00_06_0259.

QXD 9/30/2005 2:55 PM Page 719

Primary T-cell lymphoma of the central

SMALL ANIMALS/
nervous system in a dog

EXOTICS
Sam N. Long, BVSc; Pamela E. J. Johnston, BVM&S, PhD; T. James Anderson, BVM&S, PhD

anisocytosis and anisokaryosis (Fig 1). Most of these

cells had a round hyperchromatic nucleus and a mod-
' Primary T-cell lymphoma is a rare form of CNS neo- erate amount of basophilic cytoplasm that in some
plasia. cells was finely vacuolated. Some cells appeared to be
' Diagnosis may be aided by use of cytologic exam- lymphoid, whereas others appeared to be
ination of CSF. macrophages. Although there was no direct evidence of
' Primary CNS T-cell lymphoma should be considered malignancy, neoplasia could not be excluded as a pos-
in a patient with multiple cranial nerve abnormali- sibility. However, on the basis of the lack of a mass, as
ties, even if results of imaging studies are consid- indicated by results of CT scanning, it was felt that an
ered normal. inflammatory process was more likely, and a tentative
diagnosis of encephalitis of unknown cause was made.

A 5-year-old sexually intact female Labrador

Retriever was referred to the Glasgow University
Veterinary Hospital with a 5-day history of obtunda-
Treatment for neurologic signs was initiated by
administration of prednisone (0.5 mg/kg [0.23 mg/lb] of
body weight, PO, q 12 h). Treatment for signs of nausea
tion, vomiting, left-sided head tilt, facial asymmetry, and vomiting was initiated by administration of acepro-
and ataxia. Physical examination revealed respiratory mazine (0.01 mg/kg [0.005 mg/lb], SC, q 6 h) and meto-
rate within reference range and signs of depression. clopramide (1 mg/kg [0.46 mg/lb], IV, q 24 h).
Rectal temperature was 38.8 C (101.8 F). Thoracic aus- Treatment for the presumptive aspiration pneumonia
cultation revealed increased respiratory sounds, partic- was initiated with metronidazole (10 mg/kg [4.5 mg/lb],
ularly in the cranioventral lung fields, without inspira- IV, q 12 h) and amoxycillin-clavulanic acid (20 mg/kg
tory or expiratory stridor. Excessive pooling of saliva [9.1 mg/lb], IV, q 6 h). In addition, a gastrostomy tube
and ptyalism was detected. A full ophthalmic examina- was placed, and parenteral feeding with a commercial
tion was performed and did not reveal abnormalities. concentration dietb was initiated. Despite treatment, the
Neurologic examination revealed vestibular ataxia with clinical signs and general demeanour of the dog contin-
staggering to the left side, but proprioceptive deficits ued to deteriorate, and 5 days after admission the dog
were not detected. Cranial nerve examination con- had right-sided vestibular signs and facial nerve paraly-
firmed a left-sided head tilt, horizontal nystagmus with sis in addition to previous signs. At this point, the own-
the fast phase to the right, left-sided facial nerve paral- ers chose to have the dog euthanatized.
ysis, and bilaterally decreased gag reflex. On the basis On gross postmortem, the dog was found to have
of neurologic findings, a lesion within the left side of a dilated esophagus. The cranioventral portions of the
the caudal fossa of the calvarium or polyneuropathy lungs on both sides were red-purple, firm, and
affecting cranial nerves was considered likely. atelectatic. All other tissues, including brain and spinal
Results of routine hematologic and biochemical cord, appeared grossly normal. The brain and spinal
analyses and serologic tests for Toxoplasma and cord were removed and fixed in neutral-buffered 10%
Neospora spp were unremarkable. Evaluation of tho- formalin. Specimens of spleen, kidney, liver, lymph
racic radiographs revealed megaesophagus; the dilated nodes, tonsil, mammary gland, esophagus, lung, vagus
portion of the esophagus extended from the thoracic nerve, and left recurrent laryngeal nerve were obtained
inlet to the esophageal hiatus. A mixed alveolar-inter- and fixed in neutral-buffered 10% formalin. Bone mar-
stitial pattern was also evident in the cranioventral row specimens were not obtained.
region of the lungs, consistent with aspiration pneu- Histologic sections of all tissues were stained with
monia of several days’ duration. Results of computed H&E and examined. Immunohistochemical staining
tomography (CT) scanning of the skull before and was performed on paraffin-embedded sections by use
after injection of iodinated contrast materiala were
of a modified avidin-biotin-peroxidase method.1
unremarkable. Analysis of CSF obtained via cerebel-
Immunohistochemical staining was performed with
lomedullary cistern tap revealed WBC count of 46
antibodies against B-cell antigen (CD79a), T-cell anti-
cells/µl (reference range, 0 to 5 cells/µl) and protein
gen (CD3), IgG, IgM, IgA, and macrophage antigen
concentration of 1,060 mg/L (reference limit,
(MAC337). Stains were used on formalin-fixed, paraf-
< 250 mg/L). Cytologic examination of the CSF
revealed a population of round cells with substantial fin-embedded sections of brain and spinal cord.
Human tonsil tissue and specimens stained without
primary antibody were used as positive and negative
From the Departments of Small Animal Clinical Studies (Long,
Anderson) and Histopathology (Johnston), University of Glasgow
controls, respectively.
Veterinary School, Bearsden Rd, Bearsden, Glasgow G61 1QH, Brain and spinal cord were examined at multiple
Scotland, UK. levels. Abnormalities were found only in those sections
The authors thank Alma Jenkins for technical assistance. obtained from the area of the middle and caudal cere-

JAVMA, Vol 218, No. 5, March 1, 2001 Scientific Reports: Clinical Report 719
00_06_0259.QXD 9/30/2005 2:55 PM Page 720

SMALL ANIMALS/
EXOTICS

Figure 1—Photomicrograph of a centrifuged cytologic prepara- Figure 3—Photomicrograph of a dorsal nerve root of the first
tion of CSF of a dog with T-cell lymphoma. Notice pleomorphic cervical spinal cord segment of a dog with T-cell lymphoma.
cells. May-Grünwald giemsa stain; bar = 25 µm. Notice selective infiltration of nerve root without invasion of the
spinal cord parenchyma. H&E stain; bar = 500 µm.

level of the first spinal cord segment, there was focal

infiltration of neoplastic cells into the dorsal funiculus
from the dorsal horn of the first spinal nerve.
Neurons in the nucleus of cranial nerve V on the
left side of the brain at the level of the caudal cerebel-
lar peduncle had moderate chromatolysis. Some chro-
matolytic neurons were also seen in the nuclei of cra-
nial nerves VII and VIII and in the nucleus of the soli-
tary tract but only on the left side of the brain. The
nucleus of the solitary tract on the left side also had
moderate gliosis, without evidence of infiltration by
neoplastic cells. Vacuolation within myelin sheaths of
the afferent tracts of cranial nerve VIII was seen in
some locations, only on the left side of the brain stem.
Immunohistochemical staining revealed strong
Figure 2—Photomicrograph of a section of the choroid plexus of
the fourth ventricle of a dog with T-cell lymphoma. Notice infil- staining of all neoplastic cells by use of antiCD3 anti-
tration of neoplastic cells with medium-sized to large pleomor- body (Fig 4). Staining was not detected by use of
phic central nuclei with coarsely stippled chromatin. H&E stain; antiCD79a, antiMAC337, antiIgG, antiIgM, or antiIgA
bar = 50 µm.
antibodies. A diagnosis of primary CNS T-cell lym-
bellar peduncles, lateral apertures of the fourth ventri- phoma was established.
cle, and the first 3 cervical spinal cord segments. Sections from the cranioventral areas of the lungs
Microscopic examination revealed a diffuse cellular had alveoli and bronchioles that were filled with fibri-
infiltrate in the leptomeninges, choroid plexus of the
fourth ventricle, nerve roots of cranial nerves V, VII,
VIII, IX, X and XI, and the dorsal and ventral nerve
roots of the first 3 cervical spinal nerves (Fig 2 and 3).
The infiltrate was composed of neoplastic round cells
that each had a moderate amount of basophilic cyto-
plasm that was sometimes finely vacuolated. Cells had
medium-sized to large pleomorphic central nuclei with
coarsely stippled chromatin and prominent nucleoli.
Cell size varied between 7 and 10 µm. There was mod-
erate to substantial anisokaryosis. The mitotic index
was low.
Within nerve roots, the infiltrate was composed of
densely packed sheets of cells, whereas within the
meninges, the cells were more sparsely arranged. On the
left side of the brain stem the nerve roots and choroid
plexus were more heavily infiltrated than on the right
Figure 4—Photomicrograph of a section of the choroid plexus
side, except for the nerve root of cranial nerve X, which (inset, negative control tissue) of a dog with T-cell lymphoma.
was infiltrated on both sides to a similar degree. At the Immunohistochemical antiCD3 stain; bar = 100 µm.

720 Scientific Reports: Clinical Report JAVMA, Vol 218, No. 5, March 1, 2001
00_06_0259.QXD 9/30/2005 2:55 PM Page 721

nous material and high numbers of degenerate cells. chemical studies indicate that many cases of neoplastic

SMALL ANIMALS/
There was extensive necrosis of alveolar walls. A diag- reticulosis in dogs are in fact B-cell lymphomas.12,13
nosis of fibrinonecrotic bronchopneumonia consistent Results of a recent study indicate that, in contrast,

EXOTICS
with aspiration pneumonia was made. Neoplastic cells T cells are the predominant lymphocyte found in
were not seen in the lungs or other nonneural tissues. lesions of GME, suggesting T-cell mediated delayed-
Central nervous system tumors are a fairly com- type hypersensitivity as a possible pathogenic mecha-
mon cause of neurologic dysfunction in animals; an nism.11 However, other researchers have theorized that
incidence of 14.5 intracranial neoplasms/y/100,000 GME involves an initial inflammatory process that may
dogs at risk has been reported.2 Primary CNS lym- be transformed to a neoplastic process, or alternatively
phoma may involve the CNS without systemic disease.3 that GME may represent an immune response to a neo-
In humans, most primary CNS lymphomas are of B-cell plastic process.14
origin, whereas primary T-cell lymphoma is exceeding- The dog reported here was referred with clinical
ly rare.4 Findings are similar in dogs and cats; to the signs related primarily to cranial nerve dysfunction.
authors’ knowledge, only a single case in a cat has been The clinical signs induced by CNS tumors are general-
reported.5 Until more reports with immunohistochem- ly nonspecific and depend on the location of the lesion.
ical verification are published, however, the true pro- Signs associated with CNS tumors include seizures,
portion of CNS lymphomas that are derived from T cranial nerve deficits, paresis, changes in behavior, cir-
cells will remain speculative. cling, and endocrine dysfunction. Animals with diffuse
Lymphoma within the CNS develops most com- leptomeningeal metastasis of nonlymphoid tumors are
monly as part of a multicentric or systemic process and more likely to have clinical signs mimicking meningi-
only rarely as a primary tumor.5-7 In cats and humans, tis (pyrexia, seizures, altered consciousness, signs of
5% of patients with systemic lymphoma have CNS neck pain, and lethargy).15 However, in humans, cra-
involvement, predominantly of the leptomeninges.3 In nial nerve deficits are commonly seen with metastasis
contrast, CNS lymphoma in dogs seems to rarely affect of carcinomas to the meninges.15 Cranial nerve deficits
the leptomeninges.8,9 In the dog described here, the have also been reported in association with lymphoma
CNS lymphoma was likely a primary tumor, because in dogs, either because of direct infiltration of nerves16
neoplastic cells were confined to the CNS. This case is or as part of a paraneoplastic polyneuropathy.17,18 In 1 of
also unusual, because the tumor was predominantly these dogs, bilateral trigeminal nerve deficits were the
located in the leptomeninges and cranial and cervical referring signs, and patchy neoplastic infiltration of
nerve roots, without invasion into the brain or spinal lymphoid cells was identified along the meninges at
cord parenchyma. various sites, although no discrete foci of tumor cells
Immunohistochemical evaluation of lymphomas were identified.17 Other tumors of leukocyte origin
allows them to be classified as derived from T cells or cause cranial nerve deficits, usually because of neo-
B cells. In humans, primary CNS T-cell lymphomas plastic infiltration of the affected nerves.19-21 In addi-
have a propensity for involvement of the lep- tion, neurologic signs have been reported with malig-
tomeninges alone.4 A similar pattern of neoplastic infil- nant angioendotheliomatosis.22,23 Similar to the case
tration was observed in the dog described here. reported here, vestibular disease and facial paresis have
The principal immunocytochemical characteristic been reported with meningiomas of the cerebellopon-
that differentiates B-cell lymphoma from T-cell lym- tine angle.24 In the dog described here, the cranial
phoma is the expression of cytoplasmic immunoglobu- nerve deficits were likely the result of infiltration of the
lin of 1 light-chain type. Ideally, B-cell tumors should nerves themselves, and the relatively mild histologic
express IgM or IgG and κ or λ light chains.3 In humans, changes seen in the brain support a theory of neuronal
diagnosis of tumor type is aided by polymerase chain and white matter degeneration secondary to axonal
reaction confirmation of the monoclonal nature of neo- loss. Clinical signs indicated that lesions developed on
plastic cells. Lymphocyte surface receptor genes may 1 side of the brain stem before the other, which corre-
aid in the diagnosis and definitive classification of lym- lated well with the asymmetric pathologic findings. It
phoma in dogs, and recently a family of Vβ T-cell is also interesting that vestibular signs in this dog were
receptor genes have been identified that have been predominantly peripheral in nature, consistent with
used to detect T-cell lymphoma by use of the poly- the primary involvement of cranial nerves and relative
merase chain reaction.10 sparing of the brainstem.
Lymphoma may be confused with other CNS dis- Diagnosis in the dog reported here was complicat-
eases, particularly inflammatory conditions such as ed by the equivocal CSF findings and the apparently
granulomatous meningoencephalitis (GME). Gen- normal results of CT scanning. Leptomeningeal or
erally, lymphoma cells are less differentiated and more dural metastasis may cause linear enhancement in
isomorphic than those seen in dogs with GME.3 In areas that correlate anatomically with the meninges, as
dogs with GME, lesions are localized predominantly in viewed after injection of contrast material, but may be
the white matter of brain and spinal cord but may difficult to detect with CT.25 Leptomeningeal or dural
involve the leptomeninges.11 Granulomatous menin- metastasis is better detected by use of magnetic reso-
goencephalitis was previously known as inflammatory nance imaging because of its superior soft-tissue detail;
reticulosis, part of a spectrum of reticuloses that were however, abnormal enhancement of the meninges may
divided into neoplastic and inflammatory forms. also occur with a variety of other pathologic processes,
Lymphoma now replaces neoplastic reticulosis as a including infection, inflammation, irritation, and
more appropriate term, and results of immunohisto- spontaneous intracranial hypotension.21 Conversely,

JAVMA, Vol 218, No. 5, March 1, 2001 Scientific Reports: Clinical Report 721
00_06_0259.QXD 9/30/2005 2:55 PM Page 722

imaging studies of some diffuse meningeal tumors 7. Lefbom BK, Parker GA. Ataxia associated with lymphosar-
SMALL ANIMALS/

have not detected lesions.26 coma in a dog. J Am Anim Hosp Assoc 1995;207:922–923.
8. Britt JO Jr, Simpson JG, Howard EB. Malignant lymphoma
Diagnosis of meningeal neoplasia is often made on
EXOTICS

of the meninges in two dogs. J Neurooncol 1984;94:45–53.

the basis of CSF analysis. Results are usually reliable 9. Couto CG, Cullen J, Pedroia V, et al. Central nervous sys-
when leptomeningeal involvement is present, especial- tem lymphosarcoma in the dog. J Am Vet Med Assoc 1984;184:
ly for lymphoreticular tumors,9,15,26,27 and animals with 809–813.
lymphoma involving the leptomeninges will generally 10. Dreitz MJ, Ogilvie G, Sim GK. Rearranged T lymphocyte
have atypical plasmacytoid mononuclear cells in the antigen receptor genes as markers of malignant T cells. Vet Immunol
Immunopathol 1999;69:113–119.
CSF.15 However, pleomorphic leukocytes in CSF may 11. Kipar A, Baumgartner W, Vogl C, et al. Immuno-
complicate the differentiation of primary CNS inflam- histochemical characterization of inflammatory cells in brains of
matory disease from CNS neoplasia.28 In humans, dogs with granulomatous meningoencephalitis. Vet Pathol 1998;
immunohistochemical staining of CSF aids the differ- 35:43–52.
entiation of neoplasia from inflammatory CNS dis- 12. Vandevelde M, Fatzer R, Frankhauser R. Immuno-
eases. Unfortunately, this was not available in the case histochemical studies on primary reticulosis of the canine brain. Vet
Pathol 1981;18:577–588.
reported here. If the correct diagnosis had been avail- 13. Fenner WR. Diseases of the brain. In: Ettinger SJ, Feldman
able, the use of corticosteroids could have been avoid- EC, eds. Textbook of veterinary internal medicine. 5th ed.
ed, because corticosteroids would have influenced the Philadelphia: WB Saunders Co, 2000;553–561.
efficacy of further treatment. 14. Vandevelde M. Primary reticulosis of the central nervous
To the authors’ knolwedge, this is the first case system. Vet Clin North Am Small Anim Pract 1980;10:57–63.
report of a primary T-cell lymphoma involving the lep- 15. Fenner WR. Metastatic neoplasms of the central nervous
system. Semin Vet Med Surg (Small Anim) 1990;5:253–261.
tomeninges and the cranial and cervical nerve roots in 16. Pfaff AD, March PA, Fishman C. Acute bilateral trigeminal
a dog. T-cell lymphoma should be considered as a dif- neuropathy associated with nervous system lymphosarcoma in a
ferential diagnosis if cranial nerve deficits are evident dog. J Am Anim Hosp Assoc 2000;36:57–61.
and abnormal pleomorphic mononuclear cells are 17. Hobbs SL, Cobb MA. A cranial neuropathy associated with
detected in the CSF. Immunohistochemical staining of multicentric lymphosarcoma in a dog. Vet Rec 1990;127:525–526.
cells in the CSF may help to differentiate lymphoma 18. Presthus J, Teige J Jr. Peripheral neuropathy associated with
lymphosarcoma in a dog. J Small Anim Pract 1986;27:463–469.
from inflammatory CNS conditions and aid in the early 19. Carpenter JL, King NW, Abrams KL. Bilateral trigeminal
diagnosis and treatment of this disease. nerve paralysis and Horner’s syndrome associated with myelomono-
cytic neoplasia in a dog. J Am Vet Med Assoc 1987;191:1594–1596.
a
Conray 420, Sodium Iothalamate, Mallinckrodt Medical (UK) Ltd, 20. Christopher MM, Metz AL, Klausner J, et al. Acute
Northampton, England. myelomonocytic leukemia with neurologic manifestations in the
b
Hill’s a/d Prescription Diet, Hill’s Pet Nutrition Ltd, Hatfield, dog. Vet Pathol 1986;23:140–147.
England. 21. Vernau KF, Terio RA, LeCouteur RA, et al. Acute B-Cell
lymphoblastic leukemia with meningeal metastasis causing prima-
ry neurologic dysfunction in a dog. J Vet Intern Med 2000;14:
References 110–115.
1. Hsu SM, Raine L, Fanger H. Use of avidin biotin peroxidase 22. Summers BA, de Lahunta A. Cerebral angioendothe-
complex (ABC) in immunoperoxidase techniques: a comparison liomatosis in a dog. Acta Neuropathol (Berl) 1985;68:10–14.
between ABC and unlabeled antibody (PAP) procedures. J Histochem 23. Dargent FJ, Fox LE, Anderson WL. Neoplastic angioen-
Cytochem 1981;29:577–580. dotheliomatosis in a dog: an angiotropic lymphoma. Cornell Vet
2. Gavin PR, Fike JR, Hoopes PJ. Central nervous system 1988;78:253–262.
tumors. Semin Vet Med Surg (Small Anim) 1995;10:180–189. 24. Fischer A. What is your neurologic diagnosis? J Am Vet Med
3. Johnson GC. Genesis and pathology of tumors of the ner- Assoc 1994;205:687–689.
vous system. Semin Vet Med Surg (Small Anim) 1990;5:210–222. 25. Kraft SL, Gavin PR. Intracranial neoplasia. Clin Tech Small
4. Graham DI, Lantos PL. Primary T cell lymphomas. In: Anim Pract 1999;14:112–123.
Graham DI, Lantos PL, eds. Greenfield’s neuropathology. 6th ed. New 26. Bagley RS. Common neurologic diseases of older animals.
York, Oxford University Press, 1997;773–774. Vet Clin North Am Small Anim Pract 1997;27:1451–1486.
5. Fondevila D, Vilafranca M, Pumarola M. Primary central ner- 27. Rosin A. Neurologic disease associated with lymphosarco-
vous system T cell lymphoma in a cat. Vet Pathol 1998;35:550–553. ma in ten dogs. J Am Vet Med Assoc 1982;181:50–53.
6. Summers BA, Cummings JF, de Lahunta A. Principles of 28. Carrillo JM, Sarfaty D, Greenlee P. Intracranial neoplasm
neuropathology. In: Summers BA, Cummings JF, de Lahunta A, eds. and associated inflammatory response from the central nervous sys-
Veterinary neuropathology. 1st ed. St Louis: Mosby, 1995;1–54. tem. J Am Anim Hosp Assoc 1986;22:367–373.

722 Scientific Reports: Clinical Report JAVMA, Vol 218, No. 5, March 1, 2001