Hematopathology / PRIMARY LYMPH NODE PLASMACYTOMAS Primary Lymph Node Plasmacytomas (Plasmacytic Lymphomas)

To determine whether primary lymph node plasmacytoma (PLNP) is a distinct entity among other types of plasma cell neoplasia, we analyzed a large series of PLNPs from 2 large lymphoma registries to compare histologic, immunophenotypic, and clinical features of PLNPs, nonnodal extramedullary plasmacytomas, and multiple myeloma. Twenty-five PLNPs (clinical data on 15 cases) were compared with 10 non–lymph node plasmacytomas and 51 cases of multiple myeloma; 36 cases of reactive plasmacytoses were used as controls. The histologic features of PLNP and other extramedullary plasmacytomas were similar. The histologic features of PLNPs were more immature than those of reactive plasmacytoses and less immature than in multiple myeloma. The immunophenotype of PLNPs significantly differed from that of reactive plasmacytoses, other extramedullary plasmacytomas, and multiple myeloma. PLNPs did not progress to multiple myeloma, unlike other extramedullary plasmacytomas, even though survival in PLNPs and other extramedullary plasmacytomas was similar. Our findings suggest that PLNPs may be distinct from other plasma cell dyscrasias. Primary lymph node plasmacytomas (PLNPs) are rare malignant neoplasms.1-20 PLNPs represent 2% of all extramedullary plasmacytomas,19 0.5% of lymph node malignant neoplasms,8 and only 0.08% of all plasma cell malignant neoplasms.20 PLNPs can be diagnosed only after exclusion of metastatic multiple myeloma (which metastasizes to lymph nodes in up to 40% of cases of advanced-stage disease)19 and metastatic upper respiratory tract plasmacytomas (which represent 76% of extramedullary plasmacytomas and infiltrate cervical lymph nodes in approximately 15% of cases).19 A total of 33 PLNPs have been described, primarily as single case reports, 7 of them arising in Castleman disease (CD), plasma cell type.9,21-26 The 26 PLNPs not associated with CD1-7,10-20 showed a male preponderance (16 of 26 cases). The average age of the 26 patients was 56 years (range, 10-72 years). The tumors usually manifested as a localized mass (22 of 26 cases) in the neck (18 of 20 cases). Except for neck swelling, most PLNPs were asymptomatic. The lymph nodes excised were large (average size, 4.6 cm; range, 2-10 cm) and frequently were associated with serum monoclonal proteins (6/14 cases [43%]). The serum monoclonal proteins and plasma cell immunoglobulin expression were usually of IgG kappa type (3 of 6 reported cases of paraproteinemia). Focal amyloid deposition has been reported in 2 of 5 cases. Most patients had localized disease at initial diagnosis and were cured (15 of 21 cases) after local excision or radiation therapy. Patients with disseminated disease frequently died of disease (2 of 4 cases), which spread by lymph node rather than osseous metastasis. Only 1 described patient with disseminated disease has been cured with myeloma-type chemotherapy (melphalan and prednisone).10 Am J Clin Pathol 2001;115:119-126 119 © American Society of Clinical Pathologists Menke et al / PRIMARY LYMPH NODE PLASMACYTOMAS PLNP reported in association with CD manifests with systemic lymphadenopathy typical of plasma cell CD (cervical, 2; axillary, 2; mediastinal, 1; and inguinal, 1).21-26 PLNP in CD occasionally metastasizes to bone, forming solitary sclerotic bone lesions (typically has demonstrated lambda-restricted monoclonal proteins or plasma cell immunoglobulin expression). PLNP with CD may be associated with a POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, monoclonal proteins, and skin changes; 2 of 6 cases).27 The size of lymph nodes (average, 5.8 cm; range, 3.5-10 cm), age of patients (average, 50 years; range, 38-56 years), male predominance (4 men, 2 women), frequency of amyloid deposition (1 of 3 cases), and frequency of serum monoclonal protein (3 of 5 cases) in PLNPs with CD are similar to PLNPs not associated with CD. Patients with PLNP and CD (especially those with POEMS syndrome) have disease that is often refractory to treatment with corticosteroids, chemotherapy, or radiation. In the Kiel classification,28 primary lymph node plasmacytomas were designated as plasmacytic lymphomas. The Revised European-American Classification of Lymphoid Neoplasms (also called the REAL classification),29 however, lists extramedullary plasmacytomas and multiple myeloma as a single-spectrum disease with similar histologic, immunophenotypic, and clinical behavior. We investigated whether PLNP may be a distinct type of plasma cell neoplasia. We collected a large series of PLNPs from 2 large lymphoma registries to compare histologic, immunophenotypic, and clinical features of PLNP with other plasma cell neoplasms and reactive plasmacytosis. Materials and Methods We identified cases of PLNP from 2 large lymphoma registries (aggregate of about 200,000 lymphomas) through the computer archives. Disseminated multiple myeloma and extramedullary plasmacytoma were excluded by clinical examination for adjacent mucosal tumors, radiographic evidence of regional tumors, skeletal surveys, and bone marrow biopsy. Lymphoplasmacytic lymphomas with extensive plasmacytic differentiation30 were excluded by histologic identification of tumefactive masses of plasma cells without intervening small B lymphocytes typical of lymphoplasmacytic lymphomas. According to Greipp et al,31 plasma cell morphologic features were described as mature, immature, or plasmablastic ❚Image 1❚. Mature plasma cells had dense clock-faced nuclear heterochromatin with abundant eccentric cytoplasm and inconspicuous nucleoli. Immature plasma cells showed more open, vesicular nuclear chromatin and prominent single nucleoli. Plasmablasts were distinguished by open vesicular nuclear chromatin with a prominent single nucleus and scant eccentric cytoplasm. Plasmablastic morphologic features were established by the histologic recognition of more well-differentiated plasma cell areas. Immunohistologic staining was performed according to the method of Hsu et al.32 A broad panel of antibodies against a wide variety of hematopoietic antigens ❚Table 1❚ was used that defined heavy-chain and light-chain restriction via cytoplasmic immunoglobulin staining, B-cell antigen expression, and expression of other hematopoietic antigens previously described in plasmacytomas or multiple myeloma.33,34 PLNPs that failed to show restricted immunoglobulin lightchain expression had paraffin blocks analyzed for immunoglobulin heavy-chain receptor gene rearrangements. DNA for gene rearrangement analysis was extracted from three 10-μm sections. The DNA was amplified with polymerase chain reaction, mineral oil was removed, and samples were extracted with chloroform. Ten microliters of each sample was run on a 3% agarose gel for 2 hours at 75 V. Fractions were stained with ethidium bromide and photographed. To detect a monoclonal JH rearrangement, the seminested primer system35 was used (primers Fr3A, LJH, and VLJH). Polymerase chain reaction was done with Taq polymerase (GIBCO, Life Technologies, Rockville, MD) and performed in a thermocycler (Stratagene, LaJolla, CA). All primers were synthesized on a DNA synthesizer (Applied Biosystems, Foster City, CA). The histologic, immunophenotypic, and clinical features of PLNPs were compared with those of an ageand sexmatched group of 10 non–lymph node extramedullary plasmacytomas, 51 previously published cases of multiple myeloma, and a control group of 36 reactive plasmacytoses. Extramedullary plasmacytomas were clonal as defined by restricted immunoglobulin light-chain expression or immunoglobulin gene rearrangement. Multiple myeloma cases were determined by the histologic identification of primary plasma cell neoplasms of bone accompanied by multiple lytic bone lesions and monoclonal serum or urine proteins. Reactive plasmacytoses were composed of dense aggregates of polyclonal plasma cells. Comparisons between groups for histologic and immunohistologic data were made with nonparametric methods for small sample sizes (Fisher exact test), and survival comparisons were made with the log-rank test and Kaplan-Meier survival curves.