Although there is a close association between Langerhans cell histiocytosis and malignant neoplasms, simultaneous occurrence of lymphoblastic lymphoma and Langerhans cell histiocytosis in the same lymph node is an extremely rare finding. Herein, we describe such a case in a 26-year-old woman who presented with progressive cervical lymphadenopathy. The lymphoma cells have an immature T-cell phenotype (terminal deoxynucleotidyl transferase+, HLA-DR+, CD34+, CD38+, and CD7+) with expression of both CD3 and CD79a on immunohistochemical stain. The Langerhans cells are present focally with the characteristic morphologic features and immunophenotype (CD1a+ and S100+). The significance of CD79a coexpression in T-cell lymphoblastic lymphoma and the association between lymphoblastic lymphoma and Langerhans cell histiocytosis are discussed.
Lymphoblastic lymphoma comprises approximately 5% of all non-Hodgkin lymphomas. It occurs predominantly in adolescents and young adults, who present with a rapidly enlarging mediastinal mass and frequent central nervous system involvement. Besides age and cytogenetics, immunophenotype is also a very important prognostic factor.1,2
Langerhans cells represent a distinct subset of a large family of dendritic cells and possess characteristic morphologic, immunophenotypic, ultrastructural, and functional features. They are generally believed to originate in the bone marrow. Proliferation of Langerhans cells can manifest in various forms, ranging from a solitary eosinophilic granuloma in the lung or bone to disseminated Langerhans cell histiocytosis involving multiple organ systems.
There appears to be a frequent association between Langerhans cell histiocytosis and malignant neoplasms, either concurrently or sequentially.3,4 Only 2 cases of simultaneous occurrence of Langerhans cell histiocytosis and lymphoblastic lymphoma have been reported in the literature.5 Herein, we present an additional case of coexistent Langerhans cell histiocytosis and lymphoblastic lymphoma, with the latter expressing both CD3 and CD79a.
REPORT OF A CASE
The patient was a 26-year-old, white woman who presented with a 1-month history of a progressively enlarged left cervical lymph node. She denied having fever, night sweats, or weight loss. The results of physical examination were unremarkable except for left cervical lymphadenopathy. A chest x-ray film showed no evidence of mediastinal enlargement. An abdominal computed tomographic scan was negative for lymphadenopathy and organomegaly. The complete blood cell count was within normal range (white blood cells, 8.5 × 109/L; hemoglobin, 127 g/L; platelet, 315 × 109/L). A biopsy of the left cervical lymph node was performed. Subsequently, a bone marrow biopsy for staging was also performed, and a portion of the bone marrow aspirate was sent for flow cytometric immunophenotyping. The patient was treated with standard chemotherapy but continued to have residual disease at the last follow-up.
MATERIALS AND METHODS
The lymph node was sectioned and fixed in 10% buffered formalin and embedded in paraffin. Histologic sections were stained routinely with hematoxylin-eosin for morphologic evaluation.
Immunohistochemical stains were performed on the lymph node using the standard avidin-biotin peroxidase complex technique.6 The antibodies included CD1a, CD3, CD4, CD5, CD8, CD20, CD45, CD79a, CD68, CD99, myeloperoxidase, and terminal deoxynucleotidyl transferase (TdT) (Dako Corporation, Carpinteria, Calif).
Flow cytometric immunophenotyping was performed on the bone marrow aspirate using a 4-color FACSort flow cytometer (Becton Dickinson, San Jose, Calif) with various combinations of the following antibodies: CD2, CD3, CD4, CD5, CD7, CD8, CD10, CD11a, CD11b, CD11c, CD13, CD14, CD15, CD16, CD19, CD20, CD22, CD23, FMC-7, CD33, CD34, CD38, CD45, CD56, and HLA-DR conjugated with different fluorochromes (Becton Dickinson). Briefly, the bone marrow aspirate was washed with 1× phosphate-buffered saline. Red blood cells were then lysed with ammonium chloride. After washing with 1× phosphate-buffered saline, the cells were incubated with appropriate fluorochrome-labeled antibodies and isotype controls. Approximately 10 000 events were acquired and the cells of interest were analyzed using Cellquest (version 3.1, Becton Dickinson).
The lymph node biopsy specimen showed 2 populations of abnormal cells. Most cells were medium in size with very dispersed chromatin and high mitotic rate, which largely replaced the lymph node architecture (Figure 1). On immunohistochemical staining, these neoplastic cells were positive for CD3, TdT, and CD99 (Mic-2 or O13) but were negative for CD1a, CD4, CD5, CD8, CD20, CD68, and myeloperoxidase. Interestingly, these cells were also strongly and uniformly positive for CD79a (Figure 2). The remainder of the lymph node contained large, pale histiocytes with nuclear grooves, characteristic of Langerhans cells (Figure 1). These cells were present predominantly in a single large focus but were also scattered elsewhere throughout the lymph node. As expected for Langerhans cells, these histiocytes were characteristically positive for CD1a and S100 (Figure 2). No admixed eosinophils were observed.
A staging bone marrow biopsy specimen revealed a normocellular marrow with focal involvement by lymphoblastic lymphoma. There was no evidence of Langerhans cell histiocytosis in the bone marrow. Flow cytometric studies demonstrated that the neoplastic cells had a very immature T-cell immunophenotype expressing HLA-DR, CD34, CD38, and CD7. No other commonly used B-cell, T-cell, natural killer, or myeloid lineage markers were detected (data not shown). The morphologic, immunohistochemical, and flow cytometric immunophenotyping results supported the diagnosis of lymphoblastic lymphoma derived from T-cell lineage.
This case report raises 2 interesting questions. First, what is the clinical and pathologic implication of CD79a expression in T-cell lymphoblastic lymphoma? Second, what is the relationship between Langerhans cell histiocytosis and lymphoblastic lymphoma?
CD79a is a B-cell surface molecule that is closely associated with immunoglobulins. Through this interaction, CD79a is involved in the mediation of intracellular signal transduction after antigen binding in B cells, similar to CD3 in T cells. In addition, CD79a is expressed during the entire span of B-cell development. Because of these features, CD79a is considered to be a very sensitive and specific marker for the identification of B-cell neoplasms. It is particularly useful when only paraffin-embedded tissues are available, because CD79a is resistant to routine tissue processing.7
The specificity of CD79a for B-cell neoplasms, however, has been challenged recently. By immunohistochemistry, Pilozzi et al8 reported that CD79a was expressed in approximately 40% of cases of T-cell acute lymphoblastic leukemia/lymphoma, and its expression was further confirmed by Western blotting in some of these cases. These authors then performed gene rearrangement studies for the immunoglobulin heavy chain and T-cell receptor γ genes in T-cell acute lymphoblastic leukemia/lymphoma with or without CD79a expression detected by immunohistochemistry. They found that all cases showed T-cell receptor γ gene rearrangement and none showed solely immunoglobulin heavy chain gene rearrangement, regardless of CD79a expression.9 Blakolmer et al10 reported the expression of CD79a in a few peripheral T-cell lymphomas by immunohistochemical staining, and the clonal T-cell origin in these cases was confirmed by polymerase chain reaction for the T-cell receptor γ gene. Using flow cytometric immunophenotyping, Lai et al11 also identified 3 of 8 cases of T-cell acute lymphoblastic leukemia/lymphoma coexpressing CD79a. In our case, CD79a is uniformly and strongly positive on the T-lymphoblastic lymphoma cells (Figure 2, B). All these results demonstrate that CD79a can be expressed in both mature and immature T-cell lymphoproliferative disorders.
The clinical implication of CD79a expression in T-cell acute lymphoblastic leukemia/lymphoma is uncertain. In the small series reported by Lai et al,11 all 3 cases occurred in patients younger than 18 years, and all had unusual cytogenetic abnormalities and a poor response to treatment. Additional prospective study with more cases is needed to confirm their findings.
Increasing evidence indicates a close relationship between Langerhans cell histiocytosis and malignant neoplasms.3,4 Most lymphomas associated with Langerhans cell histiocytosis reported in the literature are classic Hodgkin disease, with a few being non-Hodgkin lymphoma. In most of these cases, lymphoma and Langerhans cell histiocytosis are present in the same lymph node, with the latter being microscopic lesions. Both diseases are frequently diagnosed concurrently, and none of the cases meet the diagnosis of systemic Langerhans cell histiocytosis. All patients described are treated with either chemotherapy and/or radiation therapy for their lymphoma. Among those patients with follow-up information, the presence of Langerhans cell histiocytosis does not influence the treatment outcome of malignant lymphoma. These findings suggest that Langerhans cell histiocytosis represents a reaction to the coexisting malignant lymphoma rather than a true neoplastic process.
Only 2 cases of composite lymphoblastic lymphoma and Langerhans cell histiocytosis have been described in the literature.5 Except for the fact that the status of CD79a expression is unknown, the lymphoblastic lymphoma cells have the same immature T-cell phenotype as the case described herein. In contrast to our case, however, the 2 patients appear to have a less aggressive clinical course, suggesting that Langerhans cell histiocytosis is not associated with the clinical behavior of the coexisting lymphoblastic lymphoma.
In summary, CD79a should not be used as a sole marker for B-lineage differentiation. Whenever possible, flow cytometry should be performed to characterize the immunophenotype. The coexistence of Langerhans histiocytosis and lymphoblastic lymphoma is most likely incidental, and the presence of Langerhans cell histiocytosis should not influence the treatment choice for lymphoblastic lymphoma.
Reprints: Shiyong Li, MD, PhD, Department of Pathology and Laboratory Medicine, Emory University School of Medicine, 1364 Clifton Rd NE, Atlanta, GA 30322 (firstname.lastname@example.org)