This study is one of the first evaluations of SD-OCT
to the corneal edema. for corneal diseases in private practice conditions. This
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

8. 8 famose
(a1) (a2)
Figure 9. Major stromal alterations. (a1,a2)
intense corneal edema due to glaucoma in a cat.
Collagen lamellae are separated by aqueous
hyporeflective material. (b1,b2) Feline Bullous
keratopathy. Stromal structure is replaced by
(b1) (b2)
‘pockets’ of fluid.
(a1) (a2)
(b1) (b2)
Figure 10. Endothelial alterations. (a1,a2)
persistent pupillary membrane with endothelial
attachment (grey arrows). (b1,b2) retrokeratic
precipitates (green arrows) with heterogeneity of
(c1) (c2) anterior chamber. (c1,c2) anterior synechia (green
arrows).
means, we have evaluated the pathological conditions that The SD-OCT imaging resolution was excellent at all
came through our clinic during the time of the study. depths, and we were able to measure the central cornea
Although a wide range of conditions has been examined, thickness with a resolution of 5 lm. This measurement
many other conditions such as extended endothelial required the device to be positioned axially in the center
dystrophies or eosinophilic keratitis were not evaluated of the cornea. The results obtained in the ‘pachymetry’
because we were not presented with those conditions mode were compatible with available data.18,19 However,
during the study. statistical analysis of these data was not the aim of this
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

9. sd-oct evaluation of canine and feline cornea 9
(a1) (a2)
(b1) (b2)
Figure 11. Surgical disorders. The red arrow
represents the section area. (a1,a2) conjunctival
graft in a dog. The conjunctiva (green arrows) is
denser than the cornea and has blood vessels
(yellow arrows). (b1,b2) traumatic cornea wound.
Blood clot is present near the edge of the wound.
(c1,c2) Surgical corneal wound in a cat. Wound (c1) (c2)
edge coaptation is shown by white arrows.
study and has not been performed. The influence of the of high reflectivity in the anterior stroma.23 We were able
corneal curvature on the pachymetry value is known in to measure the thickness of these high-density regions in
humans,20 but has not been measured in this study. As is vivo using SD-OCT. These observations were similar to
the case with humans, the central trajectory of IR illumi- human data produced from analysis of presumed calcified
nation was accompanied by a central reflection that could lesions.24,25 Precise localization in the subepithelial
disturb tissue analysis. This reflection was not systematic stroma, specific reflectivity, and evaluation of the deposits
and was related to the perpendicular position of the inci- thickness provided a noninvasive confirmation of clinical
dent light in relation to the corneal surface.7 diagnosis. The deepest stromal lesions, Florida spots, were
In all pathological cases studied, qualitative and quanti- also characterized by the presence of an infiltrating com-
tative analyses of the lesions were possible using ponent without any epithelial modification. SD-OCT
SD-OCT. In the case of chronic superficial corneal ulcer- imaging of major stromal changes such as complete cor-
ation, we were able to detect both epithelial detachment neal edema and bullous keratopathy produced spectacular
and hyperplasia, as well as the increased reflectivity of the pictures showing profound changes in the stromal archi-
anterior stroma. The analysis of bacterial keratitis high- tecture that could not be evaluated by slit-lamp examina-
lighted the presence and the intensity of cellular infiltra- tion. What we call feline bullous keratopathy is named, in
tion, corneal edema, and tissue destruction. Measurements human ophthalmology, corneal hydrops, a condition asso-
made on the different compartments of the cornea allowed ciated with endothelial rupture or detachment.26 SD-OCT
us to monitor the development of these lesions over time. examination failed to show endothelial structures in bul-
Our observations were similar to those performed in lous keratopathy because it was not technically possible to
human ophthalmology for which SD-OCT is now a useful have the entire corneal thickness on the same scan, so
additional technique in the evaluation of bacterial kerati- results could, therefore, not be compared with human
tis.21,22 However, in some cases, the hyper-reflectivity of data.27 Similarly, when imaging endothelial lesions, the
lesions limited the assessment of deep corneal areas, and SD-OCT examination provided an accurate evaluation as
as a result, evaluation of the entire stroma was not always long as the corneal thickness was 1300 lm, and thus,
possible. absorption was minimal. This limitation can be further
Images of corneal dystrophy and degeneration were cor- investigated by imaging endothelial dystrophies and
related with available histologic data by identifying areas degenerations, which were not included in our study. This
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

10. 10 famose
limitation could also be handled by the use of UBM that investigative field that is comparable to the existing field
uses high-frequency ultrasound, which does not depend in human ophthalmology. In our study, the advantages of
on corneal transparency. UBM can scan extremely thick this technique were found in the ability of giving a quali-
corneas but with a lower resolution than OCT.6 tative and quantitative evaluation of all cornea layers, in
We also used SD-OCT for evaluation and follow-up of vivo, in real time. The rapid acquisition time and the
surgical procedures. The preoperative analysis of corneal absence of contact with the ocular surface make the
sequestra in cats was not reliable. Our aim was to antici- method insensitive to eye movements and compatible
pate the choice of the surgical technique and of the prog- with the fragility of the cornea to be examined. Further-
nosis by preoperative evaluation. In two cases, the size and more, the production of quantitative images and measure-
position of the lesion were accurately measured, while in ments allows us to monitor clinical situations in an
other cases, the reflectivity of the pigmented lesion pro- objective manner, particularly in the context of microbial
duced a posterior shadow that made analysis of the deep keratitis.
stroma impossible. However, this lack of information However, we encountered three types of difficulties in
about the depth of the sequestrum did not change either the application of SD-OCT to corneal evaluation. First,
the treatment choice or the outcome of the surgical proce- focusing must be very precise, and it was very difficult to
dure. OCT was used after the keratectomy to measure the use the device on conscious animals. This fact led us to
corneal thickness and to detect residual hyper-reflective use SD-OCT on sedated animals only, which added the
areas. The shadow effect was also observed in the evalua- necessity of permission, from the owners, for repeated
tion of corneal foreign bodies. Serial scans were required sedations for follow-up imaging sessions. Additionally, due
to overcome this difficulty and accurately quantify the to the fine focus requirements, small lesions were not eas-
depth of penetration. ily detected during the examination, as was the case for
Preoperative and post-operative evaluation during the corneal wounds and foreign bodies. Serial scans were
superficial keratectomies and conjunctival or biomaterial required to obtain useable evaluation images in such cases.
(such as porcine intestine submucosa) grafts was possible Finally, for certain disorders, opacity of corneal structures
with SD-OCT. The thin graft material was transparent obscures image acquisition and interpretation, and thus,
enough to allow unobstructed imaging of underlying accurate analysis of corneal sequestra was difficult using
structures and the measurement of their dimensions. SD-OCT. In these cases, SD-OCT presented the same
These observations were similar to those made in humans limitations as slit-lamp examination. For highly edematous
where OCT is used frequently in the operative scope of lesions, examination was limited by corneal thickness for
corneal surgery to aid in choice of treatment strat- which it was not possible, with our device, to scan the
egy.2,5,7,11,25,28 Currently, more data are needed to support entire cornea. The use of UBM, which does not depend
this kind of protocol, but we believe that increasingly on corneal transparency but has a lower resolution than
widespread use of OCT will improve treatment strategies OCT, could help to evaluate such corneal conditions.
in the future. For example, the ability to accurately evalu- We were driven to complete this study because of the
ate corneal thickness before or during the surgical proce- possibility of obtaining equipment that is compatible with
dure will allow the veterinary surgeon to use novel veterinary practice. Here, we demonstrate the successful
surgical tools like lasers or corneal cross-linking, which use of the SD-OCT technique for the imaging and evalua-
should only be used when minimal residual corneal thick- tion of canine and feline cornea in clinical conditions for a
ness can be guaranteed. wide range of corneal diseases. This technique is not
The analysis of corneal wounds proved more difficult. intended to replace careful slit-lamp examination. The
Perforation areas were not always visible in SD-OCT results and images presented here show that SD-OCT
because they required the beam to be oriented accurately optical analysis has a resolution comparable to low-magni-
along the axis of the wound. In addition, the IR absorp- fication histologic images, and the images obtained are in
tion due to corneal edema limited the imaging of deep agreement with available clinical and histologic data in the
structures in most of the cases. In these conditions, the literature. In most of the cases, images provided us with
use of SD-OCT adds little to no value to the slit-lamp quantitative information that completed the slit-lamp
examination. examination. The major advantage of this technique is
The SD-OCT images of healthy corneas in carnivores real-time, in vivo, contactless evaluation of animal corneal
were comparable to images of the human cornea, where structures, and SD-OCT corneal evaluation in pathologi-
different layers are easily distinguishable.6 These results cal and surgical conditions is very promising diagnostic
correlate strongly with a previous study performed on a tool for therapeutic decision-making and for follow-up of
smaller group.29 However, the SD-OCT image of Desc- corneal healing.
emet’s membrane and the endothelium fades as the thick-
ness of the cornea increases.
ACKNOWLEDGMENTS
The use of SD-OCT for corneal evaluation in normal
and pathological conditions in dogs and cats opens an None.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11

11. sd-oct evaluation of canine and feline cornea 11
15. Soliman W, Mohamed T. Spectral domain anterior segment
REFERENCES
optical coherence tomography assessment of pterygium and
1. Konstantopoulos A, Hossain P, Anderson DF. Recent advances in pinguecula. Acta Ophthalmologica 2010; 90: 461–465.
ophthalmic anterior segment imaging: a new era for ophthalmic 16. Sorbara L, Maram J, Fonn D et al. Metrics of the normal cornea:
diagnosis? British Journal of Ophthalmology 2007; 91: 551–557. anterior segment imaging with the Visante OCT. Clinical and
2. Ramos JLB, Li Y, Huang D. Clinical and research applications Experimental Optometry 2010; 93: 150–156.
of anterior segment optical coherence tomography – a review. 17. Moutsouris K, Dapena I, Ham L et al. Optical coherence
Clinical and Experimental Ophthalmology 2009; 37: 81–89. tomography, Scheimpflug imaging, and slit-lamp biomicroscopy
3. Hirano K, Ito Y, Suzuki T et al. Optical coherence tomography in the early detection of graft detachment after descemet
for the noninvasive evaluation of the cornea. Cornea 2001; 20: membrane endothelial keratoplasty. Cornea 2011; 30: 1369–1375.
281–289. 18. Gilger BC, Whitley RD, McLaughlin SA et al. Canine corneal
4. Huang D, Swanson EA, Lin CP. Optical coherence tomography. thickness measured by ultrasonic pachymetry. American Journal of
Science 1991; 254: 1178–1181. Veterinary Research 1991; 52: 1570–1572.
5. Can I, Bayhan HA, Celik H et al. Anterior segment optical 19. Moodie KL, Hashizume N, Houston DL et al. Postnatal
coherence tomography evaluation and comparison of main clear development of corneal curvature and thickness in the cat.
corneal incisions in microcoaxial and biaxial cataract surgery. Veterinary Ophthalmology 2001; 4: 267–272.
Journal of Cataract and Refractive Surgery 2011; 37: 490–500. 20. Wirbelauer C, Thannhauser L, Pham DT. Influence of corneal
6. Dada T, Sihota R, Gadia R et al. Comparison of anterior curvature on central and paracentral pachymetry with optical
segment optical coherence tomography and ultrasound coherence tomography. Cornea 2009; 28: 254–260.
biomicroscopy for assessment of the anterior segment. Journal of 21. Konstantopoulos A, Kuo J, Anderson D et al. Assessment of the
Cataract and Refractive Surgery 2007; 33: 837–840. use of anterior segment optical coherence tomography in
7. Doors M, Berendschot TJM, De Brabander J et al. Value of microbial keratitis. American Journal of Ophthalmology 2008; 146:
optical coherence tomography for anterior segment surgery. 534–542.
Journal of Cataract and Refractive Surgery 2010; 36: 1212–1229. 22. Mantopoulos D, Cruzat A, Hamrah P. In vivo imaging of
8. Doors M, Tahzib NG, Eggink FA et al. Use of anterior segment corneal inflammation: new tools for clinical practice and
optical coherence tomography to study corneal changes after research. Seminars in Ophthalmology 2010; 25: 178–185.
collagen cross-linking. American Journal of Ophthalmology 2009; 23. Crispin SA. Crystalline corneal dystrophy in the dog.
148: 844–851. Histochemical and ultrastructural study. Cornea 1988; 7: 149–
9. Hashemi H, Asgari S, Mehravaran S et al. The Distribution of 161.
Corneal Thickness in a 40- to 64-Year-Old Population of 24. Wirbelauer C, Pham DT. Imaging and quantification of calcified
Shahroud, Iran. Cornea 2011; 30: 1409–1413. corneal lesions with optical coherence tomography. Cornea 2004;
10. Kim HY, Budenz DL, Lee PS et al. Comparison of central 23: 439–442.
corneal thickness using anterior segment optical coherence 25. Kim TI, Hong JP, Ha BJ et al. Determination of treatment
tomography vs ultrasound pachymetry. American Journal of strategies for granular corneal dystrophy type 2 using Fourier-
Ophthalmology 2008; 145: 228–232. domain optical coherence tomography. British Journal of
11. Ma JJK, Tseng SS, Yarascavitch BA. Anterior segment optical Ophthalmology 2010; 94: 341–345.
coherence tomography for transepithelial phototherapeutic 26. Stiles J, Townsend WM. Feline ophthalmology. In: Veterinary
keratectomy in central corneal scarring. Cornea 2009; 28: 927–929. Ophthalmology, 4th edn. (ed. Gelatt KN) Blackwell Publishing,
12. Matonti F, Hoffart L, Alessi G et al. Tomographie par cohrence
e Ames, 2007; 1095–1164.
optique en domaine spectral du segment antrieur: la 3e
e 27. Basu S, Vaddavallin PK, Vemuganti GT et al. Anterior Segment
ß
dimension. Journal franc ais d’ophtalmologie 2009; 32: 727–734. Optical Coherence Tomography Features of Acute Corneal
13. Nubile M, Dua HS, Lanzini M et al. In vivo analysis of stromal Hydrops. Cornea 2012; 31: 479–485.
integration of multilayer amniotic membrane transplantation 28. Schallorn JM, Tang M, Li Y et al. Optical coherence
in corneal ulcers. American Journal of Ophthalmology 2011; 151: tomography of clear corneal incisions for cataract surgery.
809–822. Journal of Cataract and Refractive Surgery 2008; 34: 1561–1565.
14. Qiu X, Gong L, Sun X et al. Age-related variations of human 29. Rosolen SG, Riviere ML, Lavillegrand S et al. Use of a
tear meniscus and diagnosis of dry eye with Fourier-domain combined slit-lamp SD-OCT to obtain anterior and posterior
anterior segment optical coherence tomography. Cornea 2011; 30: segment images in selected animal species. Veterinary
543–549. Ophthalmology 2012; 15: 105–115.
© 2013 American College of Veterinary Ophthalmologists, Veterinary Ophthalmology, 1–11