Abstract
A primary intracerebral plasmacytoma was identified in a 7-year-old spayed female Boston Terrier. Grossly, a well-demarcated, 2 cm in diameter, roughly spherical tumor was in the rostral aspect of the left cerebral hemisphere. Histologically, the neoplasm was composed of sheets of round cells with distinct plasmacytoid features and marked anisocytosis and anisokaryosis. Cells were positive for vimentin, CD18, CD79a, and lambda light-chain, and negative for kappa light chain, cytokeratin, lysozyme, glial fibrillary acidic protein, and S100 protein. Clonally rearranged B-cell antigen receptor genes were detected by PARR (polymerase chain reaction for antigen receptor rearrangements), confirming clonal proliferation of B lymphocytes. Although primary solitary intracerebral plasmacytoma is rare in dogs and other species, it should be included in the differential diagnosis for central nervous system round-cell neoplasms. Clonality testing can be utilized to support the histological diagnosis of this neoplasm type.
A 7-year-old, spayed female Boston Terrier was presented to the emergency service of North Carolina State University Veterinary Teaching Hospital (NCSU-VTH) with a 1-week history of recurrent grand mal seizures, head pressing and circling, and unsuccessful symptomatic treatment with potassium bromide, phenobarbital, diazepam, and ciprofloxacin. Upon presentation, the dog was ataxic and alternated between periods of depression and alertness. A neurologic examination localized the lesion in the left cerebral hemisphere. During the night, the dog became bradycardic, decompensated, and was euthanatized.
At necropsy, the rostral two thirds of the left cerebral hemisphere was enlarged with flattening of gyri, consistent with brain edema and swelling. Within the left frontal lobe, involving the cortex and underlying white matter, there was a tan, soft, moderately well-demarcated, approximately 2 cm in diameter, roughly spherical mass that displaced the longitudinal fissure and contacted the overlying meninges over an approximately 1 cm × 1 cm area. No significant gross lesions were observed in other organs at necropsy. Tissues were fixed in 10% neutral buffered formalin and processed routinely for histopathology.
Histologically, the cerebral cortex contained a nonencapsulated, well-demarcated, expansile, moderately cellular mass of haphazardly arranged round cells that were packed in sheets supported by scant fibrovascular stroma. The neoplastic cells had distinct cell borders, abundant finely granular eosinophilic cytoplasm, and often distinct plasmacytoid features with a juxtanuclear clear golgi zone (Fig. 1). Eccentric nuclei were round and occasionally cleaved with coarse stippled chromatin and 1 to 2 prominent nucleoli, especially in the larger cells. The cells had marked anisocytosis (up to 60 µm in diameter) and anisokaryosis. Moderate numbers of multinucleated cells were seen with up to 4 nuclei per cell. There were 4 mitotic figures per 10 high power fields (400×). Small perivascular aggregates of mature lymphocytes and plasma cells were randomly distributed throughout the neoplasm. The meninges overlying the neoplasm were markedly thickened (up to 300 µm) and composed of dense mature collagenous stroma with few blood vessels. There were mildly increased numbers of astrocytes in the neuroparenchyma adjacent to the neoplasm. Neoplastic cell infiltrates were absent in the pancreas, adrenal gland, thyroid gland, pituitary gland, liver, kidney, spleen, heart, lung, and gastrointestinal tract. The histologic differential diagnosis of this neoplasm included plasma cell tumor, histiocytic sarcoma, and (less likely) lymphoma.

Brain; dog. Plasmacytoma. The neoplastic cells have variable nuclear size, number, and chromatin content. Plasmacytoid differentiation is most apparent in the smaller cells. The giant cells have a large nucleus or multiple nuclei with abundant cytoplasm and paranuclear pallor. HE. Bar = 25 µm.
Moderate to intense diffuse cytoplasmic immunoreactivity of more than 70% of the neoplastic cells was observed with antibodies to lambda light-chain and CD18 (Fig. 2). Occasionally, with lambda light-chain and in a moderate number of cells with CD18, more intense immunoreactivity was present along the cell membrane. Moderate immunoreactivity with a similar distribution was present in approximately 20% of the neoplastic cells with CD79a and in approximately 50% of the neoplastic cells with vimentin. Other immunohistochemical markers (kappa light-chain, cytokeratin, lysozyme, glial fibrillary acidic protein [GFAP], and S100 protein) were not detected in the neoplastic cells. No hyaline deposits were seen, and a Congo red stain did not reveal the presence of amyloid. CD18 immunohistochemical expression, indicating leukocytic lineage; CD79a, indicating B-cell lineage; and the marked plasmacytoid differentiation with lambda light-chain expression were diagnostic for plasma cell tumor.

Brain; dog. Plasmacytoma. Most neoplastic cells have granular cytoplasmic reactivity of variable intensity with anti-human lambda light-chain immunohistochemistry. Polyclonal rabbit anti-human lambda light-chains, Ventana Basic DAB kit, counterstained in hematoxylin. Bar = 50 µm.
Formalin-fixed, paraffin-embedded, 20-µm-thick sections of the tumor were submitted for polymerase chain reaction for antigen receptor rearrangements (PARR) analysis (NCSU-VTH, Raleigh, NC). A clear, repeatable amplicon that did not smudge with heating/cooling was produced using B-cell receptor primers whereas no amplicons were produced using T-cell receptor primers, indicating that the lesion resulted from clonal proliferation of B cells (Fig. 3).

Brain; plasmacytoma; dog. B- and T-cell PARR analysis: Lanes: L = Marker, 1 = no DNA, 2 = +B-cell control unheated, 3 = +B-cell control heated/cooled, 4, 5 = brain tumor sample unheated, 6, 7 = brain tumor sample heated/cooled, 8 = no DNA, 9 = +T-cell control unheated, 10 = +T-cell control heated/cooled, 11, 12 = brain tumor sample unheated, 13, 14 = brain tumor sample heated/cooled.
Ultrastructural examination of a formalin-fixed, paraffin-embedded tumor sample provided evidence of dilated rough endoplasmic reticulum, which when abundant is an ultrastructural feature of plasma cells. In this case, paraffin processing was associated with prominent ultrastructural artifacts.
Plasma cell neoplasms are monoclonal proliferations of mature B cells with monoclonal production of immunoglobulin light-chain and/or heavy-chain proteins. 8 Plasma cell neoplasia manifests either as extramedullary plasmacytomas, solitary osseous plasmacytomas, or multiple myelomas. Solitary osseous plasmacytomas are considered an early manifestation of multiple myeloma. 6 Extramedullary plasmacytomas are common canine neoplasms that typically occur in the skin and gastrointestinal tract, particularly the oral and rectal mucosa. Single cases of extramedullary plasmacytoma have uncommonly been reported in the lung, larynx, spleen, kidney, liver, and uterus. 1,3,5,6 Primary (extramedullary) intracranial plasmacytoma is exceedingly rare in humans or animals, with only 2 cases reported in dogs. 8,9,10 Gross, histologic, immunohistochemical and ultrastructural features were described in 1 of the 2 reported canine intracerebral plasmacytomas. 9 The second case of cerebral plasmacytoma was part of a large study evaluating and comparing immunohistochemical markers, and no specific information regarding this case was presented. 8 A cerebral plasma cell tumor was reported in a cat, but only the brain was examined, and it could not be determined whether the tumor was solitary or part of multiple myeloma. 4
Canine plasmacytomas have been classified histologically in 5 types: hyaline, mature, cleaved, asynchronous, and polymorphous-blastic. 6 This classification system is useful in standardizing the nomenclature of plasmacytomas; however, it cannot be used as a tumor grading system. 6 Following this classification system, this plasmacytoma is most consistent with the asynchronous type, with prominent nuclear-cytoplasmic maturation asynchrony, clear perinuclear halo, and frequent giant cells.
The histologic and immunohistochemical features of this neoplasm resemble those of the canine case reported by Sheppard et al. 9 Both cases were negative for GFAP, S100, lysozyme, and cytokeratin. The present case was positive for vimentin while the previous case was not. Kappa and lambda light-chain immunohistochemistry was not performed in the previously reported case. Immunohistochemistry for immunoglobulin (Ig) heavy-chains was not performed in the present case. In the previously reported canine case, neoplastic cells were strongly positive for Ig A, and, to a lesser extent, for Ig M and Ig G.
Stringent criteria exist in human medicine for the antemortem diagnosis of solitary plasmacytomas: 1) histologic diagnosis of plasma cell tumor with evidence of a clonal population, 2) no detectable bone marrow lesions and clonal population of plasma cells, 3) no bony lesions on survey skeletal radiographs or magnetic resonance imaging of the vertebral column and pelvis, and 4) no hematologic or biochemical alterations of end-organ damage (i.e., anemia, hypercalcemia, or renal failure) that could be caused by a plasma cell proliferative disorder. 7 Such exhaustive investigations were not performed antemortem or postmortem in the present case. Although a detailed necropsy and basic hematologic and biochemical evaluations were performed, we cannot entirely exclude the possibility of multiple myeloma since a small occult bony neoplastic mass may have remained undetected.
PARR utilizes genomic DNA and PCR primers that are specific to the V(D)J splice junctions of B- and T-cell receptor gene segments in lymphocytes. The main advantage of this test is its ability to distinguish monoclonal/oligoclonal (neoplastic) lymphoid proliferation from polyclonal/pseudoclonal (reactive) proliferation, although clonality alone does not prove neoplasia or necessarily imply malignancy. PARR using IgH major B-cell receptor (BCR) and T-cell receptor (TCR-γ) primers confirmed the B-cell origin of the neoplastic cells in this lesion. Antigen receptor gene clonal rearrangement, detected by PCR, has been reported in 2 cases of myeloma and 1 plasmacytoma in dogs; however, the sensitivity and specificity of this PCR is unknown for plasma cells. 2 The primer used for the PARR analysis in the present case is identical to that used in the published report. 2
Based on gross, histologic, and immunohistochemical features and PARR assay evidence of clonality, this neoplasm was diagnosed as a solitary cerebral plasmacytoma. Though rare, this tumor should be in the differential diagnosis for round-cell tumors of the brain.
