DOI: 10.1111/j.1365-3164.2009.00760.
Canine cutaneous epitheliotropic T-cell lymphoma with
Blackwell Publishing Ltd
vesiculobullous lesions resembling human bullous
mycosis fungoides
Petra Bizikova*, Keith E. Linder†,
Steven E. Suter*, Arnaud J. Van Wettere†
and Thierry Olivry*
*Department of Clinical Sciences, College of Veterinary Medicine,
North Carolina State University, Raleigh, North Carolina, USA
†Department of Population Health and Pathobiology, College of
Veterinary Medicine, North Carolina State University, Raleigh,
North Carolina, USA
Correspondence: Thierry Olivry, Department of Clinical Sciences,
College of Veterinary Medicine, Veterinary Teaching Hospital,
North Carolina State University, 4700 Hillsborough Street, Raleigh,
NC 27606, USA. E-mail: thierry_olivry@ncsu.edu
Source of Funding
This study is self-funded.
Conflict of Interest
No conflict of interest has been declared.
Abstract
The broad spectrum of clinical signs in canine cutaneous
epitheliotropic T-cell lymphoma mimics many inflam-
matory skin diseases and is a diagnostic challenge.
A 13-year-old-male castrated golden retriever crossbred
dog presented with multifocal flaccid bullae evolving
into deep erosions. A shearing force applied to the skin
at the periphery of the erosions caused the epidermis
to further slide off the dermis suggesting intraepidermal
or subepidermal separation. Systemic signs consisted
of profound weight loss and marked respiratory distress.
Histologically, the superficial and deep dermis were
infiltrated by large, CD3-positive neoplastic lymphocytes
and mild epitheliotropism involved the deep epidermis,
hair follicle walls and epitrichial sweat glands. There
was partial loss of the stratum basale. Bullous lesions
consisted of large dermoepidermal and intraepidermal
clefts that contained loose accumulations of neutrophils
mixed with fewer neoplastic cells in proteinaceous
fluid. The lifted epidermis was often devitalized and
bordered by hydropic degeneration and partial epidermal
collapse. Similar neoplastic lymphocytes formed small
masses in the lungs associated with broncho-invasion.
Clonal rearrangement analysis of antigen receptor
genes in samples from skin and lung lesions using
primers specific for canine T-cell receptor gamma
(TCRγ) produced a single-sized amplicon of identical
sequence, indicating that both lesions resulted
from the expansion of the same neoplastic T-cell
population. Macroscopic vesiculobullous lesions
with devitalization of the lesional epidermis should
© 2009 The Authors. Journal compilation © 2009 ESVD and ACVD. 20; 281–288
be included in the broad spectrum of clinical signs
presented by canine cutaneous epitheliotropic T-cell
lymphoma.
Accepted 10 February 2009
Introduction
Canine cutaneous epitheliotropic T-cell lymphoma is a
rare progressive neoplastic skin disease affecting the
epidermis, adnexal structures and mucosae. It is caused
predominantly by the proliferation of neoplastic γδTCR+,
CD3+ and CD8+ T-lymphocytes.1
Due to the substantial heterogeneity of its clinical and
microscopic features, several different variants have
been described in the literature. ‘Classic’ epitheliotropic
lymphoma is the most common type seen predominantly
in older dogs,2,3 with median age of 10 years.4,5 Three
clinical stages are recognized: patch, plaque and tumour;
however, an overlap of these stages is commonly seen.
Canine epitheliotropic lymphomas that present at a tumour
stage from the onset are described as the ‘d’emblée
form’.3,6 The coexistence of cutaneous epitheliotropic
lymphoma, spleen and lymph node involvement and
presence of circulating tumour cells in the peripheral blood
is described in human medicine as Sézary syndrome. This
form is very rare in dogs.7–9 Another variant of cutaneous
epitheliotropic lymphoma in dogs, called pagetoid reticulosis,
is described as an exfoliative erythroderma with alopecia,
erosions and ulcers without obvious neoplastic masses.
Mucocutaneous junctions and footpads are often affected
while generalized skin lesions are a common accompanying
feature.3,10
Canine epitheliotropic lymphoma is considered to be
a ‘great imitator’ because of its considerable clinical
pleomorphism. Indeed, early lesions are often misdiagnosed
as inflammatory skin disease.2 Four main clinical phenotypes
are usually seen in this entity: (i) exfoliative erythroderma,
(ii) mucocutaneous form, (iii) tumoural form and (iv) ulcerative
form affecting the oral cavity.6,11
To emphasize the clinical heterogeneity of this condition,
this paper reports a patient with unusual flaccid vesiculobullous
lesions diagnosed with an epitheliotropic T-cell lymphoma
confirmed by histopathology, immunohistochemistry and
polymerase chain reaction (PCR) for antigen receptor gene
rearrangement (PARR). Interestingly, the clinical signs and
histopathology closely resemble human bullous mycosis
fungoides (MF).
281
Bizikova et al.
Case report
History
A 13-year-old male neutered golden retriever crossbred
was presented originally to the referring veterinarian with
a 2-month history of mildly pruritic skin lesions on the face and
left shoulder. The lesions were described as erythematous
alopecic scaly patches with thickening and depigmentation
of the facial lesion. The dog was treated orally with cefalexin,
(Karalex Pharma, Woodcliff Lake, NJ, USA) (29.4 mg/kg
twice daily for 6 weeks) and chlorpheniramine, (Amneal
Pharmaceuticals, Paterson, NJ, USA) (4.0 mg per dog twice
daily for 1 month) without marked improvement. Moreover,
new lesions of similar character developed on the ventral
abdomen, pressure point areas and in the axillae. A skin
sample obtained by biopsy at that time revealed a lymphocytic
interface dermatitis and folliculitis with moderate epidermal
hyperplasia. Results of a complete blood count, serum
chemistry and urinalysis performed the same day were
within normal limits, and the antinuclear antibody (ANA) test
was negative. Treatment with prednisone, (West-ward
Pharmaceuticals, Eatontown, NJ, USA) (0.6 mg/kg twice daily
orally for 2 weeks), and tetracycline, (IVAX Pharmaceuticals,
Miami, FL, USA) and niacinamide (Major Pharmaceuticals,
Livonia, MI, USA) (500 mg per dog of each drug every 8 h
orally for 1 month), was initiated due to suspected discoid
lupus erythematosus. Ciprofloxacin (Teva Pharmaceuticals,
North Wales, PA, USA) (29.4 mg/kg once daily orally for
1 month) was added to prevent secondary bacterial infec-
tion of the eroded and ulcerated areas. However, due to
further worsening of skin lesions and general health, the
patient was referred to a dermatologist 3 weeks later.
Clinical presentation
At the time of presentation, the dog was still receiving the
ciprofloxacin, tetracycline and niacinamide treatment.
The dog was quiet but responsive with a subnormal
body condition (score: 1 of 9), normal temperature and
increased heart and respiration rates (130 and 48 per
minute, respectively). He was unable to walk unsupported
due to profound weakness and prominent muscle atrophy,
and exhibited severe respiratory distress with increased
bronchial sounds over the entire thorax. Palpation of the
282
Figure 1. Line diagrams illustrating the
distribution of the skin lesions (grey colour) at
the time of presentation to the dermatologist.
peripheral lymph nodes and abdomen did not reveal any
abnormalities.
The skin examination revealed multifocal, asymmetrically
distributed alopecia with flaccid vesiculobullous lesions
evolving into deep erosions that were predominantly
localized to the pressure points and friction areas of
the axillary and inguinal regions and inter-digital areas
(Figures 1 and 2a). A shearing force applied to the skin at
the periphery of the erosions caused the epidermis to
further slide off the dermis – a phenomenon known as the
direct Nikolskiy sign (Figure 2b) (see editors note).12 Addi-
tionally, a depigmented area on the caudal aspect of the
nasal planum (approximately 0.5 cm) merged into a well-
demarcated patch of alopecia with crusts and scarring that
affected the cranial third of the dorsal muzzle (Figure 2c).
An erythematous alopecic plaque with partial depigmenta-
tion expanding to the mucocutaneous junction was present
on the right side of the muzzle (Figure 2d). The oral cavity
and other mucosal surfaces were unaffected.
Clinical assessment and plan
The main differential diagnoses for the skin lesions included
epitheliotropic lymphoma, paraneoplastic pemphigus,
pemphigus vulgaris, necrolytic migratory erythema (hepato-
cutaneous syndrome) and bullous systemic lupus
erythematosus. Secondary bacterial infection of the
ulcerated areas was also suspected.
Pulmonary neoplasia (primary or metastatic) and
pneumonia of infectious origin were the major differential
diagnoses for the severe respiratory distress.
The following differential diagnoses were considered
for the profound cachexia, muscle atrophy and weakness:
cancer cachexia and inflammatory polymyositis. The latter
is a multifactorial condition that was reported in rare cases
of canine multicentric lymphoma.13
A complete blood count with haematology differential
showed only mild nonregenerative anaemia. A serum
chemistry analysis was unremarkable. Further investiga-
tions including urinalysis, thoracic radiographs and skin
cytology and biopsy of the skin were proposed to the
owner; however, because of the severe respiratory distress,
poor prognosis and marked discomfort of the patient, the
client requested euthanasia of the patient.
© 2009 The Authors. Journal compilation © 2009 ESVD and ACVD.
 Canine bullous mycosis fungoides

Figure 2. Skin lesion spectrum found during the examination: (a) An eroded area previously affected with a flaccid vesiculobullous lesion. (b) The
consequence of a shearing force applied to the skin at the periphery of the erosions; the epidermis has slid further off the dermis – the direct Nikolskiy
sign. (c) A depigmentation of the caudal aspect of the nasal planum merged into a well-demarcated patch of alopecia with crusts and scarring affecting
the cranial third of the dorsal muzzle. (d) An erythematous, partially alopecic plaque with partial depigmentation extending to the mucocutaneous junction.

Post-mortem examination
Gross examination revealed a thin body condition with
marked atrophy of axial and appendicular skeletal muscles
and skin lesions compatible with those described above.
Upon opening the thoracic cavity, multiple (40 to 60), 0.3
to 2.0 cm tan-white, firm, spherical masses were dis-
tributed multifocally throughout the parenchyma of the
right and left lung lobes. These nodules had a nearly flat
surface where they contacted the pleura (Figure 3).
Mediastinal lymph nodes were slightly enlarged (approxi-
mately two- to threefold). In the liver, there were a few,
scattered, slightly raised tan nodules of variable size (0.5
to 1.5 cm). A larger (6.0 cm) friable and spherical mass
with a tan to pale mahogany colour was found in the left
lateral liver lobe. The heart appeared grossly normal.

Histopathology
A perivascular to coalescing infiltrate of large neoplastic
lymphocytes expanded the superficial and deep dermal
stroma and extended a short distance into the panniculus
(Figure 4a). Segments of intimate epidermal contact by
neoplastic cells were accompanied by multifocal areas of
epitheliotropism in the deep epidermis (Figure 4d) and
in occasional hair follicles and epitrichial sweat glands.
Nuclear profiles ranged from round to slightly oval, were
sometimes slightly indented and exhibited moderate
anisokaryosis. Zero to four mitotic figures were seen per
Figure 3. Post-mortem examination – lungs: Examples of the tan-white,
firm, spherical masses (0.3 to 2.0 cm, arrows) in the parenchyma of a
lung lobe.
one high-power microscopic field (×400) in most areas. The
cells had small to moderate amounts of grey cytoplasm and
moderate anisocytosis. Discrete Pautrier’s micro-aggregates
were not a feature. Vesicular and bullous lesions consisted
of dermoepidermal and deep intraepidermal clefts that

© 2009 The Authors. Journal compilation © 2009 ESVD and ACVD. 283
Bizikova et al.

Figure 4. Histopathology: (a) A superficial and deep dermal infiltrate of large neoplastic lymphocytes expands the dermis and superficial pannic-
ulus. Epidermal lifting (arrow) leading clinically to vesiculobullous lesions is apparent even at this low magnification. Haematoxylin and eosin (H&E),
bar = 5.0 mm. (b) Jumbling and loss of basal keratinocytes due to necrosis lead to early intrabasal clefting at the margins of bullous lesions (arrow-
heads). H&E, bar = 100 μm. (c) Bullous lesions consist of large dermoepidermal and deep intraepidermal clefts containing loose accumulations of
neutrophils admixed with fewer neoplastic cells in proteinaceous fluid. The lifted epidermis is undergoing necrosis (asterisk). H&E, bar = 200 μm.
(d) Multifocal areas of epitheliotropism present in the deep portion of the epidermis (arrowheads). H&E, bar = 100 μm. (e) Immunohistochemical
staining of a serial histological section, of the one depicted in image Figure 4d, exhibits diffuse membranous staining of the epitheliotropic and
dermal neoplastic cells with CD3, which confirmed the T-cell origin of the neoplasm (positively stained cells are brown), bar = 100 μm.

contained loose accumulations of neutrophils admixed with
fewer neoplastic cells in proteinaceous fluid (Figure 4c).
At marginal areas and away from bullous lesions variable
basal keratinocyte death and loss, mild to severe, was
evident with vacuolation and disorganization of remaining
basal cells (Figure 4b). Vesicles appeared to form first in
the basal layer and/or in the immediate supra-basal layers
due to loss of keratinocytes. Centrally in vesicles and bullae,
where the dermis formed the base, the lifted epidermis
was usually extensively collapsed and/or completely necrotic
often with significant loss of the viable layers of epidermis.
Remaining cells exhibited hydropic degeneration usually
in marginal areas. These changes were superimposed
on a diffusely hyperplastic epidermis with moderate to
marked, laminated, orthokeratotic hyperkeratosis of the
stratum corneum, which often formed the central roof
of developed bullous lesions. A periodic acid-Schiff stain
revealed an intact basement membrane zone below bullae,

284 © 2009 The Authors. Journal compilation © 2009 ESVD and ACVD.

which supported the clinically observed deep erosions
and correlated with the lack of significant haemorrhages
in most lesions. Neoplastic cells were tightly perivascular
and tracked vascular structures, but clear vaso-invasion
was not observed.
In the lungs, similar neoplastic lymphocytes formed
spherical to ovoid, nonencapsulated masses that were
located near bronchi and medium-sized vessels, which
often coalesced and tracked these structures where cells
were intimately associated with the adventitia. Broncho-
invasion involved the submucosa in several areas. In the
heart, a few scattered microscopic foci of neoplastic
lymphocytes were observed in the myocardium of the left
ventricle. Intrathoracic lymph nodes were oedematous
and had mild reactive changes but infiltration by neoplastic
cells was not identified. The neoplastic process did not
involve the spleen, bone marrow or any other organs
examined during the complete post-mortem evaluation.
Histology confirmed multifocal marked hepatocellular
nodular hyperplasia and a focal cystadenoma in the liver.
Immunohistochemistry
Immunohistochemical staining using a routine immuno-
histochemical technique employing a rabbit polyclonal
antiserum specific for a conserved sequence of the
epsilon chain of CD3 (Dako, Carpinteria, CA, USA) supported
the T-lymphocyte origin of the neoplasm as there was strong,
diffuse membranous staining of over 95% of neoplastic
cells in the skin and lung, including epitheliotropic cells in
the epidermis (Figure 4e).14
Indirect immunofluorescence
Indirect immunofluorescence was performed to exclude
the possible coexistence of an autoimmune blistering skin
disease. For this purpose, the patient serum was tested
using normal and salt-split canine lip tissue as described
previously.15,16
Using both substrates, circulating IgG, IgA or IgE
autoantibodies against keratinocyte or basement membrane
antigens at 1 : 10 serum dilution were not detected.
PCR for antigen receptor rearrangements (PARR)
PCR for antigen receptor rearrangements (PARR) was
performed as previously described17,18 with minor modifica-
tions for its use in paraffin-embedded tissues. Briefly, 1.2 mL
xylene was added to four 20-μm sections from skin and
lung lesions and vortexed vigorously. After centrifugation
at 8000 g, the supernatant was removed and the tissue was
washed twice with 1.2 mL 100% ethanol. The samples were
then incubated at 37 °C for 15 min for ethanol removal.
Next, 180 μL tissue lysis buffer ATL plus 20 μL proteinase
K (Qiagen DNeasy Kit, Qiagen, Valencia, CA, USA) was
added, and the samples were incubated overnight at
56 °C. DNA isolation proceeded the next day as per the
manufacturer’s instructions (Qiagen DNeasy Kit). DNA was
quantified using a NanoDrop ND 1000 spectrophotometer
(NanoDrop Technologies, Wilmington, DE, USA). Amplicons
were separated through a 10% PAGE gel, stained with ethi-
dium bromide, and photographed using a multiwavelength
illuminator (Alpha Innotech, San Leandro, CA, USA).
For amplicon sequence analysis, the PCR products
were cloned (USERTM Friendly Cloning Kit, New England
© 2009 The Authors. Journal compilation © 2009 ESVD and ACVD.
Canine bullous mycosis fungoides
Figure 5. PCR for antigen receptor gene rearrangement (PARR):
This technique produced a single sized amplicon (125 bp) from
skin and lung tissue samples. Abbreviations: (–) negative control,
(+) positive control, a case of T-cell lymphoma, (h) heated sample,
(nh) nonheated sample.
Biolabs, Ipswich, MA, USA), amplified in DH-5α cells and
selected via blue/white X-Gal screening using standard
techniques. Plasmid DNA was isolated from white colonies
(QIAprep Spin Miniprep Kit) as per the manufacturer’s
instructions and sequenced using M13/pUC forward and
reverse sequencing primers supplied with the cloning
kit (Cogenics, Morrisville, NC, USA). The sequences of
three clones from each site (skin and lung) were analysed
(MegAlign, DNASTAR, Madison, WI, USA).
Using this technique, a single-sized amplicon of identical
sequence from skin and lung tissue samples was produced
(Figure 5).
Discussion
Human bullous MF is an extremely rare condition and, as
such, the diagnosis should be questioned always. Indeed,
coincidental vesiculobullous lesions can arise from other
non-neoplastic causes; for example, an autoimmune
blistering disease can coexist with lymphoma. Interest-
ingly, the simultaneous occurrence of MF and pemphigus
foliaceus or bullous pemphigoid has been described
previously in the human medical literature.19
In fact, the main clinical differential diagnoses in the
present case included several autoimmune blistering skin
diseases such as pemphigus vulgaris, paraneoplastic
pemphigus and bullous systemic lupus erythematosus.
Clinical signs suggestive of pemphigus vulgaris and
paraneoplastic pemphigus were the presence of flaccid
vesicles evolving rapidly into deep erosions and ulcers and
a positive direct Nikolskiy sign. However, the lack of
mucosal involvement and the relative asymmetry of the
lesions in this patient were in contrast with the characteristic
clinical picture of both of these conditions.20–22 Bullous
systemic lupus erythematosus is an extremely rare entity
that has been described only once in dogs.23 The lack of
other supportive clinical signs compatible with systemic lupus
erythematosus made this diagnosis unlikely in this dog.
Necrolytic migratory erythema was originally considered
as a probable diagnosis because of the age, progressive
weight loss and the presence of the erosive lesions on
pressure points and friction areas. The development of the
erosions in this condition is best explained by the prominent
ballooning degeneration of keratinocytes in the superficial
layer of stratum spinosum and granulosum that eventually
leads to separation.24,25 The fact that the erosions were
deep, and characteristic lesions on the footpads, lips and
eyelids were absent, made the diagnosis of necrolytic
migratory erythema unlikely.
285
Bizikova et al.
Finally, the presence of a plaque-like lesion on the upper
lip and depigmentation of the caudal portion of the nasal
planum led to the inclusion of epitheliotropic lymphoma on
the list of differential diagnoses; however, the coexistence
of unique vesiculobullous lesions in this patient was not
typical of this entity.
In this dog, the histopathology of skin lesions confirmed
the diagnosis of epitheliotropic lymphoma with unusual
intrabasal and suprabasal clefting that progressed to bulla
with necrosis of the deep layers of the epidermis. To the
best of the author’s knowledge, except for one personal
observation mentioned in a recent dermatohistopathological
publication, a similar combination of clinical and histological
lesions has not been described in the veterinary literature.6
Intraepidermal and subepidermal clefting are distinctive
histological features of several autoimmune blistering skin
diseases and therefore negative immunofluorescence
test results have been proposed as one of the diagnostic
criteria for human bullous MF.19 The indirect immunofluo-
rescence test utilizing normal and salt-split lip tissue in this
dog failed to detect circulating antibodies against keratino-
cytes or basement membrane ruling out the coexistence
of epitheliotropic lymphoma and pemphigus vulgaris or
other autoimmune blistering skin disease.
Similarly, dogs with paraneoplastic pemphigus usually
exhibit circulating antibodies against keratinocytes and
basement membrane,26,27 and the absence of the antibodies
and the histological and clinical incompatibility ruled out
this diagnosis definitively.
The mechanism of blister formation in human bullous
MF is not clear and it can only be speculated in this patient.
One of the proposed mechanisms for the development
of subepidermal and intrabasal separation is an excessive
subepidermal aggregation of neoplastic lymphocytes under
the epidermis with prominent epitheliotropism in the basal
cell layer eventually leading to loss of cohesion between
basement membrane and basal cells. Moreover, the release
of cytokines and chemokines from the neoplastic cells
may also interfere with normal epidermal cohesion and
eventually lead to keratinocyte cell death and formation of
intrabasal or intraepidermal blisters.19 The latter mechanism
could explain the blister formation in the dog as there was
individual basal cell necrosis and prominent necrosis of
the lifted epidermis.
Another proposed mechanism for intraepidermal blister
development is the coalescence of the Pautrier’s micro-
aggregates,19 but these were not a feature of this case.
Bacterial culture is recommended in cases of human
bullous MF because some bacteria are known to produce
exfoliative toxins responsible for superficial acantholysis.28
Although bacterial culture was not performed, it is unlikely
that bacterial infection alone – even if caused by exfoliative
toxin-producing bacteria – could explain the depth of the
clefting observed in the canine samples.
The pulmonary involvement in canine epitheliotropic
T-cell lymphoma is a very rare feature reported previously
in only one dog with Sézary syndrome.7 This diagnosis
could not be supported in the present case due to a lack
of lymphocytosis and absence of atypical neoplastic
lymphocytes in the peripheral blood. Regrettably, neither
flow cytometry nor PARR of peripheral blood could be
performed.
286
Finally, PARR using TCRγ primers confirmed that the skin
and lung lesions were the result of a neoplastic expansion
of the same T-cell clone. This method is now used to detect
lymphoid malignancies in dogs with a reported sensitivity
and specificity of approximately 91% and 96%, respectively.17
Factors affecting the sensitivity of this test include the
number of the rearrangements that bind to the TCRγ
primers, the number of neoplastic cells in the sample, and
the quantity of ‘background’ immune cells in the neoplastic
lesion. In contrast to studies reported previously, the PARR
analysis was performed on formalin-fixed paraffin-embedded
tissue. This technique is widely used in human medicine,
and previous studies confirmed an identical sensitivity
regardless of tissue fixation (formalin-fixed, paraffin-embedded
versus fresh-frozen tissue) and biopsy site when the
histological degree of confidence was very high (i.e.
diagnostic).29
To increase specificity of this technique, the lung and
skin tissue samples were duplicated to rule out pseudo-
clonality, and heating and re-annealing were used to rule
out any potential oligo- or polyclonality. To the authors’
best knowledge, this is the first reported use of PARR
followed by amplicon sequence analysis in veterinary
medicine to establish that cutaneous and lung lesions are
derived from an identical neoplastic clone. This technique
may be useful in the future to definitively prove that
lesions of lymphoid origin at varying sites are the result of
the proliferation of one neoplastic clone, multiple clones,
or that they are not neoplastic.
In conclusion, this report presents a clinically unusual
case of canine epitheliotropic T-cell lymphoma with
vesiculobullous lesions resembling human bullous MF, which
was diagnosed by histopathology, immunohistochemistry
and PARR. Moreover, the patient fulfilled the majority
of the criteria that have been proposed for diagnosis of
human bullous MF.19
These criteria include:
1 Clinically apparent vesiculobullous lesions with other
findings compatible with epitheliotropic lymphoma.
2 Typical histological features of epitheliotropic lymphoma
with intrabasal, intraepidermal or subepidermal blisters.
3 Negative immunofluorescence to rule out concurrent
autoimmune blistering diseases.
4 Negative evaluation for other possible causes includ-
ing a lack of previous therapy that can lead to blister
formation.
This case report also emphasizes the clinical heterogeneity
of canine epitheliotropic lymphoma and the suitability of
its description as a ‘great imitator’.
Therefore, macroscopic vesiculobullous lesions with
devitalization of the lesional epidermis should be included
in the broad spectrum of clinical signs exhibited by canine
cutaneous epitheliotropic T-cell lymphoma.
Acknowledgements
The authors thank Sandra Horton and Monica Mattmuller
for careful preparation and processing of the histological
and immunohistochemical slides.
© 2009 The Authors. Journal compilation © 2009 ESVD and ACVD.
 Canine bullous mycosis fungoides

13. Evans J, Levesque D, Shelton GD. Canine inflammatory myopathies:

Editor’s note
Readers may note that the spelling Nikolskiy has been
used rather than the more common Nikolsky – this has
been adopted in recognition of the correct spelling of
the famous dermatologist who described this technique
which bears his name. For more information see Grando
SA, Grando AA, Glukhenky BT et al. (2003) History and
clinical significance of mechanical symptoms in blistering
dermatoses: A reappraisal. Journal of the American
Academy of Dermatology 48: 86–92.
References
1. Moore PF, Affolter VK, Olivry T. The use of immunological reagents
in defining the pathogenesis of canine diseases involving prolifera-
tion of leucocytes. In: Kwochka KW, Willemse T, von Tscharner C,
eds. Advances in Veterinary Dermatology, Vol. 3. Oxford, UK:
Butterworth-Heinemann, 1998: 103–15.
2. Wilcock BP, Yager JA. The behavior of epidermotropic lymphoma
in twenty-five dogs. The Canadian Veterinary Journal La Revue
Veterinaire Canadienne 1989; 30: 754–6.
3. Moore PF, Olivry T, Naydan K. Canine cutaneous epitheliotropic
lymphoma (mycosis fungoides) is a proliferative disorder of CD8+
T cells. American Journal of Pathology 1994; 144: 421–9.
4. Williams LE, Rassnick KM, Power HT et al. CCNU in the treatment
of canine epitheliotropic lymphoma. Journal of Veterinary Internal
Medicine 2006; 20: 136–43.
5. Risbon RE, de Lorimier LP, Skorupski K et al. Response of canine
cutaneous epitheliotropic lymphoma to lomustine (CCNU): a
retrospective study of 46 cases (1999–2004). Journal of Veterinary
Internal Medicine 2006; 20: 1389–97.
6. Gross TL, Ihrke PJ, Walder EJ et al. Lymphocytic tumors. In: Skin
Diseases of the Dog and Cat: Clinical and Histopathologic Diagnosis,
2nd edn. Oxford, UK: Blackwell Science, 2005: 866–93.
7. Foster AP, Evans E, Kerlin RL et al. Cutaneous T-cell lymphoma with
Sézary syndrome in a dog. Veterinary Clinical Pathology 1997; 26:
188–92.
8. Thrall MA, Macy DW, Snyder SP et al. Cutaneous lymphosarcoma
and leukemia in a dog resembling Sézary syndrome in man.
Veterinary Pathology 1984; 21: 182–6.
9. DeBoer DJ, Turrel JM, Moore PF. Mycosis fungoides in a dog:
demonstration of T-cell specificity and response to radiotherapy.
Journal of the American Animal Hospital Association 1990; 26:
566–72.
10. Johnson JA, Patterson JM. Canine epidermotropic lymphoprolifera-
tive disease resembling pagetoid reticulosis in man. Veterinary
Pathology 1981; 18: 487–93.
11. Scott DW, Miller WH Jr, Griffin CE. Neoplastic and nonneoplastic
tumors. In: Small Animal Dermatology. 6th edn. Philadelphia, PA:
W.B. Saunders Company, 2001: 1236–414.
12. Grando SA, Grando AA, Glukhenky BT et al. History and clinical
significance of mechanical symptoms in blistering dermatoses: a
reappraisal. Journal of the American Academy of Dermatology
2003; 48: 86–92.
a clinicopathologic review of 200 cases. Journal of Veterinary
Internal Medicine 2004; 18: 679–91.
14. Mason D, Cordell J, Brown M et al. Detection of T cells in paraffin
embedded tissue using antibodies against a peptide sequence
from the CD3 antigen. Journal of Clinical Pathology 1989; 42:
1194–200.
15. Favrot C, Dunston SM, Deslandes J et al. Effect of substrate selec-
tion on indirect immunoflorescence testing of canine autoimmune
subepidermal blistering diseases. Canadian Journal of Veterinary
Research 2002; 66: 26–30.
16. Olivry T, Dunston SM, Schachter M et al. A spontaneous canine
model of mucous membrane (cicatricial) pemphigoid, an autoimmune
blistering disease affecting mucosae and mucocutaneous
junctions. Journal of Autoimmunity 2001; 16: 411–21.
17. Burnett RC, Vernau W, Modiano JF et al. Diagnosis of canine
lymphoid neoplasia using clonal rearrangements of antigen receptor
genes. Veterinary Pathology 2003; 40: 32–41.
18. Vernau W, Moore PF. An immunophenotypic study of canine
leukemias and preliminary assessment of clonality by polymerase
chain reaction. Veterinary Immunology and Immunopathology
1999; 69: 145–64.
19. Bowman PH, Hogan DJ, Sanusi ID. Mycosis fungoides bullosa:
report of a case and review of the literature. Journal of the American
Academy of Dermatology 2001; 45: 934–9.
20. Olivry T. Spontaneous model of pemphigus vulgaris. In: Chan LS,
ed. Animal Models of Human Inflammatory Skin Diseases. Boca
Raton, FL: CRC Press, 2004: 263–73.
21. Elmore SA, Basseches J, Anhalt GJ et al. Paraneoplastic pemphigus
in a dog with splenic sarcoma. Veterinary Pathology 2005; 42: 88–
91.
22. Lemmens P, DeBruin A, Demeulemeester J et al. Paraneoplastic
pemphigus in a dog. Veterinary Dermatology 1998; 9: 127–34.
23. Olivry T, Savary KCM, Murphy KM et al. Bullous systemic lupus
erythematosus (type I) in a dog. Veterinary Record 1999; 145:
165–9.
24. Byrne KP. Metabolic epidermal necrosis-hepatocutaneous syndrome.
Veterinary Clinics of North America – Small Animal Practice 1999;
29: 1337–55.
25. Gross TL, Ihrke PJ, Walder EJ et al. Necrotizing diseases of the
epidermis. In: Skin Diseases of the Dog and Cat: Clinical and
Histopathologic Diagnosis. 2nd edn. Oxford, UK: Blackwell Science,
2005: 75–104.
26. DeBruin A, Muller E, Wyder M et al. Periplakin and envoplakin are
target antigens in canine and human paraneoplastic pemphigus.
Journal of the American Academy of Dermatology 1999; 40: 682–5.
27. Nishifuji K, Olivry T, Ishii K et al. IgG autoantibodies directed against
desmoglein 3 cause dissociation of keratinocytes in canine
pemphigus vulgaris and paraneoplastic pemphigus. Veterinary
Immunology and Immunopathology 2007; 117: 209–21.
28. Nishifuji K, Sugai M, Amagai M. Staphylococcal exfoliative toxins:
‘molecular scissors’ of bacteria that attack the cutaneous defense
barrier in mammals. Journal of Dermatological Science 2008; 49:
21–31.
29. Li N, Bhawan J. New insights into the applicability of T-cell receptor
gamma gene rearrangement analysis in cutaneous T-cell lymphoma.
Journal of Cutaneous Pathology 2001; 28: 412–8.

Résumé Les symptômes du lymphome T épithéliotrope canin présentent une large diversité et peuvent
mimer de nombreuses dermatoses inflammatoires ce qui constitue ainsi un défit diagnostique. Un croisé
Golden Retriever mâle de 13 ans présente une dermatose bulleuse multifocale évoluant en lésions profondes.
Une pression appliquée en périphérie des érosions fait glisser l’épiderme sur le derme, suggérant un clivage
intra-épidermique ou sub-épidermique. Les signes généraux consistent en une perte de poids importante
et une détresse respiratoire marquée. Histologiquement, le derme superficiel et le derme profond présentent
un infiltrat de larges lymphocytes néoplasiques, CD3+ avec un épitheliotropisme modéré de l’épiderme
profond, de la paroi des follicules pileux et des glandes sudoripares épitrichiales. Une perte partielle de la
couche basale est notée. Les bulles consistent en de larges clivages dermo-épidermiques et intra-épidermiques
contenant des neutrophiles et quelques cellules néoplasiques au sein d’un exsudat protéique. L’épiderme

© 2009 The Authors. Journal compilation © 2009 ESVD and ACVD. 287
Bizikova et al.

soulevé est souvent dévitalisé et bordé par une dégénérescence hydropique et un collapsus épidermique
partiel. Des lymphocytes néoplasiques identiques forment de petites masses pulmonaires associées à une
invasion bronchique. Le clonage des gènes codant pour les récepteurs antigéniques dans les échantillons
de lésions cutanées et pulmonaires utilisant des sondes spécifiques pour les récepteurs gamma des cellules
T (TCRγ) produisent des séquences identiques d’amplifications de même taille, indiquant que les deux
lésions sont issues de la même population de cellules T tumorales. Les lésions vésiculo-bulleuses macroscopiques
avec dévitalisation de l’épiderme devraient être inclues dans l’éventail des signes cliniques du lymphome
T épithéliotrope canin.

Resumen El amplio espectro de signos clínicos en el linfoma epiteliotrópico cutáneo canino mimetiza
muchas enfermedades inflamatorias de la piel y es un diagnóstico complicado. Un perro macho castrado
de 13 años de edad de raza cruzada Golden Retriever se presentó con bullas flácidas multifocales que
evolucionaron a erosiones profundas. La aplicación de tracción en la piel de la periferia de las erosiones
producía mayor separación de la epidermis, indicando una separación intraepidermal o subepidermal. Los
signos sistémicos consistieron en pérdida de peso marcada y respiración anormal. Histológicamente la
dermis superficial y profunda estaba infiltrada por linfocitos neoplásicos CD3 positivos de gran tamaño y con
ligero epiteliotropismo en la epidermis profunda, paredes de los folículos pilosos y glandulas sudoríparas.
Había una pérdida parcial del estrato basal. Las lesiones bullosas se formaban por separación dermo-
epidermal e intraepidermal extensa que contenía acúmulos de neutrófilos mezclados con un número menor
de células neoplásicas en fluido proteináceo. La epidermis que se había separado aparecía a menudo
devitalizada y rodeada de degeneración hidrópica y colapso parcial de la epidermis. Linfocitos semejantes
formaban pequeñas masas en los pulmones asociadas con invasión bronquial. El análisis de clonalidad en
el alineamiento de los genes del receptor antigénico en muestras de piel y pulmón, utilizando iniciadores
específicos para el receptor gamma de linfocitos T (TCRγ), produjo un solo amplicón de idéntica secuencia,
indicando que ambas lesiones resultaron de la expansión de la misma población neoplásica. Lesiones
vesiculobullosas a nivel macroscópico con devitalización de la epidermis deben incluirse en el amplio espectro
de signos clínicos con los que puede presentarse el linfoma cutáneo epiteliotrópico canino.

Zusammenfassung Die große Bandbreite an klinischen Symptomen des caninen kutanen epitheliotropen
T-Zell Lymphoms kann viele entzündliche Hauterkrankungen nachahmen und ist eine diagnostische
Herausforderung. Ein 13 Jahre alter, kastrierter Golden Retriever-Mischlingsrüde wurde mit multifokalen
schlaffen Blasen, die sich zu tiefen Erosionen entwickelten, vorgestellt. Scherkräfte, die an den Rändern
der Erosionen auf die Haut ausgeübt wurden, bewirkten, dass die Epidermis sich noch mehr von der Dermis
abhob, was auf eine intra-epidermale oder sub-epidermale Spaltung hindeutete. Die systemischen
Symptome bestanden aus einem markanten Gewichtsverlust und deutlicher Atemnot. Histologisch waren
die oberflächliche und die tiefe Dermis mit großen, CD3-positiven neoplastischen Lymphozyten infiltriert
und in der tiefen Epidermis, in den Wänden der Haarfollikel und in den epitrichen Schweißdrüsen bestand
ein geringgradiger Epitheliotropismus. Das Stratum basale war teilweise verloren gegangen. Die bullösen
Veränderungen bestanden aus großen dermo-epidermalen und intra-epidermalen Spalten, die lose
Ansammlungen von Neutrophilen gemischt mit wenigen neoplastischen Zellen in proteinöser Flüssigkeit
enthielten. Die abgehobene Epidermis war oft devitalisiert und von hydropischer Degeneration und
teilweisem epidermalen Kollaps begrenzt. Ähnliche neoplastische Lymphozyten bildeten kleine Umfangs-
vermehrungen in den Lungen aus, was mit einer Invasion der Bronchien einherging. Bei der klonalen
Rearrangement Analyse der Antigen-Rezeptorgene aus Haut- und Lungengewebsproben mittels Primers,
die spezifisch für den caninen T-Zell Rezeptor Gamma (TCRγ) waren, wurde ein einziges Amplikon gleicher
Größe und identischer Sequenz produziert, was darauf hinwies, dass beide Veränderungen aus der Expansion
derselben neoplastischen T-Zellen Population resultierten. Makroskopische vesikulo-bullöse Läsionen mit
Devitalisierung der läsionalen Epidermis sollten in die große Bandbreite der klinischen Symptome, die von
einem caninen kutanen epitheliotropen T-Zell Lymphom präsentiert werden können, inkludiert werden.

288 © 2009 The Authors. Journal compilation © 2009 ESVD and ACVD.