Morphologic, immunohistochemical, and molecular

characterization of hepatosplenic T-cell lymphoma

in a dog
Elizabeth A. Cienava, Kirstin F. Barnhart, Raquel Brown, Joanne Mansell, Robert Dunstan, Kelly Credille

Abstract: A 13-year-old neutered male Jack Russell Terrier (Parson Russell Terrier) was presented to the Texas Veterinary Medical
Center with a history of lethargy, depression, vomiting, and fever. The dog had mildly regenerative anemia, severe
thrombocytopenia and low antithrombin activity. Marked splenomegaly was found on physical examination and imaging
studies, and malignant round cell neoplasia and marked extramedullary hematopoiesis were diagnosed on aspirates of the
spleen. The dog underwent exploratory laporatomy and splenectomy. Because of a rapid decline in clinical condition post-
surgery, the dog was euthanized. Splenic and hepatic biopsies were submitted for histopathologic evaluation. A neoplastic
population of round cells was found throughout the splenic parenchyma and within hepatic sinusoids. The neoplastic cells
stained strongly positive for CD3 (T-cell marker) and were negative for CD79a (B-cell marker) and lysozyme (histiocytic marker).
A diagnosis of T-cell lymphoma was confirmed by assessment of T-cell clonality using canine-specific polymerase chain reaction–
based techniques. Although expression of the cd T-cell receptor was not evaluated, this case shares many similarities with a rare
syndrome in humans known as hepatosplenic cd T-cell lymphoma. (Vet Clin Pathol. 2004;33:105–110)

2004 American Society for Veterinary Clinical Pathology

Key Words: Canine, gene rearrangement, hepatosplenic lymphoma, immunohistochemistry, T-cell lymphoma

Case Presentation titers were negative. Because of persistent lethargy,

A 13-year-old neutered male Jack Russell Terrier
(Parson Russell Terrier) was presented to the referring
veterinarian with a 2-week history of lethargy, de-
pression, and intermittent vomiting. Physical examina-
tion revealed pyrexia (1048F), pale mucous membranes,
and marked splenomegaly. CBC results showed mild
lymphopenia (680/lL, reference interval 1.0–4.8 3 103/
lL), mild, normocytic normochromic nonregenerative
anemia (PCV 27.3%, reference interval 32–55%); and
clumped platelets. There was a mild decrease in total
serum protein concentration (5.5 g/dL, reference
interval 5.6–7.9 g/dL) with albumin concentration at
the low end of normal (3.0 g/dL, reference interval
3.0–4.5 g/dL).
Splenomegaly was confirmed with abdominal
radiography and ultrasonography. Splenic aspirates
were submitted to a diagnostic laboratory and inter-
preted as extramedullary hematopoiesis (EMH) with no
other abnormalities noted. The dog was treated initially
with 100 mg of doxycycline PO twice daily for
presumptive ehrlichiosis; however, serum Ehrlichia canis
fever, and a continued decline in PCV, the dog was
referred to the Texas Veterinary Medical Center (TVMC)
at Texas A&M University.
On admission to the TVMC, the dog was depressed,
had a mildly elevated body temperature of 102.98F, and
weighed 10.9 kg, with a body condition score of 3.5/5.
Heart rate and respiratory rate were within normal
limits. Marked splenomegaly was detected on ab-
dominal palpation. Mucous membranes were pale
and a grade 2/6 left heart base murmur was present.
Results of a CBC indicated moderate anemia (PCV
21.2%, reference interval 37.0–55.0%) that was slightly
regenerative on the basis of an absolute reticulocyte
count of 127,600/lL (reference interval , 80,000
reticulocytes/lL) and slight polychromasia and moder-
ate anisocytosis on the blood smear. The absolute nRBC
count was 1925/lL (20 nRBCs per 100 WBCs, total
WBC count 9400/lL). Plasma protein concentration
was slightly decreased at 5.6 g/dL (reference interval
6.0–8.0 g/dL). The neutrophil count was normal with
a mild left shift (band neutrophils 564/lL, reference
interval 0–300/lL). Platelet number was markedly

From the Departments of Veterinary Pathobiology (Cienava, Barnhart, Mansell, Dunstan, Credille) and Small Animal Medicine (Brown), College of Veterinary
Medicine, Texas A&M University, College Station, TX. Corresponding author: Elizabeth A. Cienava, DVM, Department of Veterinary Pathobiology, Texas A&M
University, Veterinary Medical Teaching Hospital, Clinical Pathology, Bldg 1085, Room 2020, College Station, TX 77843-9988 (ecienava@cvm.tamu.edu). ª2004
American Society for Veterinary Clinical Pathology

Vol. 33 / No. 2 / 2004 Veterinary Clinical Pathology Page 105

 Hepatosplenic T-Cell Lymphoma in a Dog
decreased at 34,000/lL (reference interval 200,000–
500,000/lL).
The BUN concentration was decreased (5 mg/dL,
reference interval 8.0–29.0 mg/dL) with a slight
hypoproteinemia (total plasma protein 5.3 g/dL,
reference interval 5.7–7.8 g/dL) and mild hypoalbumi-
nemia (albumin 2.3 g/dL, reference interval 2.4–3.6 g/
dL). Mild metabolic acidosis (total CO2 18 mmol/L,
reference interval 21.0–28.0 mmol/L) also was present.
Urine specific gravity was 1.036 with mild proteinuria
and bilirubinuria. Urine and blood cultures were
negative. Prothrombin and partial thromboplastin times
were within reference intervals, but antithrombin (AT)
activity was decreased at 59% of the control value
(reference interval . 84% of control value). Assays for
fibrin degradation products or d-dimers were not
performed. The dog remained depressed, lethargic,
and febrile (104.18F). Two units of plasma were
administered because of the low AT activity. The platelet
count declined to 30,000/lL on the day of admission.
Imaging studies included thoracic radiographs, an
echocardiogram (both performed on the day of admis-
sion), and abdominal ultrasound (on day 2). Thoracic
radiographs were unremarkable. Echocardiogram find-
ings included a thickened mitral valve with mild
regurgitation, consistent with chronic degenerative
valve disease. Endocarditis could not be excluded, and
the patient was started on broad-spectrum antibiotics.
Abdominal ultrasound revealed an enlarged, hyper-
echoic spleen and a small amount of free fluid within
the abdominal cavity. Differential diagnoses for the
splenomegaly included infectious disease, EMH, hyper-
splenism, and neoplasia. Splenic aspirates were ob-
tained for cytologic evaluation.
Cytologic evaluation
Splenic aspirates were highly cellular with abundant
nucleated cells and erythrocytes and a proteinaceous
background. Marked EMH with erythroid predomi-
nance and a few megakaryocytes was noted. The
lymphocyte population was heterogeneous but con-
sisted primarily of small lymphocytes. The number of
medium and large lymphocytes and plasma cells
appeared mildly increased. Erythrophagocytosis occa-
sionally was seen. No etiologic agents were found.
In addition to EMH, a population of large round
discrete cells was found individually and in loose
aggregates. The cells had moderate anisocytosis and
anisokaryosis with variable N:C ratios and small to
moderate amounts of pale to slightly basophilic
cytoplasm (Figure 1A). Rarely, the cells contained small
round cytoplasmic vacuoles. Nuclei were round to oval,
occasionally cleaved, and frequently eccentric, with fine
Page 106 Veterinary Clinical Pathology
to coarsely stippled chromatin and prominent bizarre
nucleoli (Figure 1B). Frequently, binucleated and mul-
tinucleated cells with moderate to marked anisokar-
yosis were noted. Occasionally, macrokaryotic cells
were seen, some with atypical mitoses.
The cytologic interpretation included EMH, re-
active lymphoid hyperplasia, and malignant round cell
neoplasia. On the basis of the dog’s clinical presenta-
tion, and the anatomic location and morphologic
characteristics of the neoplastic cells, the primary
differential diagnosis was maliganant histiocytosis.
In addition, hypersplenism was considered likely
because of the severe splenomegaly, marked splenic
EMH, erythrophagocytosis, and peripheral cytopenias
(anemia and thrombocytopenia).
Bone marrow aspirates showed marked erythroid
hyperplasia, moderate to marked megakaryocytic
hyperplasia, mild myeloid hyperplasia, and focal
plasmacytosis. A small amount of abdominal fluid
was obtained, and although the volume was insufficient
for complete analysis, cytologic evaluation suggested
mild suppurative inflammation. No erythrophagocyto-
sis or hemosiderophages were seen. Neoplastic cells
were not identified in the peripheral blood, bone
marrow, or abdominal fluid samples.
Clinical outcome
Approximately 48 hours after presentation, the dog’s
condition worsened. The dog became extremely de-
pressed and body temperature increased to 105.28F. The
PCV declined to 15.2%. Possible explanations for the
progressive anemia included hemorrhage, immune-
mediated disease, and hypersplenism. In addition, the
platelet count declined to 15,000/lL on day 3. The cause
of the thrombocytopenia was likely multifactorial and
attributed to disseminated intravascular coagulation
(DIC), hypersplenism, or secondary immune-mediated
disease.
Because of the rapid decline in the dog’s condi-
tion and the markedly enlarged spleen, exploratory
laporatomy and splenectomy were performed. The dog
was supported before and during surgery with whole
blood transfusions (total blood transfused 750 mL), and
the PCV remained at around 28%. Gross abnormalities
noted at surgery included an enlarged nodular spleen
and asymmetrical hepatomegaly. The spleen and biop-
sies of the liver, jejunum, and mesenteric lymph node
were submitted for histologic evaluation.
The day after surgery the dog became severely
dyspneic. On the basis of thoracic radiographs, pulmo-
nary thromboembolism was presumed. The dog failed
to respond to treatment and was euthanized. A
necropsy was not performed.
Vol. 33 / No. 2 / 2004
 Cienava, Barnhart, Brown, Mansell, Dunstan, Credille

scattered megakaryocytes were observed. Erythropha-

gocytosis was rare and macrophages occasionally con-
tained hemosiderin.
In multiple sections of liver, the sinusoids and veins
contained a mixed population of cells similar to those
found in the spleen. The majority of the mixed popu-
lation consisted of large neoplastic round cells with
pleomorphic and vesicular nuclei, multiple prominent
nucleoli, and scant cytoplasm (Figure 2B). A second
population of round cells consistent with histiocytes,
numerous nucleated erythrocytes, and occasionally,
megakaryocytes, were identified. Kupffer cells were
frequent and prominent.
Immunohistochemistry was performed on forma-
lin-fixed sections of spleen and liver, and, in both
tissues, the population of neoplastic round cells had
strong, diffuse cytoplasmic staining for CD3 (Figure 3A
and B). Neoplastic round cells in the liver and spleen
did not stain for CD79a or lysoszyme. The large
histiocytic cells stained positive for lysozyme and were
negative for CD3 and CD79a.

Figure 1. Fine-needle aspirate of spleen. A pleomorphic population of

large round atypical neoplastic cells can be seen. Diff-Quik, (A) 360
objective, (B) 3100 objective.

Histopathology

The spleen measured approximately 10 3 4 cm and

multiple sections of the spleen were examined histo-
logically. Numerous aggregates of fibrin and diffuse
hemorrhage were present throughout the splenic
parenchyma, markedly distending the sinuses. Focal
thrombosis of a large artery was noted. Only small
periarteriolar lymphoid sheaths remained. The splenic
sinuses were diffusely infiltrated by large, round to
oval, neoplastic cells (Figure 2A). The cells had large
pleomorphic nuclei with single or multiple prominent
basophilic nucleoli and an average of 5 mitotic figures
per high-power field. Occasionally, atypical mitotic
figures were noted. Some of the neoplastic cells were
binucleated. A second population consisted of large,
round to oval cells with reniform nuclei and abundant Figure 2. Histologic sections of spleen and liver. Spleen (A). Atypical
neoplastic round cells (arrows) are distending the splenic sinuses. Liver
amphophilic foamy cytoplasm, consistent with benign
(B). Neoplastic round cells within hepatic sinusoids are similar to those
histiocytes. Moderate numbers of erythroid cells and found in the spleen. Hematoxylin and eosin, 340 objective.

Vol. 33 / No. 2 / 2004 Veterinary Clinical Pathology Page 107

 Hepatosplenic T-Cell Lymphoma in a Dog
Figure 3. Immunohistochemical staining of spleen (A) and liver (B) for
CD3. Note the strong positive diffuse cytoplasmic staining of the large
atypical neoplastic cells in both tissues. 320 objective.
The mesenteric lymph node had lymphoid deple-
tion with medullary sinus histiocytosis and erythro-
phagocytosis. No neoplastic cells were identified in the
lymph node. Jejunal biopsies were unremarkable.
On the basis of the morphologic and immunohis-
tochemical evaluation, a diagnosis of a hepatosplenic T-
cell lymphoma with marked EMH was made.
Gene rearrangement
T-cell clonality was assessed by canine-specific, poly-
merase chain reaction (PCR)-based techniques. PCR
amplification of DNA isolated from splenic aspirates of
the patient was performed as described1 using specific
primers obtained from Dr. Anne Avery (Colorado State
University, Fort Collins, CO). A single distinct band was
obtained with the T-cell–specific primer set, indicating
T-cell clonality and confirming the diagnosis of T-cell
lymphoma (Figure 4). A faint smear was obtained with
the B-cell–specific primer set, indicating a residual
population of reactive B-lymphocytes.
Page 108 Veterinary Clinical Pathology
Discussion
Lymphoma is the most common neoplasm of the
hematopoietic system in dogs and has an annual
incidence of approximately 33 cases per 100,000 dogs.2
The etiology of lymphoma remains unclear, but factors
such as immune dysfunction, chemical carcinogens,
chromosomal abnormalities, and viral etiologies have
been explored. Classification of lymphoproliferative
disease is challenging and has undergone significant
revision as new diagnostic techniques have become
available. Immunophenotyping has improved our
ability to distinguish between canine B- and T-cell
lymphomas and has lead to more-specific classification
schemes. Because of variations in biologic behavior,
differentiating between B- and T-cell lymphomas can
have significant prognostic implications. Dogs with
T-cell lymphoma have a lower complete response to
chemotherapy and shorter remission and survival times
when compared with dogs with B-cell tumors.3,4
According to a recent morphologic and immunologic
study of 140 cases of canine lymphoma over a 4-year
period, the total incidence of T-cell lymphoma was
32.8% (46/120 cases).5 All 46 cases of T-cell lymphoma
expressed a CD3þ CD79a- phenotype.5
Unlike typical cases of canine T-cell lymphoma, the
clinical progression of disease in this dog was aggres-
sive and acute and characterized by high fever. The
most severe hematologic abnormalities in this case were
anemia and thrombocytopenia. The exact cause of
anemia was unclear; blood loss secondary to thrombo-
cytopenia, hypersplenism, and immune-mediated de-
struction were considered, although there was no strong
evidence to support the latter. Cytokine-induced bone-
marrow suppression secondary to neoplasia also may
have contributed. Thrombocytopenia has been reported
in 30–50% of dogs with lymphoma and usually is
caused by immune-mediated destruction or bone-
marrow suppression caused by inhibitory cytokines
(ie, interferon gamma) produced by neoplastic T
lymphocytes.6 The low platelet count in this case was
likely because of splenic sequestration and possible
platelet consumption. Lymphocytosis and lymphopenia
are seen with about equal frequency in canine lympho-
mas.7 The mild left shift in this case was attributed to
inflammation and tissue necrosis caused by prolifera-
tion and infiltration of tumor cells. Neutropenia may
have been related to increased tissue demand as a result
of splenic sequestration.
An unusual aspect of this case was the restricted
anatomic distribution of the tumor to spleen and liver,
with marked splenomegaly. Most dogs with multi-
centric lymphoma have bilateral peripheral lymphade-
nopathy.3 Another unusual feature of this case was the
Vol. 33 / No. 2 / 2004
 Cienava, Barnhart, Brown, Mansell, Dunstan, Credille
Figure 4. Polymerase chain reaction–based assay for B- and T-cell
clonality using DNA isolated from splenic aspirates. Cmu is a positive
internal control, which amplifies a 260 base-pair segment of the constant
region of immunoglobulin M. A water blank was used as a negative
control (data not shown). A single distinct band located at approximately
90 base pairs in the T-cell lane indicates T-cell clonality. A faint smear
centered on 130 base pairs in the B-cell lane indicates a residual
population of reactive B-cells.
bizarre, pleomorphic morphology of the neoplastic
cells. Because of cell morphology, malignant histiocy-
tosis was a differential diagnosis; although the neo-
plastic cells were not phagocytic, cytophagia is not
a consistent feature of malignant histiocytosis.8 Malig-
nant histiocytosis can be difficult to distinguish from
large, pleomorphic T-cell lymphomas, such that addi-
tional diagnostic testing, including immunohistochem-
istry and molecular analysis, is needed to confirm a
T-cell neoplasm.
Molecular testing now is considered the gold
standard for determining B- and T-cell clonality in
dogs.1,9 Clonality assays may be used in conjunction
with cytology, histology, and immunophenotyping and
can be performed on biopsy samples, cavity fluids, fine-
needle aspirates, bone marrow, and peripheral blood.
Other applications of clonality testing include determi-
nation of B- or T-cell lineage, staging of lymphoma, and
detection of residual disease after therapy.1 Clonality in
canine lymphoma is demonstrated through PCR am-
Vol. 33 / No. 2 / 2004 Veterinary Clinical Pathology
plification of unique somatic recombination rearrange-
ments that occur in the T-cell receptor (TCR) gene for
T-lymphocytes and in the immunoglobulin heavy chain
gene for B-lymphocytes.1,10 The rearrangements occur
by excision and insertion of a variable number of
random nucleotides between the V, D, and J segments,
resulting in the production of many unique proteins.1
Because non-neoplastic lymphoid populations contain
a variety of lymphocyte clones, PCR amplification of the
VDJ region generates many different-sized products,
which appear as a broad smear on gel electrophoresis.
In contrast, neoplastic lymphoid populations are mono-
clonal in most cases, and PCR amplification results in
a single distinct band, as seen in this case. Burnett et al1
demonstrated high sensitivity for the canine clonality
assay, with clonal rearrangement detected when 0.1–
10% of DNA was derived from neoplastic cells, depend-
ing on the tissue. In humans, PCR-based assays detect
clonal TCR rearrangements that account for as little as
0.1–1% of all DNA in the sample.11,12
A small percentage (9%) of canine lymphomas
may give false negative clonality assay results,
possibly due to failure of the primers to bind, deleted
antigen receptor genes, or NK-derived lymphomas
that do not have rearranged antigen receptor genes.1
False-positive clonal T-cell receptor rearrangements
have been found in several dogs with positive E canis
titers. The T-cells in these dogs may be transformed T-
cells that have not yet progressed to a clonal
population, or may be an antigen-specific population
that will eventually recede.
A rare syndrome termed hepatosplenic cd T-cell
lymphoma in human patients shares many features with
the present case, including clinical presentation, hema-
tologic abnormalities, anatomic distribution, morpho-
logic features of the neoplastic cells, and the aggressive
nature of the disease. According to the Revised
European–American Lymphoma (REAL) classification,
hepatosplenic cd T-cell lymphoma is a form of peripheral
T-cell lymphoma that most likely arises from cd T-cells in
the splenic red pulp.13 The syndrome is characterized by
predominant involvement of the liver and spleen with
minimal lymphadenopathy, and it often follows an
aggressive course.13,14 Bone marrow involvement, if
present, is minimal. Most cases affect young adult males,
and patients invariably die within 2 years of diagnosis
despite therapeutic measures.13 Clinical signs in humans
include weight loss, decreased appetite, fatigue, fever,
and night sweats.13,14 Physical examination typically
reveals hepatosplenomegaly with no peripheral lymph-
adenopathy. Clinical laboratory data often includes
anemia and thrombocytopenia with unremarkable
serum chemistry results. Anemia and thrombocytopenia
usually are caused by hypersplenism or bone marrow
Page 109
 Hepatosplenic T-Cell Lymphoma in a Dog
infiltration.13 In one unusual case of hepatosplenic cd
T-cell lymphoma, the patient developed immune-
mediated hemolytic anemia and thrombocytopenia.13
The first confirmed case of canine hepatosplenic
cd T-cell lymphoma recently was reported.6 The
diagnosis was confirmed by histopathology and
immunohistochemistry, which revealed cd T-cell re-
ceptor expression by neoplastic T lymphocytes.6 Like
the dog in this report, the dog with confirmed
hepatosplenic cd T-cell lymphoma presented with an
aggressive disease that included lethargy, fever, and
splenomegaly. The dog also was anemic and throm-
bocytopenic, had infiltration of the spleen and liver
with neoplastic T-cells, and lacked peripheral lymph-
adenopathy. In contrast to the dog in this report, the
dog with confirmed hepatosplenic cd T-cell lympho-
ma had bone marrow involvement. It was not
possible to definitively determine whether the T-cell
lymphoma in the present case consisted of pro-
liferating cd T-cells because frozen tissue samples
were not collected for immunophenotypic analysis.
However, the parallel between the rare entity of
hepatosplenic cd T-cell lymphoma in humans and the
clinical and anatomic distribution of the T-cell
lymphoma described in our patient is compelling.
Acknowledgment
The authors thank Dr Anne Avery (Colorado State University,
Fort Collins, CO) for the T-cell and B-cell primer sets used for the
PCR amplification of DNA in this study.
References
1. Burnett R, Vernau W, Modiano J, et al. Diagnosis of canine
lymphoid neoplasia using clonal rearrangements of antigen
receptor genes. Vet Pathol. 2003;40:32–41.
Page 110 Veterinary Clinical Pathology
2. Teske E. Canine malignant lymphoma: a review and
comparison with human non-Hodgkin’s lymphoma. Vet Q.
1994;4:209–219.
3. Meuten DJ. Tumors in Domestic Animals. 4th ed. Ames, IA:
Iowa State University Press; 2002:128–144, 169.
4. Greenlee PG, Filippa DA, Quimby F, et al. Lymphomas in
dogs. A morphologic, immunologic and clinical study. Cancer.
1990;66:480–490.
5. Fournel-Fleury C, Ponce F, Felman P, et al. Canine T-cell
lymphomas: a morphological, immunological, and clinical
study of 46 new cases. Vet Pathol. 2002;39:92–109.
6. Fry MM, Vernau W, Pesavento PA, Brömel C, Moore PF.
Hepatosplenic lymphoma in a dog. Vet Pathol. 2003;40:
556–562.
7. Madewell BR. Hematological and bone marrow cytological
abnormalities in 75 dogs with malignant lymphoma. J Am
Anim Hosp Assoc. 1986;22:235–240.
8. Raskin R, Meyer D. Atlas of Canine and Feline Cytology. 1st ed.
Philadelphia, PA: WB Saunders Co; 2001:108–120, 180–182.
9. Ioachim H, Ratech H. Ioachim’s Lymph Node Pathology. 3rd ed.
Philadelphia, PA: Lippincott Williams & Wilkins; 2002.
10. Vernau W, Moore PF. An immunophenotypic study of canine
leukemias and preliminary assessment of clonality by poly-
merase chain reaction. Vet Immunol Immunopathol. 1999;69:
145–165.
11. Bachelez H. The clinical use of molecular analysis of clonality
in cutaneous lymphocytic infiltrates. Arch Dermatol. 1999;135:
200–202.
12. Bourguin A, Tung R, Galili N, et al. Rapid, nonradioactive
detection of clonal T-cell receptor gene rearrangements in
lymphoid neoplasms. Proc Natl Acad Sci U S A. 1990;87:8536–
8540.
13. Motta G, Vianello F, Menin C, et al. Hepatosplenic gamma-
delta T-cell lymphoma presenting with immune-mediated
thrombocytopenia and hemolytic anemia (Evan’s Syndrome).
Am J Hematol. 2002;69:272–276.
14. Nosari A, Oreste PL, Biondi A, et al. Hepatosplenic cd T-cell
lymphoma: a rare entity mimicking the hemophagocytic
syndrome. Am J Hematol. 1999;60:61–65.
Vol. 33 / No. 2 / 2004