Immunohistochemical analysis of peripheral T-cell lymphoma in Japanese patients

The authors immunohistochemically analyzed the phenotype of 40 cases of peripheral T-cell lymphoma, including 12 adult T-cell leukemia/lymphoma (ATL) cases. Molecular genetic analysis of the T-cell receptor beta-chain and immunoglobulin heavy chain genes were also applied to cases of angioimmunoblastic lymphadenopathy with dysproteinemia (AILD)-like lymphoma and so-called Lennert's lymphoma. Twenty non-ATL lymphomas expressed a helper/inducer phenotype, whereas only one extranodal case expressed a suppressor/cytotoxic phenotype. Three cases had a CD4-CD8- phenotype, and two cases a CD4+CD8+phenotype. No specific relationship between morphologic characteristics (LSG classification) and phenotype was found among non-ATL lymphomas. Six of eight AILD-like lymphomas had a helper/inducer phenotype. Monoclonality of neoplastic T-cells was demonstrated in six of the seven cases of AILD-like lymphoma by molecular genetic analysis. Two cases of Lennert's lymphoma also showed a helper/inducer phenotype and rearrangement of the T-cell receptor beta-chain gene. Serologically defined ATL cases had a helper/inducer phenotype except in one case that expressed both CD4 and CD8. None of the ATL cases had the CD7 antigen in this study using WT 1 as a CD7 antibody, which is in contrast with the non-ATL lymphomas in which 13 of 25 cases expressed CD7. CD25, strongly detectable in all ATL cases, was negative or weakly expressed in non-ATL lymphomas. These facts suggest that non-ATL and ATL are in the different biologic state, probably resulting from the integration of human T-cell leukemia virus type I (HTLV-I), although both are derived from helper/inducer T-cells.     

A unique phenotype of skin-associated lymphocytes in humans. Preferential expression of the HECA-452 epitope by benign and malignant T cells at cutaneous sites.

It has been proposed that the skin is a functionally unique compartment of the immune system, although little direct evidence supporting this hypothesis has been presented. Here we show that lymphocyte populations at cutaneous sites can be differentiated from otherwise similar populations at noncutaneous sites by their preferential expression of an epitope defined by the MAb HECA-452. This MAb recognizes a predominantly 200-kd cell-surface glycoprotein present on about 16% of peripheral blood T cells, including both CD4+ and CD8+ T cells (17% and 11% HECA-452+, respectively), as well as TCR-delta-bearing T cells (32%+). Most thymocytes (99%) lacked HECA-452 antigen expression, and essentially all the HECA-452+ peripheral blood T cells were found in the adhesion molecule high, CD45R low putative memory cell subset, findings suggesting that HECA-452 expression develops peripherally as a consequence of antigenic stimulation. However, the HECA-452 antigen is not a conventional activation antigen because it was not upregulated with mitogen stimulation of peripheral blood T cells. Most significantly, among 54 diverse specimens of normal/reactive lymphoid tissues and sites of chronic inflammation, there was a clear association of lymphocyte HECA-452 expression and cutaneous location. In extracutaneous sites (n = 38) only about 5% of lymphocytes within the T-cell areas of these tissues expressed this antigen, whereas in inflammatory skin lesions (n = 16), 85% were HECA-452+. The association of HECA-452 expression and cutaneous location was also seen in a series of T-cell lymphomas. The malignant cells of 16 of 18 cases of epidermotropic (patch/plaque) stage mycosis fungoides were HECA-452+, as well as 2 of 7 nonmycosis fungoides peripheral T-cell lymphomas in skin. In contrast, this antigen was not expressed in thymic (lymphoblastic) lymphomas (n = 14), nonepidermotropic (tumor) stage mycosis fungoides (n = 5), and noncutaneous peripheral T-cell lymphomas (n = 15). Among lymphocytes, the preferential expression of the HECA-452 determinant by cutaneous T cells supports the hypothesis that the skin constitutes a immunologically unique lymphoid tissue and suggests that this molecule may play a role in either lymphocyte homing to skin or in lymphocyte interactions with the epidermis.     

The World Health Organization classification of malignant lymphoma: incidence and clinical prognosis in HTLV-1-endemic area of Fukuoka

New insights into the pathogenesis of lymphoid malignancies have been gained through novel genetic, molecular and immunological techniques. A new classification system for lymphoid malignancies, known as the new World Health Organization (WHO) classification, has been proposed recently based on these findings. The relative incidence of the subtypes of malignant lymphoma is known to differ according to geographic location. Adult T-cell leukemia/lymphoma (ATLL) is a human malignancy associated with human T-cell leukemia virus type 1 (HTLV-1), and the Kyushu islands are an HTLV-1 endemic area. To clarify the relationship between the histological classification and prognosis of lymphoid malignancies, we reclassified previous cases in our department and summarized our previous reports using the WHO classification. Of 933 cases of lymphoid malignancies, 471 (50%) were B-cell lymphoma, 396 (42%) T/natural killer (NK)-cell lymphoma and 41 (4%) Hodgkin lymphoma (HL). Analysis of clinical outcome showed favorable prognosis for HL, intermediate for B-cell lymphoma and poor prognosis for T-cell lymphoma. Among B-cell lymphomas, the commonest type was diffuse large B-cell lymphoma (n = 281; 60%). Marginal zone lymphoma of mucosa-associated lymphoid tissue (MALT) was diagnosed in 82 cases (17%), follicular lymphoma in 52 (11%) and mantle cell lymphoma in 24 (5%). Other less common lymphomas were Burkitt lymphoma (n = 9; 2%) and lymphoblastic lymphoma (n = 5; 1%). Using overall survival rates, the various B-cell lymphoma types could be divided into three broad groups for prognostic purposes: (i) low-risk group comprising follicular lymphoma and MALT; (ii) intermediate-risk group comprising diffuse large B-cell lymphoma and Burkitt lymphoma; and (iii) high-risk group comprising mantle cell lymphoma and lymphoblastic lymphoma. Among the T/NK-cell lymphomas, the commonest type was ATLL (n = 191; 48%), followed by peripheral T-cell lymphoma, unspecified (n = 83; 21%), angioimmunoblastic lymphadenopathy with dysproteinemia (AILD) (n = 38; 10%), anaplastic large cell lymphoma (ALCL) (n = 22; 6%). Less common types were lymphoblastic lymphoma (n = 17; 4%), nasal and nasal-type NK/T-cell lymphoma (n = 17; 4%), mycosis fungoides (MF) (n = 9; 2%) and other rare types. With respect to clinical prognosis, T/NK-cell lymphomas fell into three groups: (i) relative low-risk group comprising ALCL, AILD, MF and lymphoblastic lymphoma; (ii) relative intermediate-risk group comprising NK/T-cell lymphoma and unspecified lymphoma; and (iii) extremely high-risk group comprising ATLL. Among the lymphoblastic lymphomas, B-cell type and T-cell type lymphomas exhibited different clinical outcomes. We conclude that the histological, phenotypic and genotypic classification of the new WHO system should be beneficial for the clinical approach to these tumors.     

Expression of T-cell antigens on Reed-Sternberg cells in a subset of patients with nodular sclerosing and mixed cellularity Hodgkin's disease

Expression of T-cell antigens by Reed-Sternberg (RS) cells has not been detected in most studies of Hodgkin's disease (HD). The authors employed an improved method of fixation (paraformaldehyde-lysine-periodate), which sharply defined cell borders and revealed T-cell antigens on RS cells in 8 of 30 (27%) cases of HD. Antigen-specific staining was confirmed by immunoelectron microscopy. RS cells expressed T11 (8/8 cases), Leu-3 or T4 (4/8 cases), Leu-4 or T3 (3/8 cases), but not other T-cell specific antigens (Leu-1, T8, T6, 3A1). RS cells were negative for leukocyte common antigen (LCA/T200), in contrast to positive LCA/T200 staining of RS-like cells in T-cell lymphomas. RS cells in all HD cases were positive for Ki-1 and/or Leu-M1 antigens. The percentage of RS cells expressing T-cell antigens was less than 20% (2 cases), 20-50% (3 cases), or greater than 50% (3 cases). This percentage and the specific T-cell antigens expressed varied in tissues from different sites in each of 2 cases. Expression of T-cell antigens by RS cells was found in nodular sclerosis (6 of 20 cases) and mixed cellularity (2 of 5 cases) but not in lymphocyte predominance (2 cases), lymphocyte depletion (1 case), or unclassified types (2 cases). Two cases of nodular sclerosis contained areas of necrosis surrounded by sheets of lacunar cells (syncytial variant of NSHD). Two other cases were associated with cutaneous lymphoma. One of these cases was mixed cellularity HD, which appeared to be confined to the skin. In a second case, tumor cells of similar phenotype (T4+, Ki-1+) were found in skin and lymph nodes of a patient with coexistent mycosis fungoides and HD. These results are consistent with an origin of RS cells from T cells in some cases of nodular sclerosing and mixed cellularity HD. They also suggest that the same cell type, an activated helper T-cell, is involved in the pathogenesis of both skin lesions and lymphadenopathy of some patients with coexistent mycosis fungoides and HD.     

Emmprin, a cell surface inducer of matrix metalloproteinases (MMPs), is expressed in T-cell lymphomas

Degradation of the extracellular matrix by matrix metalloproteinases (MMPs) is a crucial step in tumour invasion and metastasis. In human carcinomas, tumour cell-fibroblast interactions (TFIs) have been demonstrated to play a role in the up-regulation of MMP levels in tumours, and emmprin is a surface molecule on tumour cells that stimulates nearby fibroblasts to produce MMP-1, 2, and 3. T-cell lymphomas frequently show extranodal organ involvement and skin invasion, but a role for TFIs in their invasion has not been examined in detail. This study investigated TFIs in T-cell lymphomas with special reference to emmprin expression and MMP production. Immunohistochemically, only germinal centre cells and some histiocytes expressed emmprin in non-neoplastic lymph nodes (ten cases), while all T-cell lymphomas [14 cases of adult T-cell leukaemia/lymphoma (ATLL), six cases of lymphoblastic lymphoma, seven cases of anaplastic large cell lymphoma, and nine cases of angio-immunoblastic T-cell lymphoma] expressed emmprin strongly and diffusely. FACS analysis of peripheral blood from normal individuals revealed that small fractions of B-cells, T-cells, and monocytes expressed emmprin, whereas emmprin-expressing T-cells were much increased in number, and expressed this protein to a higher level, in ATLL patients. In vitro co-cultures of emmprin-positive HTLV-1-transformed lymphocytes (MT-2) and emmprin-negative human fibroblasts enhanced the production of pro-MMP-2 (gelatinase A) and active MMP-2, compared with cultures of either cell type alone. This stimulation was inhibited by an activity-blocking peptide against emmprin. Moreover, in histopathological sections from patients with ATL skin involvement, MMP-2 was demonstrated in fibroblasts around infiltrating ATL cells, but not in fibroblasts in non-diseased areas. In conclusion, emmprin is overexpressed by T-lymphoma cells, when compared with normal counterparts, and facilitates MMP-2 production via interactions with fibroblasts, which could play a role in stromal invasion by lymphoma cells.     

Morphologic and immunologic characterization of 50 peripheral T-cell lymphomas.

Fifty T-cell lymphomas, excluding mycosis fungoides and lymphoblastic lymphoma, were studied morphologically and immunohistochemically with a panel of monoclonal antibodies reactive with T-cell differentiation antigens in fresh frozen tissue. Histologically, 36% of the lymphomas were large-cell immunoblastic, 26% were diffuse large-cell, 22% were diffuse mixed small and large-cell, and 16% were monomorphic medium-sized-cell lymphomas. By immunologic studies, 64% of the lymphomas were of helper phenotype, 12% were of cytotoxic/suppressor phenotype, 8% expressed both helper and cytotoxic/suppressor suppressor antigenic markers, and 16% lacked detectable markers for either helper or cytotoxic/suppressor cells. There was no correlation between histologic category and immunophenotype. A common finding, and one which may prove to be helpful in the diagnosis of T-cell lymphomas, was the loss of one or more of the pan-T antigens Leu 1, 4, and 5 or the T-cell antigen Leu 9 in 32 cases. The expression of Leu 1 and Leu 9 was lost in 46% of cases, expression of Leu 4 was lost in 26%, and expression of Leu 5 was lost in 24%. About three-quarters of the lymphomas expressed Ia antigens.     

CD40 ligand is constitutively expressed in a subset of T cell lymphomas and on the microenvironmental reactive T cells of follicular lymphomas and Hodgkin's disease.

Although CD40 has been extensively studied in B- and T-cell non-Hodgkin's lymphomas (NHLs)/leukemias, and more recently in Hodgkin's disease (HD), little is known about the expression of its ligand (CD40L) in lymphoproliferative disorders other than T-cell NHLs/leukemias. A series of 121 lymphoma/leukemia samples, including 35 cases of HD, 34 T-cell and 39 B-cells NHLs, 2 cases of adult T-cell leukemia/lymphoma, and 11 cases of T-cell acute lymphoblastic leukemia, were evaluated for CD40L expression by immunostaining of frozen tissue sections and flow cytometry with the anti-CD40L monoclonal antibody M90. CD40L was constitutively expressed by neoplastic cells in 15 of 36 (42%) T-cell NHLs/adult T-cell leukemia/lymphomas, almost invariably those displaying the CD4+/CD8- phenotype, whereas no CD40L-expressing tumor cells could be found in B-cell NHL and HD. Among T-cell acute lymphoblastic leukemias, CD40L was detected only on 2 cases displaying a stem-cell-like phenotype. In follicular B-cell lymphomas a large number of CD40L-expressing CD3+/CD4+ T lymphocytes were found admixed with tumor cells within the neoplastic follicles and in their surrounding areas. In the nonfollicular B-cell lymphomas, CD40L-positive CD3+/CD4+ T lymphocytes were few or absent. In all HD subtypes other than the nodular lymphocytic predominance, CD40L-expressing CD3+/CD4+ T lymphocytes were numerous in the HD-involved areas and were mainly located in close proximity to the Reed-Sternberg cells. Our data indicate that in human lymphomas CD40L is preferentially expressed by a restricted subset of T-cell lymphomas, mostly with CD4 immunophenotype. Finally, we have provided morphological evidence that CD40L may play an important role in the cell contact-dependent interaction of tumor B-cells (CD40+) within the neoplastic follicles or Reed-Sternberg cells (CD40+) in HD-involved areas and the microenvironmental CD3+/CD4+/CD40L+ T lymphocytes.     

Monoclonal antibodies reactive in routinely processed tissue sections of malignant lymphoma, with emphasis on T-cell lymphomas

The immunoreactivity of eight monoclonal antibodies was evaluated on 45 routinely processed lymphomas (22 T-cell lymphomas, 11 B-cell lymphomas, and 12 cases of Hodgkin's disease). Two antibodies reactive with leukocyte common (T200) antigens (PD7/26 and 2B11) stained most of the B- and T-cell lymphomas but did not stain the Reed-Sternberg cells and variants in Hodgkin's disease. Two antibodies known to stain B cells (LN-1 and LN-2) reacted with some of the B-cell lymphomas, but LN-2 also reacted with the neoplastic cells in six of 22 T-cell lymphomas and with the Reed-Sternberg variants in eight of 12 cases of Hodgkin's disease. The granulocyte antibody anti-Leu M1 reacted with most cases of Hodgkin's disease but also reacted with two of 11 B-cell non-Hodgkin's lymphomas. An antibody to epithelial membrane antigen (anti-EMA) stained some cases of T-cell lymphoma, B-cell lymphoma, and Hodgkin's disease. Leu 7 was expressed in one T-cell lymphoma and in one case of Hodgkin's disease. A novel antibody reactive with T cells (L60) stained all cases of T-cell lymphoma but also stained some cases of B-cell lymphoma and one case of Hodgkin's disease. We conclude that none of these antibodies, when used alone on routinely fixed paraffin-embedded material, is completely sensitive and specific for T-cell lymphoma, B-cell lymphoma, or Hodgkin's disease. However, a panel of antibodies is useful in distinguishing Hodgkin's disease from non-Hodgkin's lymphoma and in suggesting the B- or T-cell phenotype of non-Hodgkin's lymphomas.     

Phenotyping of proliferating lymphocytes in angioimmunoblastic lymphadenopathy and related lesions by the double immunoenzymatic staining technique.

Biopsy specimens of lymph nodes with the histologic characteristics of angioimmunoblastic lymphadenopathy with dysproteinemia (AILD) were obtained from 9 cases (4 cases of AILD and 5 cases of AILD-like T lymphoma [AILD-T]) and histologically analyzed by the use of a double immunoenzymatic staining technique with the combination of a monoclonal antibody against lymphocyte membrane antigen and that against human DNA polymerase alpha (pol alpha), which is detectable in the nucleus of the cells in G1, S, and G2 phases. In all 9 cases, the pol alpha + proliferating cells had a peripheral T-cell phenotype with T11 and Leu-4 antigens, whereas proliferating B cells with B1 antigen were rarely observed. As for T-cell subset antigens, the proliferating T cells had T4+ helper/inducer phenotype in 7 cases, while T8+ suppressor/killer T cells proliferated in 2 cases, although a significant number of T4+ proliferating cells were also recognized. The study on malignant lymphomas that evolved in the 2 cases showed that the T-subset antigens on major proliferating tumor cells were the same as those found in the preceding AILD lesions, suggesting that lymphoma T cells originate from the AILD lesion. The results suggested that AILD without histologic manifestations of malignancy and AILD-T may be a neoplastic disease derived from either subset of peripheral T cells.     

HTLV-Associated Diseases: Human Retroviral Infection and Cutaneous T-Cell Lymphomas

An array of neurologic, oncologic, and autoimmune disorders are associated with infection with the human pathogenic retroviruses human T-cell leukemia virus types I and II (HTLV-I, II), as well as the human immunodeficiency viruses (HIV). The cutaneous T-cell lymphomas, mycosis fungoides (MF) and its hematogenous variant Sezary Syndrome (SS), share similar clinical and pathological features to HTLV-I-associated adult T-cell leukemia (ATL) and speculation of a retroviral link to MF and SS, especially in areas non-endemic for ATL, has lead to an intensified search for HTLV- and HIV-like agents in these diseases. To further explore a potential role for human retroviruses in MF and SS, skin biopsy-derived or peripheral blood mononuclear cell-derived DNA from 17 patients (MF, n=7; erythrodermic MF (EMF), n=5; SS, n=5) from the North Eastern United States were screened using gene amplification by PCR and a liquid hybridization detection assay. Previously published primers and probes for HTLV-I (LTR, gag, pol, env, and pX), and our own primers and probes for HTLV-I (gag, pol, and env), HTLV-II (pol and env) and HIV-I (gag and pol) were employed. Serum antibodies to HTLV-I were negative in all but one EMF patient. The single HTLV-I seropositive patient carrying a diagnosis of EMF generated positive amplified signals for all of the eight HTLV-I regions tested. Ultimately, this individual evolved to exhibit clinical manifestations indistinguishable from ATL. The other 16 patients were negative for all 12 HTLV and HIV retroviral regions. Our findings suggest that none of the known prototypic human retroviruses are associated with seronegative MF and SS. The uniformly positive results for HTLV-I in the seropositive patient suggests that this patient initially presented with a smoldering form of ATL and illustrates the difficulty that sometimes may be encountered in the differential diagnosis of MF, SS, and ATL based solely on clinical and histopathological criteria.     

CCR4 in cutaneous T-cell lymphoma: Therapeutic targeting of a pathogenic driver

New treatments are needed for patients with cutaneous T-cell lymphoma (CTCL), particularly for advanced mycosis fungoides (MF) and Sezary syndrome (SS). The immunopathology of MF and SS is complex, but recent advances in tumor microenvironment understanding have identified CCR4 as a promising therapeutic target. CCR4 is widely expressed on malignant T cells and Tregs in the skin and peripheral blood of patients with MF and SS. The interaction of CCR4 with its dominant ligands CCL17 and CCL22 plays a critical role in the development and progression of CTCL, facilitating the movement into, and accumulation of, CCR4-expressing T cells in the skin, and recruiting CCR4-expressing Tregs into the tumor microenvironment. Expression of CCR4 is upregulated at all stages of MF and in SS, increasing with advancing disease. Several CCR4-targeted therapies are being evaluated, including “chemotoxins” targeting CCR4 via CCL17, CCR4-directed chimeric antigen receptor-modified T-cell therapies, small-molecule CCR4 antagonists, and anti-CCR4 monoclonal antibodies. Only one is currently approved: mogamulizumab, a defucosylated, fully humanized, anti-CCR4, monoclonal antibody for the treatment of relapsed/refractory MF and SS. Clinical trial da1ta confirm that mogamulizumab is an effective and well-tolerated treatment for relapsed/refractory MF or SS, demonstrating the clinical value of targeting CCR4.     

Classification of cell lineage and anatomical site, and prognosis of extranodal T-cell lymphoma – natural killer cell, cytotoxic T lymphocyte, and non-NK/CTL types

Due to their minority among the non-Hodgkin lymphomas, classification of extranodal T-cell lymphomas, including those of the natural killer (NK) cell type, has long been controversial and unclear, and the clinical outcome is not well clarified. Recently, new well-defined disease entities have been described based on tumor cell biology combined with anatomical site, clinical features, Epstein-Barr virus (EBV) status, and cell lineage as determined by immunophenotype and genotype. Cytological features are usually not specific, and there are no morphologic correlates with the classification of extranodal T/NK-cell lymphomas. From a human T-cell lymphotropic virus type 1 (HTLV-1) endemic area in Japan, we report here the analysis of 144 cases of extranodal T-cell lymphoma, from which fresh tissues were available. As the clinicopathological features were known, we simply reclassified the cases according to cell lineage and anatomical site. The extranodal T-cell lymphomas were classified into three types on the basis of cell lineage: (1) natural killer cell (NK) type [sCD3-, CD56+, T-cell receptor gene (TCR) germline], (2) cytotoxic T lymphocyte (CTL) type [sCD3+, TIA-1+, TCR rearranged, CD8+/-, CD4-/+], and (3) non-NK/CTL type [sCD3+, TIA-1-, TCR rearranged, CD4+/-, CD8-/+]. In addition to cell lineage, the anatomical site and clinical features were added for subclassification. NK type tumors (35 cases) included the lymphoblastic type, nasal/nasal-type NK lymphoma, and NK leukemia. The CTL type (46 cases) included anaplastic large cell lymphoma (ALCL), cutaneous type, intestinal, gamma delta T-cell type, and an unspecified type. The non-NK/CTL type (63 cases) included adult T-cell leukemia/lymphoma (ATLL), mycosis fungoides (MF), and an unspecified type. With the exception of ATLL and MF, most extranodal T-cell lymphomas had a cytotoxic phenotype of NK type or CTL type and were often associated with EBV infection. MF and the unspecified type within the non-NK/CTL tumors, with the exception of ATLL, had a favorable prognosis. However, NK and CTL types, with the exception of ALCL, were associated with a poor prognosis. Our results indicate that anatomical site and cell lineage are useful predictors of clinical outcomes of extranodal T-cell lymphomas.     

Leu-M1 antigen expression in acute leukaemias

Leu-M1 is a differentiation antigen present in human myelomonocytic cells. Seventy-seven acute leukaemias were retrospectively stained with anti-Leu-M1 using the immunoperoxidase technique on Bouin-fixed paraffin-embedded sections. The subjects were 44 acute lymphoblastic leukaemias (ALL) and 33 acute myeloid leukaemias (AML) previously characterized by cytochemical and immunologic (cell suspension) methods. Leu-M1 was positive in all the AML and in half of the ALL cases. These results suggest that Leu-M1 does not allow differentiation between AML and ALL. For the ALL cases Leu-M1 was positive in 15/28 B-cell types, 4/12 T-cell type and 3/4 'null'-cell type cases. Thus, this antibody is of no assistance in defining types B, T, or 'null' in ALL. Leu-M1 was also studied on paraffin sections of 34 high grade malignant lymphomas. The antibody was negative in all 13 B-cell lymphomas (lymphoblastic: 6; immunoblastic: 7) and in all 4 'null' cell lymphomas. It was positive in 4/9 peripheral T-cell type, the other T-cell lymphomas (lymphoblastic: 5; immunoblastic: 3) remaining negative. Thus, Leu-M1 may be positive in T-cell lymphomas but it is negative in B-cell lymphomas and is always negative in B or T lymphoblastic types. It seems that lymphoblasts are Leu-M1 negative in non-Hodgkin's lymphoma and may be Leu-M1 positive in leukaemias.     

Heterogeneity of CD3 antigen expression in T-cell lymphoma

CD 3 antigen expression was studied in a series of 98 T-cell lymphomas, using polyclonal antibodies which recognize this molecule in routinely processed, paraffin-embedded, tissue. We identified 40 cases in which CD3 was present on only a proportion of the neoplastic cells. This phenomenon of heterogeneous CD3 expression was commonest in pleomorphic T-cell lymphomas (22/42 cases) and in CD30 (Ki-1)-positive lymphomas (5/11 cases), and was less frequently observed in mycosis fungoides (4/18 cases) and not seen in T-cell lymphoblastic lymphoma (0/9 cases). CD3 expression was often related to cell morphology, with CD3 antigen being present on the smaller neoplastic cells but absent from the larger ones. The diagnostic significance of these observations is that, on occasion, it may be possible to diagnose a lymphoma as being of T-cell origin in paraffin sections by demonstrating a minor subpopulation of CD3-positive neoplastic cells.     

Immunohistochemical detection of BCL-3 in lymphoid neoplasms: a survey of 353 cases.

The bcl-3 gene at chromosome 19q13 encodes a member of the IkappaB family involved in regulating the NFkappaB pathway. Originally identified by its involvement in the rare t(14:19)(q32;q13), BCL-3 expression has never been analyzed in a wide variety of lymphomas. We assessed BCL-3 expression in 353 cases of non-Hodgkin lymphoma and Hodgkin lymphoma using formalin-fixed, paraffin-embedded tissue specimens, a monoclonal antibody specific for BCL-3, and immunohistochemical methods. Of 172 B-cell lymphomas, 10 (6%) were positive for BCL-3, including six of 23 (26%) diffuse large B-cell lymphoma, one of 17 (6%) small lymphocytic lymphoma, one of 26 (4%) follicular lymphoma, and two of 49 (4%) mantle cell lymphoma. All other B-cell neoplasms were negative, including marginal zone lymphoma (n=24, 11 extranodal, nine nodal, four splenic), Burkitt lymphoma (n=10), lymphoplasmacytic lymphoma (n=10), lymphoblastic lymphoma (n=8), and plasmacytoma (n=5). Of 111 T/NK-cell lymphomas, 25 (23%) were positive for BCL-3, including 13 of 40 (32%) anaplastic large-cell lymphoma, three of 10 (30%) angioimmunoblastic T-cell lymphoma, two of eight (25%) extranodal NK/T-cell lymphoma of nasal type, three of 12 (25%) mycosis fungoides, one of five (20%) enteropathy-type T-cell lymphoma, and two of 21 (10%) peripheral T-cell lymphoma unspecified. All other T-cell neoplasms were negative, including lymphoblastic lymphoma (n=6), prolymphocytic leukemia (n=6), and subcutaneous panniculitis-like T-cell lymphoma (n=3). Of 70 Hodgkin lymphomas, of all types, 29 (41%) were positive for BCL-3. The relatively high frequency of BCL-3 expression in some non-Hodgkin and Hodgkin lymphoma types raises the possibility that BCL-3 is involved in the pathogenesis of these tumors, and may be a target of new therapies.     

Immunoreactivity of B-cell markers (CD79a, L26) in rare cases of extranodal cytotoxic peripheral T- (NK/T-) cell lymphomas.

The monoclonal antibodies L26 (CD20) and CD79a are very useful reagents for the immunohistochemical assessment of B-cell lineage in lymphoproliferative disorders. Although very few CD20-positive peripheral T-cell lymphomas (PTL) have been reported, comprehensive analyses of CD79a reactivity in extranodal PTL and NK/T-cell lymphomas have not been performed previously. This study investigated CD79a (clone JCB117) and CD20 reactivity in 94 extranodal non-B-cell lymphomas (enteropathy-type intestinal T-cell lymphoma [n = 52], nasal NK/T-cell lymphoma [n = 11], and primary cutaneous PTL [n = 31]) and in 17 cases of nodal PTL, unspecified. In four cases (enteropathy-type intestinal T-cell lymphoma [n = 3] and nasal NK/T-cell lymphoma [n = 1]), the majority of tumor cells stained for CD79a (all CD20 negative) and one cutaneous PTL, unspecified, was CD20 positive (CD79a negative). Extensive immunophenotyping and polymerase chain reaction-based molecular analyses revealed that all five B-cell marker-positive extranodal lymphomas had a cytotoxic phenotype and did indeed represent monoclonal peripheral T-cell proliferations. To minimize the risk of misinterpretation of lymphoma cell lineage, especially in cases of extranodal, lymphoproliferative disease, we suggest the use of both CD79a and CD20 in combination with a panel of antibodies reactive to T cells, such as betaF1 and CD5, and to T cells and NK cells, such as CD3, CD2, CD56, and TIA-1.     

The T-cell chemokine receptor CXCR3 is expressed highly in low-grade mycosis fungoides

Three chemokines, Mig, IP-10, and I-TAC, are expressed highly in the epidermis. We examined the expression of the receptor for these chemokines, CXCR3, in cutaneous T-cell lymphoma. We compared CXCR3 expression with that of cutaneous lymphocyte antigen (CLA) and the activation marker CD30. CXCR3 was expressed by at least a subset of tumor lymphocytes in all 25 cases of low-grade mycosis fungoides (MF), with most cells positive in 20 cases. In progressed or transformed MF, CXCR3 expression was noted in 5 of 22 cases. In 4 of 5 MF cases with sequential biopsy specimens, large cell transformation was accompanied by loss of CXCR3 expression. In contrast, CLA was expressed in 35 of 42 MF cases with no significant differences in expression level between low-grade and transformed cases. In other lymphomas, CXCR3 was expressed in 4 of 4 cases of lymphomatoid papulosis, 3 of 4 cases of CD8+ cutaneous T-cell lymphoma, and 3 of 6 cases of systemic T-cell lymphoma in skin, but not in 10 cases of cutaneous anaplastic large cell lymphoma. CXCR3 expression was associated with epidermotropic T-cell tumors but was largely absent in dermal-based tumors. This phenotypic change likely influences the loss of epidermal localization.     

Coexpression of epithelial membrane antigen (EMA), Ki-1, and interleukin-2 receptor by anaplastic large cell lymphomas. Diagnostic value in so-called malignant histiocytosis

A group of 63 cases of anaplastic large cell lymphomas (variants of diffuse large cell lymphomas often diagnosed as "malignant histiocytosis") was characterized on both morphologic criteria and expression of epithelial membrane antigen (EMA) and Ki-1 antigen (CD30). On the basis of the reactivity of these tumors with anti-EMA and anti-Ki-1 antibodies, four subtypes could be distinguished. In the majority of cases (n = 49), neoplastic cells coexpressed EMA and Ki-1 antigens. Nineteen of these cases were tested for IL-2R, and all were positive (Type I, EMA+, Ki-1+, IL-2R+). In the second group (n = 5), the neoplastic cells expressed EMA but not the Ki-1 antigen. These cases were not tested for the presence of IL-2R (Type II, EMA+, Ki-1-, IL-2R?). There were tumors with similar morphology expressing only Ki-1 antigen (Type III, EMA-, Ki-1+, IL-2R-) or negative for both EMA and Ki-1 antigens (Type IV, EMA-, Ki-1-, IL-2R-). EMA appeared to occur predominantly on activated cells, as has been previously shown for Ki-1 antigen. Analysis using monoclonal antibodies to T-cell, B-cell, or macrophage-associated differentiation antigens showed that these tumors were heterogeneous in terms of cell lineage. Tumors coexpressing EMA, Ki-1, and IL-2R (Type I), were most commonly of T-cell origin (n = 12); the remainder in this type expressed B-cell markers (n = 4), a mixed B/T phenotype (n = 2), or no clear phenotype (n = 9). By contrast, tumors of Types II, III, and IV were mainly from B-cell origin (n = 6) or showed a mixed B/T phenotype (n = 1). Despite the fact that a significant proportion of these cases were initially classified morphologically as "malignant histiocytosis," only 3 of the 63 cases were possibly of histiocytic origin. These results confirm that true malignant histiocytosis is rare and that most tumors with histologic features currently regarded as being consistent with this diagnosis are lymphocytic in origin and express activation antigens such as EMA, Ki-1 antigen, and IL-2R.     

Diminished CD3 expression is useful for detecting and enumerating Sézary cells

We evaluated 44 peripheral blood, bone marrow, and lymph node specimens from 23 patients with the clinical diagnosis or suspicion of cutaneous T-cell lymphoma by 4-color flow cytometry for the presence of altered CD3 expression and compared the results with light microscopic evidence of involvement by mycosis fungoides (MF) or Sézary syndrome (SS). In 19 specimens (15 peripheral blood, 3 lymph node, 1 bone marrow) from 6 patients, we noted reduced CD3 expression compared with normal T lymphocytes. Diminished CD3 expression correlated with morphologic evidence of MF/SS cells in all cases, although the enumeration of MF/SS cells differed. Comparison of the enumeration of MF/SS cells in the peripheral blood by CD3+CD7- phenotyping with CD3dim phenotyping suggested that CD3dim phenotyping was the most accurate method for counting MF/SS cells. Therefore, diminished CD3 expression as shown by flow cytometry may be particularly useful for detecting and enumerating MF/SS cells in hematopoietic tissue of patients with cutaneous T-cell lymphoma.     

Immunophenotyping and gene rearrangement analysis provide additional criteria to differentiate between cutaneous T-cell lymphomas and pseudo-T-cell lymphomas.

Differentiation between cutaneous pseudo-T-cell lymphomas and cutaneous T-cell lymphomas (CTCLs) may be extremely difficult. In this study, it was investigated whether demonstration of an aberrant phenotype and detection of clonal T-cell receptor gamma (TCR gamma) gene rearrangements can be used as additional differential diagnostic criteria. Immunohistochemical studies and TCR gamma gene rearrangement analysis using a polymerase chain reaction with primers specific for V gamma 1-8 and V gamma 9 gene segments in combination with denaturing gradient gel electrophoresis (PCR/ DGGE) were performed on frozen material of 11 pseudo-T-cell lymphomas and 17 CTCLs, including 9 cases of mycosis fungoides (MF) and 8 pleomorphic CTCLs. Clonal TCR gamma gene rearrangements were found in 66% of patch/plaque-stage MF and 100% of tumor-stage MF and pleomorphic CTCL, but not in any of 10 pseudo-T-cell lymphomas studied. Aberrant expression of CD2, CD3, and/or CD5 antigens was noted in 3 of 6 (50%) cases of patch/plaque-stage MF, all three cases of tumor-stage MF, and 5 of 8 (62%) pleomorphic CTCLs, but not in any of the 11 pseudo-T-cell lymphomas. Moreover, in pseudo-T-cell lymphomas exhibiting a nodular or diffuse growth pattern, a considerable admixture with reactive CD8+ T cells (15 to 60%), B cells (up to 20%), and macrophages was a characteristic finding. In conclusion, the results of this study suggest that demonstration of clonal TCR gene rearrangements and an aberrant phenotype, as well as demonstration of many admixed CD8+ T cells and B cells can be considered as useful additional criteria in the differentiation between pseudo-T-cell lymphomas and CTCLs.