Welcome international healthcare professionals

This site is no longer supported and will not be updated with new content. You are welcome to browse and download all content already included in the site. Please note you will have to register your email address to access the site.

You are here

Composite lymphoid neoplasm of B-cell and T-cell origins: a pathologic study of 14 cases

Human Pathology, 4, 45, pages 768 - 784


We retrospectively analyzed 14 composite lymphoma/lymphoid neoplasms (CL) of B-cell/T-cell origins. These consisted of a spectrum of T-cell neoplasms in combination with different B-cell lymphomas/leukemias, with peripheral T-cell lymphoma and diffuse large B-cell lymphoma encountered most frequently for each respective neoplastic lineage. Histopathologic evaluation demonstrated 6 patterns of neoplastic distribution, including zone, inverted zone, diffuse mixed, regional/nodular mixed, compartmental, and segmental distributions. Four of 9 cases studied were positive for Epstein-Barr virus, all with a mixed pattern, suggesting that this pattern may predict an Epstein-Barr virus association. None of 14 cases was considered CL at the initial histologic evaluation. Only 6 (46.2%) of 13 cases had coexisting B-cell/T-cell neoplasms highlighted by immunohistochemistry, and the other 7 (53.8%) cases had 1 or both of the neoplastic components hidden. Flow cytometry detected both neoplastic lineages in 4 (44%) but failed to detect a clonal B-cell population in 4 (44%) and missed neoplastic T cells in 1 (11.1%) of 9 cases. Molecular testing detected clonal rearrangement of IGH/K gene in 11 (84.6%) of 13 cases, and clonal rearrangement of the TCRG/B gene in 13 (92.9%) of 14 cases, including 8 with identical amplicons detected in separate samples. CLs of B-cell/T-cell origin are heterogeneous in subtype combination and topographic pattern, often with one of the components histologically occult. A multidisciplinary approach is emphasized to establish a definitive diagnosis in these challenging cases.

Keywords: Composite lymphoma, Lymphoid neoplasm, B cell, T cell.

1. Introduction

Composite lymphoma/lymphoid neoplasm (CL) is defined as 2 or more morphologically and immunophenotypically distinct lymphomas or lymphoid neoplasms that occur in the same organ or tissue site [1] and [2]. A variety of types of CL have been described in the literature, including the combinations of essentially all the major types of lymphomas, but most reported cases are B-cell lymphoma composite with a Hodgkin lymphoma or 2 composite B-cell lymphomas of different types [2], [3], and [4]. In general, CL is an uncommon lymphoid malignancy, and in particular, those consisting of concurrent B-cell and T-cell neoplasms are quite rare. Nonetheless, an increasing recognition of the latter entity has been reflected by more cases reported in the recent literature [2], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], and [20]. This increase in reported cases is likely attributed to enhanced detection facilitated by widespread use of ancillary tests in the diagnosis of lymphoid neoplasms, including immunohistochemical analysis, flow cytometric immunophenotyping, detection of some lymphomagenic viruses, cytogenetic studies, and molecular diagnostic assays. Still, to the best of our knowledge, the pathologic features of CL of B-cell and T-cell origins have not been well characterized, and the diagnostic criteria and pathogenesis of their concurrence in the same tissue sites remain unclear, despite a few case series [2], [5], [19], and [20] and many case reports [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], and [18]. Herein, we report a series of 14 cases of CL of B-cell and T-cell origins. We specifically evaluated the topographic distributions of B-cell and T-cell components on histologic sections and analyzed the contribution of particular ancillary tests to the detection of each neoplastic component.

2. Materials and methods

2.1. Case selection

After approval of the study by the institutional review board of Duke University Medical Center, we retrieved cases of composite B-cell/T-cell lymphoma/neoplasms in our pathology database by using the search terms “lymphoma,” “leukemia,” “B-cell,” and “T-cell” in pathology reports covering the past 10-year period. Ten cases were from Duke University Medical Center, and 4 were from collaborating academic institutions, including the University of Southern California Medical Center (n = 1), Loma Linda University Medical Center (n = 1), City of Hope National Medical Center (n = 1), and Massachusetts General Hospital of Harvard University School of Medicine (n = 1). Cases with suspicious morphology but without definitive evidence of coexisting B-cell and T-cell lymphoma/leukemia supported by ancillary tests were excluded from this study. Conversely, cases with a positive ancillary test result (ie, a clonal rearrangement of the T-cell receptor [TCR] gene) but without concordant histopathology were also excluded.

2.2. Histologic and immunohistochemical studies

Lymph nodes and other tissue were fixed in formalin. All specimens were then processed routinely, embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E) stain. Four-micrometer sections from paraffin-embedded blocks were immunostained with each antibody. These mainly included the following antibodies: CD20 (1:200; Dako, Carpinteria, CA), CD21 (1:20; Dako), CD23 (1:40; Dako), CD79a (1:100; Dako), PAX5 (1:50; Biocare Med, Concord, CA), OCT2 (1:500; Santa Cruz Biotechnology, Santa Cruz, CA), CD3 (1:100; Lab Vision, Fremont, CA), CD5 (1:40; NCL), CD2 (1:20; NCL), CD4 (1:20; NCL, Buffalo Grove, IL), CD8 (1:25; Dako), CD45RO (1:50; Dako), CD43 (1:3000; D-B Pharm, Saint-Amant-Tallende, France), CD10 (1:25; NCL), BCL6 (1:20; Dako), BCL2 (1:50; Dako), CD138 (1:100; Dako), κ light chain (polyclonal antibody, Dako), λ light chain (polyclonal antibody; Dako), cyclin D1 (1:50; Lab Vision), CD30 (1:50; Dako), HHV8-LNA (1:25; NCL), and Ki-67 (1:50; Dako), using the streptavidin-biotin complex method. The corresponding positive and negative control slides for each antibody stain were carefully examined for quality assurance.

2.3. Epstein-Barr virus–encoded RNA in situ hybridization

Epstein-Barr virus (EBV)–encoded RNA (EBER) was detected by in situ hybridization (ISH). The paraffin-embedded tissue sections were dewaxed in xylene, treated with proteinase K, and hybridized with fluorescein-conjugated EBER oligonucleotide probe (Novocastra, Newcastle upon Tyne, England). The slides were incubated with anti–fluorescein isothiocyanate/alkaline phosphatase and then covered with 5-bromo-4-chloro-3-indolylphosphate, nitroblue tetrazolium, and 1 M levamisole hydrochloride. A negative control probe consisting of fluorescein-labeled oligonucleotide cocktail in hybridization solution was used in parallel. A known EBER ISH–positive tissue section was used as positive control. The stained sections were examined under light microscopy, and a positive cell was determined by its unequivocal nuclear staining.

2.4. Flow cytometric analysis

Flow cytometric immunophenotyping was performed on fresh tissue specimens collected in RPMI 1640 culture medium. Specimens were processed routinely to generate single-cell suspensions, which were then stained with premixed 4-fluorochrome–conjugated antibodies (fluorescein isothiocyanate, phycoerythrin, peridinin chlorophyll protein, and allophycocyanin) per routine leukemia/lymphoma panel protocols. The antibodies in the panel include those against leukocyte common antigen CD45, B-cell antigens (CD19, CD20, CD22, and κ and λ light chains), T-cell antigens (CD2, CD3, CD4, CD5, CD7, CD8), myeloid antigens (CD11c, CD13, CD14, CD15, CD33, and CD117), stem cell antigens (CD34 and CD123), and CD10, CD16, CD25, CD38, CD56, CD103, and HLA-DR (Becton Dickinson Biosciences, San Diego, CA). Approximately 10 000 events per tube were acquired on a flow cytometer (dual-laser FACSCalibur; Becton Dickinson Biosciences) and analyzed using the CellQuest computer software program (Becton Dickinson Biosciences).

2.5. Interphase fluorescence in situ hybridization for CCND1/IGH

The Vysis dual-color dual-fusion CCND1/IGH (11;14) translocation probe set was applied to formalin-fixed, paraffin-embedded sections. This probe set is specific for sequences on either side of the IGH J region breakpoint (14q32) and for regions flanking the common breakpoint region (major translocation cluster) that lies centromeric to the CCND1 locus (11q13). The translocation involving CCND1 at 11q13 and IGH at 14q32, t(11;14)(q13;q32), is visible by fusion of red and green probes. In total, 200 interphase nuclei were evaluated by 2 different technologists, and the percentage of positive cells was reported.

2.6. B-cell and T-cell clonality analysis

2.6.1. Immunoglobulin heavy-chain/κ light-chain gene rearrangement studies

An assay using polymerase chain reaction (PCR)–mediated amplification of the variable (V) and joining (J) regions was used to assess IGH gene rearrangements. Briefly, genomic DNA extracted from the paraffin-embedded tissues was subjected to 2 independent PCR reactions targeting framework 1 and 3 sequences within the V region. After the PCR amplification, fluorescence-labeled PCR products were resolved by capillary electrophoresis on an ABI 310 Genetic Analyzer and evaluated by GeneMapper software (Applied Biosystems Inc, Foster City, CA). To assess the IGK gene rearrangement, the BIOMED-2 multiplex PCR master mixes (Invivoscribe Technologies, Inc., San Diego, CA) were used targeting the V, intragenic, J and Kde regions of the IGK gene.

2.6.2. TCRγ/β gene rearrangement studies

BIOMED-2 oligonucleotide primers and PCR protocols for the TCRγ gene (TCRG) and TCRβ gene (TCRB) rearrangement assays by InVivoScribe Technologies (Invivoscribe Technologies, Inc.) were applied in this analysis. For the TCRG gene rearrangement, 2 multiplexed reactions (master mix A and master mix B) were performed. PCR master mix A primers amplified the V gamma 1-8 and 10 segments with all J gamma exon combinations, whereas PCR master mix B primers amplified the V gamma 9 and 11 segments with all J gamma exon combinations. For the TCRB gene rearrangement, 2 multiplexed reactions (A and B) targeted DNA sequences within the V and J regions of the rearranged TCRB gene locus, and a third reaction (C) targeted DNA sequences within the D and J regions. After PCR amplification, fluorescently labeled PCR products were resolved using capillary electrophoresis on the ABI 3130xl Genetic Analyzer. GeneMapper software (ABI) was then used to visualize and analyze PCR products.

Equal amounts of DNA extract containing 10% clonal B cells or clonal T cells from the cultured neoplastic cells were used in PCR reaction as a positive control. Negative controls included a PCR reaction with equal amounts of genomic DNA from known reactive lymph node and the reaction with distilled water instead of DNA extract.

3. Results

Clinicopathological features of the 14 cases of composite T-cell and B-cell lymphoma/neoplasm are summarized in Table 1 and Table 2. The cases were retrospectively reviewed by 5 hematopathologists (E.W., I.S., J.W., Q.H., and A.L.) independently. The diagnoses of coexisting T-cell and B-cell lymphomas/neoplasms were confirmed, respectively, according to the corresponding sections in the 2008 World Health Organization classification [21] .

Table 1 Clinical features in 14 cases of composite T-/B-cell neoplasm

Case/Sex/Age (y) Clinical presentation Diagnosis Bone marrow examination Stage Treatment Follow up (mo) Outcome
T-cell B-cell
1/F/63 History of SBCL with lymphadenopathy PTCL-NOS DLBCL PCR/TCRG+ nd nd nd nd
2/M/63 Oral cavity ulceration, gingival ulcers PTCL-NOS DLBCL nd nd For PTCL 44 Diagnosis with PBL
3/M/87 Generalized lymphadenopathy PTCL-NOS MCL nd III No nd nd
4/F/82 Multiple lung nodules, hilar lymphadenopathy, serum IgAλ paraprotein PTCL-NOS LPL 1% abnormal T cells by flow cytometry IV R-CNOP 17 Persistent disease a
5/M/18 Renal transplant from father 14 y ago, with skin and lung nodules PTLD-γ/δ TCL PTLD-EBV+DLBCL nd III-IV nd nd nd
6/M/51 3-y history of hepatosplenic T-cell lymphoma with pancytopenia Hepatosplenic T-cell lymphoma SBCL Involved by both neoplasms IV ATGAM, rituximab 10 Persistent disease a
7/M/84 16-y history of splenectomy, with anemia and thrombocytopenia T-LGL Hairy cell leukemia Involved by both neoplasms IV Cladribine nd nd

Fever, night sweat, weight loss, diarrhea, nodular skin rash, lymphadenopathy, splenomegaly PTCL-NOS EBV+DLBCL Negative III CHOP 13 Recurrent disease a
9/F/75 Submandibular/generalized lymphadenopathy PTCL-NOS EBV+DLBCL Involved by PTCL IV R-CHOP 21 Recurrent disease a
10/F/48 Mesenteric and retroperitoneal lymphadenopathy PTCL-NOS lymphoepithelioid EBV+DLBCL PCR/TCRG+ III Targretin 12 Persistent disease a

Multiple erythematous patchy/plaque/nodules in the arm, trunk, and scalp MFs Cutaneous SBCL 0.7% clonal B cells by flow cytometry T3N0M0B0x UVB/Topical steroid 6 nd
12/F/72 Diffuse lymphadenopathy AITL NLPHL <0.5% B cells by flow cytometry; PCR/TCRG+ III-IV CHOP 3 Alive
13/F/57 Multiple lung/pleural lesions ALCL, ALK negative MALT lymphoma nd nd nd nd nd
14/M/47 Inguinal lymphadenopathy ALCL, ALK negative CLL/SLL Involved by CLL/SLL IV R-EPOCH 9 In remission but PCD

a The disease was progressive and incurable by clinical evaluation.

b The cases have been previously published as case reports ([12] and [13], respectively).

Abbreviations: AITL, angioimmunoblastic T-cell lymphoma; ALCL, ALK1 negative, anaplastic large cell lymphoma, anaplastic large cell lymphoma kinase 1 negative; CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL, diffuse large B-cell lymphoma; LPL, lymphoplasmacytic lymphoma; MALT, mucosa-associated lymphoid tissue lymphoma; MCL, mantle cell lymphoma; MFs, mycosis fungoides; NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; PBL, plasmablastic lymphoma; PCD, developed plasma cell dyscrasia; PCR/TCRG+, positive for clonal rearrangement of the T-cell receptor γ gene by PCR assay; PTLD-EBV, posttransplant lymphoproliferative disorder-Epstein-Barr virus; PTCL-NOS, peripheral T-cell lymphomas, not otherwise specified; PTLD-γ/δ TCL, PTLD of γ/δ T-cell lymphoma; R-CNOP, rituximab, cyclophosphamide, mitoxantrone, vincristine, and prednisone; ATGAM, anti–thymocyte globulin; SBCL, small B-cell lymphoma; CHOP, cyclophosphamide, adriamycin, vincristine, and prednisone; T-LGL, T-cell large granular lymphocyte leukemia; UVB, ultraviolet-B phototherapy; R-EPOCH, rituximab, etoposide, prednisone, oncovin, cyclophosphamide and hydroxydaunorubicin; nd, not done.

Table 2 Pathologic features in 14 cases of composite T-/B-cell neoplasm

No. Biopsy site (other sites) Morphology a Immunohistochemistry Flow cytometry EBV status Molecular tests
T cell B cell T cell B cell T cell B cell TCRG/B IGH/K
1 Axillary lymph node Nodular, small to intermediate cells Large cells around T-cell nodules CD2+CD3+CD5+CD8+CD4CD7BCL2↓ CD20+PAX5+OCT2+MUM1+BCL2+CD30+CD43+BCL6+/− nd nd nd +*/nd +/+
2 Hard palate, buccal mucosa 85%-90%, small-large cells, diffuse, intermingled 10%-15%, large cells, diffuse, intermingled CD3+CD5−/+Ki-67 (60%) CD20+ in large cells nd nd nd +/nd +/+
3 Cervical lymph node 80%, intermediate to large cells, T-zone expansion 20%, small cells, compressed to cortex or perivascular aggregates CD3+cyclin D1+ CD20+CD79a+PAX5+CD5w+cyclin D1+/− 60% sCD3+CD5+CD4+CD2CD7CD8 10% CD20+CD5w+CD11Cw+λ+CD10CD23 Negative +/nd −/+
4 Hilar lymph node 50%, small cells, nodular aggregates surrounded by plasmacytoid cells 50%, small-intermediate plasmacytoid cells rimming T-cell nodules CD2+CD4+CD5+CD45RO+CD3CD7CD8CD43CD56CD30ALK1Granzyme B1TIA-1Ki-67 (10%-20%) CD79a+MUM1+

λ+CD139−/+CD20PAX5CD45RA−/+; clusters of cells+ for CD79a, CD20, and PAX5
40% CD45w+CD2+CD4+CD5+CD45RO+sCD3CD7CD8 Negative Negative +/nd +/nd
5 Lung nodule (skin) 70%, small-intermediate cells, diffuse infiltrate 30%, large cells, diffuse, intermingled CD2+CD3+CD5+β F1CD4CD7CD8CD56CD30Granzyme B1TdT CD79a+CD20+CD5 in large cells 55% CD2+CD3+CD5+γ/δ+CD4CD7CD8 Negative Positive in large (B) cells +*/nd +/nd
6 Bone marrow 15%, small cells, interstitial infiltrate 10%, small-intermediate cells, aggregates CD2+CD3++CD7+CD4CD5CD8 CD20+CD5w+CD23cyclin D1 10% sCD3+CD5CD4CD8 1% CD19+CD22+CD5w+κ dim+ nd +*/nd +/nd
7 Bone marrow 30%, intermediate cells, interstitial infiltrate 60%, small-intermediate cells with clear cytoplasm, interstitial infiltrate, intermingled nd nd 35% b sCD3+CD5w+CD8+CD4 15% b CD19+CD20br+CD11Cbr+CD25+FMC7+CD103+κ+CD5CD10CD23 nd +/nd nd/nd
8 Skin (lymph nodes, blood, pleural effusion) 30%, intermediate cells, intermingled in superficial-deep dermis 70%, intermediate-large, intermingled in superficial-deep dermis CD2+CD3+CD4+CD5+CD7+/−CD10 CD79a+MUM1+κ+CD20−/+PAX5CD10CD138 in intermediate to large plasmacytoid cells Involved LN: CD4/CD8 ratio of 5:1 without pan–T-cell antigen loss Pleural effusion and blood: CD45+CD19+CD22w+κw+CD38br+CD20CD138 Positive in large plasmacytoid (B) cells +*/+* +/+
9 Parotid (bone marrow, lymph nodes) 65%, small cells, diffuse, intermingled 35%,intermediate-large cells, clustered or single, intermingled CD2+CD3+CD5w+CD7+CD4+CD10BCL6BCL2↓ CD20+CD10+BCL6+BCL2+CD5 in clusters of large cells CD2+sCD3+CD7+CD5w+CD4+TCRα/β+CD10CD25CD30CD56CD57 Negative Positive in scattered large (B) cells +*/nd +/nd
10 Axillary lymph node 60%, small-intermediate cells, interfollicular, ↑epithelioid histiocytes 5%-10%, large cells, scattered, intermingled CD2+cCD3+BCL2↓ CD20+ in scattered large cells 43%, CDCD4+CD5br+CD10−/+sCD3CD7 2.5% CD19+CD10+CD5sIg Positive in scattered large (B) cells +*/nd −/+
11 Skin (2 punch from arm and 2 from scalp) 20%-50%, small-intermediate cells, papillary/perifollicular dermis 50%-80%, small, nodular infiltrate in mid-deep dermis CD2+CD3+CD4+CD5+CD7−/+CD8 CD79a+PAX5+CD5+CD20+/−CD23−/+BCL2+CD10BCL6cyclin D1 nd nd Negative +*/nd +*/+*
12 Axillary lymph node 65%, small-intermediate with clear cytoplasm, T-zone expansion 35%, nodular aggregates with scattered large “popcorn” cells CD2+CD3+CD4+CD5+CD7+β F1+BCL6+PD1+ BCL2↓ Popcorn cells: CD20+CD79a+PAX5+OCT2+BCL6+MUM1+IgMw+κw+CD5; T-cell rosette+ 25% of total, 47% of T cells CD3w+CD5br+CD4+CD7−/+TCRα/β+ Negative Positive in focal large (bystander) cells +*/nd −/+b
13 Lung/pleural nodule wedged biopsy 40%, large cells with pleomorphic nuclei and abundant eosinophilic cytoplasm in a pleural area. Demarcated from B-cell infiltrate 60%, small-intermediate lymphocytes with plasmacytoid cells, nodular infiltrate within lung parenchyma CD30+CD8+ CD2+CD3w+CD7+/−CD5CD15ALK1CD20granzyme B1perforinCD57CD56 Ki-67 (80%) CD20+CD79a+BCL2+CD5cyclin D1CD23BCL6CD10 (plasmacytoid cells with κ light-chain restriction) Ki-67 (10%) nd nd Negative +/nd +/nd
14 Inguinal lymph node 10%, large cells with pleomorphic nuclei and abundant eosinophilic cytoplasm in subcapsular sinuses 90%, small mature lymphocytes with round nuclei, diffuse with proliferation centers CD45+CD2w+CD4w+CD30+CD3CD5PAX5CD15ALK1 PAX5+CD5w+ Negative CD19+CD20+


nd −/nd nd/nd

a “%” represents the percentage of total nucleated cells in bone marrow, or the percentage of total lymphoid cells in other tissues; the estimation was aided by immunohistochemical analysis.

b The discrepancy between morphologic evaluation and flow cytometric analysis could be explained by dry-tapped bone marrow aspiration.

NOTE. In the column of molecular tests, + represents clonal rearrangement of the gene detected; −, clonal rearrangement was not detected or polyclonal pattern of amplicons; *, identical clonal amplicon was amplified from other tissue biopsy or from multiple tissue sites; **, a distinct product of IGK gene rearrangement was produced in a background of polyclonal B-cell population, but the peak height was lower than the threshold to define a positivity.

Abbreviations: nd, not done; +, positive; , negative; w+, weakly positive; BCL2↓, down-regulation of BCL2; br+, brightly positive; sCD3, surface CD3; sIg, surface immunoglobulin.

3.1. Clinical features

Among the 14 cases of composite B-cell and T-cell lymphoma/neoplasm, 8 patients were male and 6 were female ( Table 1 ). Ages ranged from 18 to 87 years, with a median age of 63 years. Although the sole pediatric patient (case 5) had a history of renal transplant and 2 patients (cases 6 and 7) had a remote history of splenectomy, none of the other 11 patients had a known history of congenital or acquired immunodeficiency or autoimmune disease. Two patients had a history of lymphoma, including small B-cell lymphoma (SBCL) in case 1 and hepatosplenic T-cell lymphoma in case 6. Six patients (cases 1, 3, 9, 10, 12, and 14) presented primarily with generalized lymphadenopathy, 2 (cases 6 and 7) with pancytopenia, and the remaining 6 with extranodal/extramedullary lesions. Among the patients with extranodal/extramedullary involvement (cases 2, 4, 5, 8, 11, and 13), 2 (cases 4 and 8) were found to have underlying lymphadenopathy and/or splenomegaly. One patient (case 2) presented with oral mucosal ulceration and another (case 11) with primary cutaneous lesions; neither demonstrated significant lymphadenopathy or organomegaly. The extranodal organs secondarily involved included the skin in 2 cases (cases 5 and 8), lung in 3 cases (cases 4, 5, and 13), and oral mucosa in 1 case (case 2). Ten patients had bone marrow examination at the time of initial diagnosis, demonstrating involvement by both neoplasms in 2 cases (cases 6 and 7), involvement by T-cell neoplasm in 1 case (case 9), involvement by B-cell neoplasm in 1 case (case 14), submicroscopic (flow cytometric) detection of clonal T cells in 1 case (case 4), submicroscopic detection of clonal B cells in 1 case (case 11), submicroscopic/molecular detection of both neoplasms in 1 case (case 12), molecular detection of clonal T cells in 2 cases (cases 1 and 10), and completely negative findings in the remaining 1 case (case 8). Clinical staging information was available in 10 cases, including advanced stage in 9 cases and early stage in the remaining 1 case (case 11).

3.2. Histology and immunohistochemistry

All the 14 cases had morphologic features suggestive of lymphoid malignancies, either because of effacement of nodal architecture or dense lymphoid infiltrates in extranodal tissues ( Table 2 ). Of these, 6 (42.9%) were considered as probable T-cell neoplasms at initial histologic evaluation, including 3 cases with distinctive nodal interfollicular (T-zone) expansion (cases 3, 10, and 12), 2 cases with cutaneous dense lymphoid infiltrate (cases 8 and 11), and 1 bone marrow biopsy with clinical history of T-cell lymphoma (case 6). Only 4 cases were initially considered to be B-cell lymphoma, owing to clinical history of organ transplant (case 5; EBV-positive diffuse large B-cell lymphoma [DLBCL]), effacement of nodal architecture by monotonous population of small mature lymphocytes (case 14; chronic lymphocytic leukemia [CLL]/small lymphocytic lymphoma [SLL]), dense lymphoid infiltrate with monocytoid appearance (case 13; extranodal lymphoma of mucosa-associated lymphoid tissue [MALT lymphoma]), or plasmacytoid morphology in perinodular areas (case 4; lymphoplasmacytic lymphoma [LPL]). The remaining 4 cases (cases 1, 2, 7, and 9) showed diffuse or interstitial infiltrate by small to large lymphoid cells without morphologic features suggestive of any lineage. Of note, none of 14 cases was considered to be CL at initial histologic evaluation. In cases 13 and 14, discrete clusters or large cells were noted on histologic section but were initially considered to be a large cell transformation of indolent B-cell lymphoma. All cases demonstrated heterogeneous lymphoid infiltration with increased vascularity, except for 2 bone marrow biopsies (cases 6 and 7), 1 lymph node biopsy (case 14), and 1 skin biopsy (case 11). The 2 bone marrow biopsies showed interstitial infiltration by small or small-intermediate lymphoid cells; case 14 showed sinusoidal distribution of large cells with nodal parenchyma effaced by proliferation of small mature lymphocytes; and in case 11, the skin biopsies displayed small to intermediate lymphoid infiltrate in papillary to deep dermis.

Of the 13 cases with immunohistochemistry performed, T-zone expansion of lymph node was highlighted in 3 cases (cases 3, 10, and 12), distinct zone distribution of T-cell and B-cell components in 1 skin biopsy (case 11), inverted zone distribution with central T cells surrounded by neoplastic B cells in 2 cases (cases 1 and 4), diffuse infiltration with 2 neoplastic components intermingled with each other in 5 cases (cases 2, 5, 6, 7, and 9), nodular distribution with mixed neoplastic T cells and B cells in 1 case (case 8), compartmental distribution with neoplastic large T cell confined to sinusoid and SBCL in nodal parenchyma in 1 case (case 14), and segmental distribution with 2 distinct neoplastic components interrupted by normal lung tissue in 1 case (case 13; Table 2 ). Of note, case 10 showed a distinct T-zone expansion where both neoplastic components were accommodated, whereas in cases 3 and 12, neoplastic B cells tended to present in B-cell nodules compressed by expanded T-zones. In case 7, no immunohistochemical analysis was performed because of the detection of both T-cell and B-cell neoplasms by concurrent flow cytometric immunophenotyping of the bone marrow aspirate. Retrospective evaluation of case 7 revealed “hairy” cells intermingled with many small lymphocytes with more condensed chromatin in the bone marrow. Therefore, 6 distinctive patterns of neoplastic distribution were categorized based on the topographic relationship between T-cell and B-cell neoplasms on histologic sections: zone ( Fig. 1 A-C and D-F), inverted zone ( Fig. 1 G-I), diffuse mixed ( Fig. 2 A-C), regional/nodular mixed ( Fig. 2 C-F), compartmental ( Fig. 2 G-I), and segmental distributions ( Fig. 2 J-L; Table 3 ).


Fig. 1 Histopathologic evaluations of CL of B-/T-cell origins: patterns of neoplastic distributions. A-C, Case 3 (zone distribution pattern in lymph node). A, Section of lymph node biopsy shows a marked expansion of interfollicular area (T-zone) with a nodular aggregate of small lymphocytes (center of the image; B-zone; H&E stain, original magnification ×100). Note medium-large cells with clear cytoplasm in T-zone (the right field of the inset) and small lymphocytes in B-zone (the left field of the inset; H&E stain, ×400). B, CD3 stain demonstrates positivity in medium-large lymphocytes in the interfollicular area and negativity in small lymphocytes in the nodular aggregate (×100). The inset shows CD20 positivity in CD3-negative small lymphocytes (×100). C, Both interfollicular medium-large cells and small lymphocytes in the nodular aggregate are positive for CD5 (×200). Note weaker staining of CD5 in small lymphocytes that are positive for CD20. D-F, Case 11 (zone distribution pattern in skin). D, Section of skin punch biopsy shows a dense lymphoid infiltrate in superficial (papillary) and deep dermis (H&E stain, ×40). Inset shows atypical lymphocytes in papillary dermis with “halo” cells lining the basal layer of epidermis (H&E stain, ×400). E, CD79a stain shows B-cell infiltrate in mid-deep dermis with a bottom heavy appearance (×40). Inset shows aberrant expression of CD5 in CD79a-positive B cells in the deep dermis (×400). F, CD4 stain shows positive cells distributed in the papillary and perifollicular dermis with a top heavy appearance (×40). Inset demonstrates CD3 staining of lymphocytes in the papillary dermis (×400). Note the positive cells with large, folded or convoluted nuclei in the papillary dermis. G-I, Case 4 (inverted zone distribution pattern). G, Intermediate magnification of a hilar lymph node section shows nodal architecture effaced by proliferation of small lymphocytes and plasmacytoid cells. Note aggregate of small lymphocytes in the right and plasmacytoid lymphocytes in the left of the field (H&E stain, ×200). H, Small lymphocytes in the nodular aggregate are positive for CD3, whereas surrounding plasmacytoid cells are negative (×100). Of note, the positive cells appear to be cytoplasmic staining, which is consistent with a negative surface CD3 per the flow cytometric analysis in this case. Inset shows the plasmacytoid cells are positive for CD79a (×100). I, The plasmacytoid cells around the T-cell nodule are restricted to λ light chain (×100) and essentially negative for κ light chain (inset, ×100).


Fig. 2 Histopathologic evaluations of CL of B-/T-cell origins: patterns of neoplastic distributions (continued). A-C, Case 9 (diffuse mixed distribution pattern). A, Section of an intraparotid lymph node shows a diffuse proliferation of heterogeneous lymphoid population with many large cells (H&E stain, original magnification ×200). B, CD20 stain demonstrates positivity in medium-large lymphocytes that are mixed with more negative cells with clear cytoplasm (×200). Inset shows large cells positive for EBV-encoded small RNA by ISH (×200). C, CD3 stain demonstrates more positive cells that are smaller than CD20 positive B cells (×100). Note a pattern of diffuse mixed distribution in B-cell and T-cell components. D-F, Case 8 (nodular mixed distribution pattern in the skin). D, Section of skin punch biopsy shows a nodular lymphoid infiltrate in mid-deep dermis (H&E stain, ×40). Inset shows medium-large lymphocytes with plasmacytoid appearance in lymphoid nodule (H&E stain, ×400). E, Many medium-large cells are positive for EBV-encoded small RNA (ISH, ×400). The cells demonstrates reactivity to anti–κ light chain (inset, ×400), but not to anti–λ light chain (data not shown). F, CD3 stain shows reactivity in heterogeneous population including medium-large cells. Note a pattern of mixed distribution in positive cells and negative plasmacytoid cells (×400). G-I, Case 14 (compartmental distribution pattern). G, Section of lymph node shows nodal architecture effaced by proliferation of primarily small lymphocytes. Nonetheless, large pleomorphic lymphocytes with abundant clear cytoplasm are noted in the subcapsular sinus (H&E stain, ×200; inset, ×400). H, CD30 stain demonstrates positivity in large lymphocytes that form a linear distribution in the subcapsular sinus (×200). Large cells in the sinus are weakly positive for CD4 (inset). I, PAX5 stain demonstrates positivity in the small lymphocytes (×400). Note large cells in the subcapsular sinus (right lower corner of the image) are negative for the stain. Inset shows coexpression of CD5 in PAX5-positive B cells (×400). Note the dual intensities of CD5 staining seen typically in SBCL with aberrant CD5 with the weakly stained population representing the neoplastic B cells and the strongly stained population signifying reactive T cells. J-L, Case 13 (segmental distribution pattern). J, Section of pleural/lung biopsy shows a lymphoid infiltrate in pleura and its extension into peripleural adipose tissue (H&E stain, ×40). Inset shows medium-large lymphocytes in the pleural lymphoid infiltrate (H&E stain, ×400). K, Many medium-large cells in the pleural and peripleural lymphoid infiltrate are positive for CD30 (×400). L, Distant from the pleural infiltrate, multiple nodular lymphoid infiltrates were noted in lung parenchyma. A high magnification of the image shows primarily small lymphocytes with monocytoid features and destruction of the respiratory epithelium in the lymphoid nodules (H&E stain, ×400).

Table 3 Summary of detection of composite T-/B-cell neoplasms

Case no. Diagnosis (T cell/B cell) Histologic pattern Immunohistochemistry Flow cytometry EBV study Clonal gene rearrangement
T cell B cell T cell B cell T cell B cell
1 PTCL/DLBCL Inverted zone + + nd nd nd + a +
2 PTCL/DLBCL Diffuse mixed + +/− nd nd nd + +
3 PTCL/MCL Zone + + + + +
4 PTCL/LPL Inverted zone + + + +
5 γ/δ TCL/DLBCL b Diffuse mixed + + + + a +
6 Hepatosplenic TCL/SBCL Diffuse mixed + + nd + a +
7 T-LGL/HCL Diffuse mixed nd nd + + nd + nd
8 PTCL/DLBCL b Nodular mixed + nd c nd c + + a +
9 PTCL/DLBCL b Diffuse mixed + + + + a +
10 PTCL/DLBCL b Regional mixed +/− + + + + a +
11 MF/CSBCL Zone +/− + nd nd + a + a
12 AITL/NLPHL Zone + + + d + a −/+
13 ALCL/MALT Segmental + + nd nd + +
14 ALCL/SLL Compartmental + + + nd nd

a Amplicons with the same nucleotide base pairs being amplified from separate tissue samples.

b EBV-positive large B-cell lymphoma.

c Flow cytometric analysis was performed on specimens other than the skin biopsy on which composite T-/B-cell lymphoma was identified by studies other than flow cytometry.

d Positive in bystander cells but not in “LP cells” seen in B-cell nodules.

Abbreviations: HCL, hairy cell leukemia; CSBCL, primary cutaneous SBCL; ALCL, Anaplastic large cell lymphoma; AITL, angioimmunoblastic T-cell lymphoma; DLBCL, diffuse large B-cell lymphoma; LPL, lymphoplasmacytic lymphoma; MALT, mucosa-associated lymphoid tissue lymphoma; MCL, mantle cell lymphoma; MFs, mycosis fungoides; NLPHL, nodular lymphocyte predominant Hodgkin lymphoma; PTCL, peripheral T-cell lymphomas; SBCL, small B-cell lymphoma; SLL, small lymphocytic lymphoma; TCL,T-cell lymphoma; T-LGL, T-cell large granular lymphocyte leukemia; +, considered to be clonal process with certain confidence; −, uncertain or negative result; +/−, highly suggested as a neoplasia; −/+, suspicious for; nd, not done.

The presence of a T-cell neoplasm was highly suggested by immunohistochemical analysis in 8 (8/13; 61.5%) cases, largely due to confirmed nodal T-zone expansion (cases 3, 10, and 12), distinct zone infiltration (case 11), sinusoidal distribution of large T cells (case 14), segmental solitary lesion of large T cells (case 13), expanded T-cell nodules with medium-large cells (case 1), or diffuse infiltration by T cells (case 2; Table 2 and Table 3). Of these 8 cases, all but 1 case (case 10) showed additional evidence of T-cell neoplasm, by pan–T-cell antigen loss (CD7 in case 1, CD5 in cases 2 and 13, and multiple T-cell antigens in case 14), abnormal morphology (cases 2, 3, 11, 13, and 14), and/or expression of cyclin D1 (case 3). In the remaining 5 cases, including 4 with diffuse infiltration without a distinct pattern of distribution (cases 6-9) and 1 with inverted distribution pattern (case 4), the alterations were insufficient for a definitive diagnosis of T-cell neoplasm, although all demonstrated loss or partial loss of pan–T-cell antigens or down-regulation of BCL2 protein.

Of the 13 cases, 9 were highly suggested to have B-cell neoplasms by immunohistochemistry, including increased large B cells highlighted by B-cell antigen markers in 3 cases (cases 1, 5, and 9), plasmacytoid cells with light-chain restriction in 3 cases (cases 4, 8, and 13), large B cells with T-cell rosette in B-cell nodules in 1 case (case 12), and aberrant CD5 expression in B cells in 2 case (cases 11 and 14). The other 4 cases (case 2, 3, 6, and 10) did not demonstrate sufficient evidence of B-cell neoplasm. Although B-cell antigens were positive on large cells in cases 2 and 10, the number of large B cells was scattered without forming significant clusters or aggregates. Interestingly, in case 12, large cells with convoluted nuclear contours resembling popcorn cells were highlighted by B-cell antigens in B-cell nodules. These popcorn-like cells appeared to be ringed by T cells, best appreciated by nuclear staining of some B-cell antigens, such as PAX5 and OCT2. In cases 3 and 6, B cells appeared to coexpress CD5, but the expression was weak, raising a possibility of nonspecific antibody deposition and thus creating diagnostic uncertainty ( Table 2 ).

3.3. ISH for EBV-encoded RNA

ISH for EBV-encoded RNA was performed in 9 cases in our series ( Table 2 ). Of these, 5 cases were positive, and 4 were negative. Among 5 positive cases, 1 case showed focal positivity in rare large cells, apparently bystander cells, in the expanded T-zone (case 12), whereas the other 4 cases demonstrated positivity in large cells that were stained for B-cell antigen (cases 5 and 8-10). Interestingly, all the latter 4 cases were categorized as mixed pattern of neoplastic distribution (diffuse mixed [ Fig. 2 B, inset] in 2 and regional/nodular mixed [ Fig. 2 E] in the other 2 cases; Table 3 ).

3.4. Flow cytometric analysis

Flow cytometric analysis was performed in 9 cases in our series ( Table 2 ). Of these, 4 cases (cases 3, 6, 7, and 10) demonstrated concomitant abnormal T-cell and monoclonal B-cell populations, supporting a diagnosis of CL of T-cell and B-cell origins; 4 cases showed abnormal T-cell population without detecting monoclonal B-cell population (cases 4, 5, 9, and 12); and 1 case demonstrated a monoclonal B-cell population without detection of abnormal T cells (case 14). In case 8, flow cytometric analysis was not performed on the initial skin punch biopsy that eventually demonstrated 2 coexisting neoplastic components; however, neoplastic T-cell or B-cell component was detected in involved lymph node or pleural effusion/blood respectively, and their immunophenotypes were consistent with what were characterized by immunohistochemistry. In all but 1 case (case 8), abnormal T-cell populations were shown to have loss or altered expression of pan–T-cell antigens ( Supplementary Fig. 4 ). Of 5 cases with monoclonal or abnormal B-cell population detected, 4 showed light-chain restriction (cases 3, 6, 7, and 14), and the remaining 1 case demonstrated a subset of B-cell population with CD10 but without surface immunoglobulin light chains in a lymph node specimen (case 10).

3.5. Molecular diagnostic studies and molecular cytogenetic analyses

PCR-based gene rearrangement studies were performed on all 14 cases, including rearrangement analysis of TCRG in all 14 cases, TCRB in 1 case (case 8), IGH in 12 cases, and IGK in 7 cases ( Table 2 ). All cases but 1 (case 14) had clonal rearrangement of the TCRG gene detected by PCR assay. Of these, 8 cases had the assays performed on other tissue specimens with lymphomatous involvement and demonstrated clonal rearrangement of the TCRG/B gene with amplicons identical to those in the primary tissue biopsy in size of nucleotides ( Fig. 3 A). Of the 12 cases with IGH gene rearrangement analysis, 9 cases were positive for clonal rearrangement, and the other 3 were negative. All 7 cases with IGK gene rearrangement analysis showed clonal rearrangement of the gene ( Fig. 3 B), including 3 cases negative for clonal IGH gene rearrangement (cases 3, 10, and 12).


Fig. 3 Clonal rearrangements of the TCRB gene (A) and IGK gene (B) detected by PCR assay in case 8. Note identical nucleotide size (258 base pairs) of clonal amplicons from skin punch biopsy (upper panel) and lymph node biopsy with evidence of PTCL (middle panel; A), and clonal amplicon of rearranged IGK gene (277 base pairs) from skin biopsy (upper panel of B). A and B (lower panels), PCR amplifications from known reactive lymphoid tissues (polyclonal control).

Interphase fluorescence in situ hybridization (FISH) for the CCND1/IGH fusion gene ( Fig. 4 A and B) was performed on case 3 to demonstrate the genetic basis for the expression of cyclin D1 in both the neoplastic T cells and small B cells by immunohistochemistry ( Fig. 4 C). Evaluation was performed in 2 separate areas: the area with sheets of large cells (corresponding to T-cell lymphoma) and that with aggregates of small lymphoid cells (corresponding to mantle cell lymphoma [MCL]). In the former area, the analysis demonstrated abnormal signals in 62% of interphase nuclei, including fusion signals of CCND1/IGH in 11.8% and amplification of the CCND1 gene (>2 red signals) and/or amplification of the IGH gene (>2 green signals) in 45.5% of interphase nuclei ( Fig. 4 B). In contrast, the latter area was shown to contain abnormal signals in 29% of interphase nuclei, including fusion signals of CCND1/IGH in 13%, amplification of the CCND1 gene and/or amplification of the IGH gene in 16% of interphase nuclei ( Fig. 4 A). Most of the amplification of CCND1 was seen as 1 extra red signal (3 copies), whereas the amplifications of IGH ranged from 3 to 7 green signals in each nucleus.


Fig. 4 Interphase FISH analysis for the CCND1/IGH fusion gene (A and B) in reference to cyclin D1 immunohistochemistry (C) in case 3. A, The area targeted on an aggregate of small lymphocytes (“B-zone”; central small cells in image C). B, The area with larger cells (“T-zone”; medium-large cells around small cell aggregate in image C). Note the positive cells each with 2 fusions of red and green fluorescent signals (FF), representing cells harboring fusion of CCND1 and IGH genes (reciprocal translocations). An isolated red signal and isolated green signal represent intact CCND1 (red) and intact IGH (green) gene, respectively, in each nucleus with 2 fusion signals (RGFF). RRGG exemplifies the nuclei with 2 intact alleles of CCND1 and IGH genes (normal). Note that many nuclei contain more than 2 red (CCND1) signals and/or more than 2 green (IGH) signals, representing amplifications of each gene. R, red; G, green; and F, fusion. C, Cyclin D1 immunohistochemical stain shows nuclear reactivity in both T-zone and B-zone, with stronger staining in some cells in B-zone. A and B, ×1000; C, ×200.

3.6. Pathologic diagnoses of 14 cases

As shown in Table 1 and Table 3, T-cell neoplasms included 7 peripheral T-cell lymphomas, not otherwise specified (PTCL-NOS), 2 anaplastic large cell lymphomas, 1 angioimmunoblastic T-cell lymphoma (AITL; case 12), 1 posttransplant lymphoproliferative disorder (PTLD) of γ/δ T-cell type (case 5), 1 mycosis fungoides (MF; case 11), 1 T-cell large granular lymphocyte leukemia (T-LGL; case 7), and 1 hepatosplenic T-cell lymphoma (case 6). The coexisting B-cell neoplasms included 6 DLBCLs, 1 CLL/SLL (case 14), 1 MCL (case 3), 1 LPL (case 4), 1 MALT lymphoma (case 13), 1 hairy cell leukemia (case 7), 2 other SBCLs (cases 6 and 11), and 1 nodular lymphocyte predominant Hodgkin lymphoma (NLPHL; case 12).

3.7. Clinical course

Eleven cases had management information available in our series ( Table 1 ). Of these, 9 patients were treated with chemotherapy, 1 patient with cutaneous CL (case 11) was treated with ultraviolet-B phototherapy plus topical regimens, and the remaining patient (case 3) received only supportive care owing to debilitated status of his medical condition. The chemotherapeutic protocols were heterogeneous but essentially targeted the T-cell neoplasm in most cases. Clinical follow-up was limited in our series, ranging from 3 to 44 months, with a median of 12 months in 9 total cases. Most showed persistence, progression, or recurrence of the disease.

4. Discussion

CL of T-cell and B-cell origin is rare, and the incidence has been estimated at a rate lower than 1% of all lymphoid malignancies, being reported in less than 100 cases [2], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], and [20], per our extensive search in the English literature. Of the reported cases, the T-cell neoplasms were most often PTCL-NOS [2], [5], [6], [7], [8], [10], [11], [13], [16], [19], and [20] and AITL [2], [5], [18], [19], and [20]; in contrast, the B-cell components were diverse, including most frequently DLBCL [2], [5], [6], [7], [8], [11], [13], [14], [16], [17], [18], [19], and [20] followed by nodal or extranodal marginal zone lymphoma [2] and [15], SLL [5], [9], and [10], Hodgkin lymphoma [2] , plasmacytoma [19] , and LPL [2] and [20]. Most of these cases presented in lymph nodes or other lymphoid organs, whereas in a minority of cases, the coexisting B-cell and T-cell neoplasms arose in extranodal tissues [2], [5], [6], [8], [12], [13], and [16]. Cases of coexisting MFs and B-cell neoplasm in skin have also been previously reported [22], [23], [24], and [25], but all of the cases described in the literature have demonstrated clinical evidence and pathologic features fulfilling the criteria for diagnosis of CLL, suggesting cutaneous involvement by CLL as a component composite with MF. Therefore, case 11 in the current series represents the first reported case of composite lymphoma of MF and cutaneous SBCL [12] . In addition, the MCL seen in case 3 of our series has never been reported to be composite with T-cell neoplasm in the past, to the best of our knowledge. To be more unique in this case, both B-cell and T-cell lymphomas expressed cyclin D1 protein, which is likely regulated by similar genetic alterations.

Histologic diagnosis of CL is challenging because the distribution of each neoplastic component is often uneven, and one dominant component can overshadow the other on the biopsy section. Particularly, in CL of T-cell and B-cell origin, a coexisting T-cell neoplasm is often overlooked because T cells are usually considered “reactive” on most occasions. On the other hand, concomitant B-cell lymphoma may be neglected when large B cells are scattered on the histologic section. Given the diagnostic difficulty, the rate reported in the literature could possibly be underestimated. With advances in diagnostic modalities, particularly their routine applications in diagnostic practice, the detection rate of CL is expected to potentially increase in the future. In most reported cases [2], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], and [20], masked neoplastic components are frequently identified by one or more of the following ancillary tests: flow cytometric immunophenotyping [2] and [5], immunohistochemical analysis [2], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], and [20], tests for certain oncogenic virus such as EBV [2], [13], [18], and [19], cytogenetic study [2] and [5], and molecular diagnostic tests [2], [5], [9], [12], [13], [18], [19], and [20]. In the present series, none of 14 cases was considered to be CL based on initial histologic evaluation of H&E sections. Only 6 (46.2%) of 13 cases with immunohistochemical analysis were highly suggested to contain coexisting T-cell and B-cell neoplasms on the sections. These include cases with either spatially separated 2 components (segmental [case 13] and compartmental [case 14] distribution) or zone/inverted zone distribution (cases 1, 2, 11, and 12). Six (46.2%) other cases had 1 neoplastic component identified, and the remaining 1 case (case 6) required flow cytometric analysis to confirm the monoclonal nature of the lymphoid infiltrate in the bone marrow (Table 2 and Table 3). Of the 6 cases with only 1 neoplastic component identified by immunohistochemistry, 4 (cases 4, 5, 8, and 9) had T-cell neoplasm unidentified, whereas in the other 2 cases (cases 3 and 10), B-cell components were overlooked because of a relatively low proportion of B cells. Of note, 4 of these cases demonstrated mixed pattern of neoplastic distribution, and the remaining 2 cases showed zone/inverted zone distribution (cases 3 and 4). Fortunately, flow cytometric analysis was performed in all but 1 case (case 8), and the cryptic neoplastic components were readily detected.

Nonetheless, the lower rate of flow cytometric detection, particularly a tendency to overlook clonal B-cell components (4/9; 44%), in our series is in contrast to that reported by Demurtas et al [2] , who detected 9% to 50% monoclonal B cells in addition to abnormal T cells in all 7 cases of composite T-cell and B-cell lymphoma. Of those cases in which B-cell neoplasms were missed by flow cytometric analysis in our series, 1 case was NLPHL (case 12) in which neoplastic large cells with nuclear convolution (popcorn cells) were scattered, much lower than 5% of total lymphoid population, and the other 3 cases (EBV+ DLBCL in cases 5 and 9, and LPL in case 4) showed only abnormal T cells without detectable clonal B-cell population, despite apparently more than 30% B cells on the sections for each case. It is uncertain if this failed flow cytometric detection of B-cell components was caused by loss of neoplastic B cells during tissue processing because of their fragility.

In contrast to CL comprising 2 distinct B-cell lymphomas, biclonalities of T-cell and B-cell components in composite B-/T-cell neoplasms can be documented by PCR-based immunoglobulin heavy-chain/κ light-chain (IGH/K) and TCRG/B gene rearrangement analysis, without micromanipulation of each neoplastic component in most cases [2], [5], [18], [19], and [20]. In the present series, 13 of 14 cases had clonal rearrangement of the TCRG/B gene, and 11 of 13 had clonal rearrangement of the IGH and/or IGK gene detected ( Table 2 ). These positive results ultimately confirmed the monoclonalities of both T-cell and B-cell infiltrates/proliferations in these cases with concomitant lymphoid neoplasms suggested or suspected by other studies. The sole negative result of TCRG gene rearrangement analysis in case 14 may be explained by a low proportion of neoplastic T cells in the sample. In this example, a distinct sinusoidal collection of large cells, which were positive for homogeneous CD30 and weak CD4 but negative for B-cell antigens, was diagnostic of anaplastic large cell lymphoma, a T-cell neoplasm.

It has been reported that clonal rearrangement products, particularly of the TCR gene, could be amplified in a certain number of biopsies that show histologic and clinical features consistent with reactive lymphadenopathy or reactive lymphoid infiltrate [26] . To overcome a possible “false”-positive clonal rearrangement of the TCR gene by the PCR approach, comparison between amplicons from separate anatomical sites was applied to be certain of a true-positive result [26] and [27]. It is currently accepted that amplicons with the same length of nucleotide base pairs are highly likely to be identical clones and thus considered to be a monoclonal process [27] and [28]. In the present series, the clonal amplicons of the TCRG/B gene rearrangement from separate tissues were amplified in 8 cases, all with nucleotide base pairs identical to the ones from primary biopsy sites. Although all these confirmed the presence of clonal T-cell population in the primary specimen and in other tissue sites, the identical sizes of amplicons ensured the coexistence of T-cell clones in 2 skin biopsies (cases 8 and 11) where concomitant T-cell lymphoma was suspected but uncertain before the molecular diagnostic tests [12] and [13]. Correspondingly, a clonal rearrangement of IGH/K gene has been detected by PCR-based assay in T-cell lymphoma with B-cell proliferation but no overt B-cell neoplasm [29] . Although this could represent an early detection of B-cell clone, the results of a “positive” gene rearrangement test result should be interpreted with caution because controversial findings have been reported with regard to clinical outcome/follow-up [30], [31], [32], [33], and [34]. Per the authors' experience, a PCR-based detection of B-cell clone should be used to support a diagnosis of concurrent B-cell neoplasm only if B cells constitute significant portion and display prominent cytological atypia/aberrant immunophenotype or they are scattered but exclusively large in size highlighted by immunohistochemistry. Nonetheless, as reflected in the inclusion criteria for the case selection in our series, PCR-based clonality tests should be applied and interpreted in a context of histopathologic suspicion. A diagnosis of CL of T-cell and B-cell origins can be made without clonal tests if the evidence is sufficient morphologically and immunophenotypically (ie, case 14), or with positive clonal test results if the evidence is insufficient but histopathologic suspicion is high (ie, cases 2, 8, and 11).

Why do morphologically and immunophenotypically 2 distinct lymphoid neoplasms coexist in the same tissue site? Are these intrinsically related or simply a coincidence? Statistical data seem to suggest intrinsic linkage between 2 lymphoid neoplasms rather than independent occurrences of 2 unrelated malignancies. For instance, DLBCL has been reported as the B-cell component in an exceedingly high proportion of CL of B-cell and T-cell origins [2], [5], [6], [7], [8], [11], [13], [14], [16], [17], [18], [19], and [20], whereas follicular lymphoma, which has a prevalence similar to DLBCL, is hardly noted in concurrence with T-cell lymphoma. In our series, 2 cases with skin biopsies from multiple anatomical sites consistently demonstrated concomitant T-cell and B-cell lymphomas, per our histopathologic evaluation and molecular diagnostic tests (cases 8 and 11 in Tables 2 ); a similar phenomenon has been reported in most relapse samples of composite T-cell/B-cell lymphoma by other series [20] . These findings support the notion of a cohabitant or symbiotic relationship between neoplastic T-cell and B-cell components, at least in some cases of CL of T-cell and B-cell origins.

Nonetheless, the exact pathogenesis of CL remains unclear at the present time despite many reported cases in the literature. A few hypothetical mechanisms have been proposed to explain why 2 or more distinct lymphoid neoplasms coexist in the same tissue location [2], [5], [19], and [35]. One possible etiology of CL is certain viral infections that presumably induce transformation of 2 or more separate neoplastic clones, particularly noted in CL of T-cell and B-cell origins. This oncogenic activity has been observed in a few viruses, but most frequently noted in EBV in immunocompromised patients [5], [7], [8], [13], [18], and [19]. In the present series, 4 (44.4%) of 9 cases with EBV study demonstrated positive viral RNA in neoplastic B cells, and all of the 4 cases (cases 5 and 8-10) were categorized as DLBCL with mixed pattern of neoplastic distribution; therefore, this pattern of neoplastic distribution may predict an EBV association with the B-cell component. In the literature, EBV-related large B-cell lymphoma has been described in association with PTCL and AITL in a simultaneous or metachronous occurrence [19] . It has been thought that EBV antigens expressed in host B cells may stimulate T-cell proliferation and, via a process of clonal selection, eventually induce neoplastic transformation of T cell in concurrence with the virus-mediated B-cell lymphomagenesis; alternatively, immunodeficiency in patients with T-cell malignancy could predispose to EBV infection that consequently transforms B cells in the host [5], [18], and [19].

However, the etiology of oncogenic viruses cannot explain the cases of CL without evidence of viral infection [20] , including 5 of 9 cases with EBV studies in our series. Many cases of CL have been reported to occur as composite with another neoplastic component subsequently after diagnosis of one lymphoma. This prompts the alternative mechanistic hypothesis that the prior treatment of a neoplasm or immunodeficiencies inherent in patients with a neoplasm may either compromise the immune surveillance in hosts or create an impaired microenvironment predisposing to the development of a second lymphoid neoplasm [24] . In our series, 2 patients had a previous diagnosis of one lymphoma component, one with SBCL (case 1), and the other with hepatosplenic T-cell lymphoma (case 6). Unfortunately, the treatment protocols were not detailed in their clinical records. Six other cases (cases 4, 7, and 11-14) in our series contained at least 1 indolent component of lymphoid neoplasm when the diagnosis of CL was established. It is thus uncertain if 1 of the 2 neoplastic components developed after the other or both neoplasias evolved approximately at the same time in parallel with each other.

Interestingly, case 3 in our series demonstrated rearrangement and replications of the CCND1 gene and up-regulation of its protein expression in both B-cell and T-cell neoplasms ( Fig. 4 ). This finding may suggest a possible oncogenic mutation on a common lymphoid progenitor that underwent divergent clonal evolutions into T-cell and B-cell neoplasms, probably via a “second hit” at other genetic loci [35] .

Other proposed mechanisms include chronic exposure to common antigens that stimulate both B-cell and T-cell proliferation, or exposure to carcinogens that could transform both B-cell and T-cell lineages [23], [24], and [35], and B-cell proliferation as well as clonal expansion mediated by cytokines from preexisting neoplastic T cells [20] . Nonetheless, more than 1 pathway may exist in this complex disease. Undoubtedly, the underlying molecular mechanisms are intriguing and remain to be further investigated in future studies.

The prognosis of patients with CL of B-cell and T-cell origins is currently not well understood. The data in the literature are dichotomous, with some demonstrating a prognosis similar to a single aggressive component such as T-cell lymphoma [5], [18], and [20], whereas others show a relatively favorable outcome [19] . The conflicting results of clinical assessment may be explained by a heterogeneous combination of neoplastic components in different reports. At the present time, optimal treatment of CL has not yet been defined. One approach is to target the more aggressive component because it may ultimately determine the overall survival in the patients [5] and [23]. Of the 10 cases with treatment information available in our series ( Table 1 ), 9 were treated primarily with protocols targeting the T-cell neoplasm, including 4 cases with the addition of rituximab; the remaining case of concomitant hairy cell leukemia/T-LGL was treated with cladribine, a purine analog used in hairy cell leukemia, because of the known indolent nature of T-LGL. Most of our patients showed persistence, progression, or recurrence of the disease with relatively short follow-up; however, the prognosis and optimal treatment need to be evaluated in future analysis.

Supplementary data

The following are the supplementary data to this article.





We thank Steven Conlon in the Department of Pathology at Duke University Medical Center for the technical assistance of photo images and figure illustration. We also thank Lisa Prewitt, PhD, in the Cytogenetic Laboratory at Duke University Medical Center for FISH analysis and Felisa Alcancia in the Clinical Laboratory at Duke University Medical Center for her help with the flow cytometric analysis.


  • [1] H. Kim, R. Hendrickson, R.F. Dorfman. Composite lymphoma. Cancer. 1977;40:959-976 Crossref.
  • [2] A. Demurtas, S. Aliberti, L. Bonello, et al. Usefulness of multiparametric flow cytometry in detecting composite lymphoma: study of 17 cases in a 12-year period. Am J Clin Pathol. 2011;135:541-555 Crossref.
  • [3] C.L. Gonzalez, L.J. Medeiros, E.S. Jaffe. Composite lymphoma. A clinicopathologic analysis of nine patients with Hodgkin's disease and B-cell non–Hodgkin's lymphoma. Am J Clin Pathol. 1991;96:81-89
  • [4] F. Fend, L. Quintanilla-Martinez, S. Kumar, et al. Composite low grade B-cell lymphomas with two immunophenotypically distinct cell populations are true biclonal lymphomas. A molecular analysis using laser capture microdissection. Am J Pathol. 1999;154:1857-1866 Crossref.
  • [5] N. Suefuji, D. Niino, F. Arakawa, et al. Clinicopathological analysis of a composite lymphoma containing both T- and B-cell lymphomas. Pathol Int. 2012;62:690-698 Crossref.
  • [6] W. Cui, F. Fan, D. Zhang, D. Garnett, L. Tilzer. Primary composite lymphoma of the larynx, composed of diffuse large B-cell lymphoma and peripheral T-cell lymphoma, not otherwise specified, presenting as left subglottic tracheal fistula, esophageal diverticulum, and neck abscess. Ann Clin Lab Sci. 2012;42:73-80
  • [7] Y. Hirose, T. Fukushima, Y. Masaki, et al. Epstein-Barr virus–associated composite lymphoma composed of peripheral T-cell lymphoma and an anaplastic variant of a diffuse large B-cell type of non–Hodgkin's lymphoma and strongly expressing p53 protein. Int J Hematol. 2004;79:260-265 Crossref.
  • [8] Y.K. Chen, E. Huang, C.C. Lin, et al. Composite lymphoma: angiocentric T-cell lymphoma (CD8+ cytotoxic/suppressor T-cell) and diffuse large B-cell lymphoma associated with EBV, and presenting clinically as a midfacial necrotizing lesion. Oral Oncol. 2004;40:353-359 Crossref.
  • [9] A. van den Berg, E. Maggio, R. Rust, K. Kooistra, A. Diepstra, S. Poppema. Clonal relation in a case of CLL, ALCL, and Hodgkin composite lymphoma. Blood. 2002;100:1425-1429
  • [10] A. Alomari, P. Hui, M. Xu. Composite peripheral T-cell lymphoma, not otherwise specified and B-cell small lymphocytic lymphoma presenting with hemophagocytic lymphohistiocytosis. Int J Surg Pathol. 2013;21:303-308 Crossref.
  • [11] S. Yamazaki, Y. Fujioka, F. Nakamura, et al. Composite diffuse large B-cell lymphoma and CD20-positive peripheral T-cell lymphoma. Pathol Int. 2011;6:662-666 Crossref.
  • [12] N.A. Whitling, R.P. Shanesmith, L. Jacob, et al. Composite lymphoma of mycosis fungoides and cutaneous small B-cell lymphoma in a 73-year-old male patient: a case study. Hum Pathol. 2013;44:670-675 Crossref.
  • [13] J.A. Papalas, P.K. Puri, S. Sebastian, E. Wang. Primary cutaneous, composite, Epstein-Barr virus–associated, diffuse large B-cell lymphoma and peripheral T-cell lymphoma. Leuk Lymphoma. 2011;52:2027-2030
  • [14] Y. Tachikawa, M. Shiratsuchi, E. Sada, et al. Composite gastrointestinal lymphoma consisting of diffuse large B-cell lymphoma and peripheral T-cell lymphoma. Int J Hematol. 2009;90:275-277 Crossref.
  • [15] W.K. Yun, Y.H. Ko, D.S. Kim, et al. Composite marginal zone B cell lymphoma and enteropathy-type T cell lymphoma of the stomach: a case report. Eur J Gastroenterol Hepatol. 2008;20:791-795 Crossref.
  • [16] A.A. Abou-Elella, T.P. Nifong. Composite EBV negative peripheral T-cell lymphoma and diffuse large B-cell lymphoma involving the ileum: a case report and a systematic review of the literature. Leuk Lymphoma. 2006;47:2208-2217 Crossref.
  • [17] Z. Kaleem, M.H. McGuire, A.C. Caracioni, et al. Composite B-cell and T-cell non–Hodgkin lymphoma of the tibia. Am J Clin Pathol. 2005;123:215-221
  • [18] Y. Xu, R.W. McKenna, M.P. Hoang, R.H. Collins, S.H. Kroft. Composite angioimmunoblastic T-cell lymphoma and diffuse large B-cell lymphoma: a case report and review of the literature. Am J Clin Pathol. 2002;118:848-854
  • [19] A. Zettl, S.S. Lee, T. Rüdiger, et al. Epstein-Barr virus–associated B-cell lymphoproliferative disorders in angloimmunoblastic T-cell lymphoma and peripheral T-cell lymphoma, unspecified. Am J Clin Pathol. 2002;117:368-379
  • [20] O. Balagué, A. Martínez, L. Colomo, et al. Epstein-Barr virus negative clonal plasma cell proliferations and lymphomas in peripheral T-cell lymphomas: a phenomenon with distinctive clinicopathologic features. Am J Surg Pathol. 2007;31:1310-1322
  • [21] S.H. Swerdlow, E. Campo, N.L. Harris, et al. WHO classification of tumours of haematopoietic and lymphoid tissues (IARC Press, Lyon, 2008) 179-351
  • [22] Y.C. Liu, J.F. Tomashefski Jr., R.P. Cleveland, S.J. Nassar, J.E. Trey. Composite cutaneous T-cell lymphoma and small B-cell lymphocytic lymphoma: morphologic, immunologic, and molecular genetic documentation of concurrent lymph node involvement. Mod Pathol. 1994;7:641-646
  • [23] P.R. Hull, A. Saxena. Mycosis fungoides and chronic lymphocytic leukaemia—composite T-cell and B-cell lymphoma presenting in the skin. Br J Dermatol. 2000;143:439-444 Crossref.
  • [24] A.B. Barzilai, H. Trau, M. David, et al. Mycosis fungoides associated with B-cell malignancies. Br J Dermatol. 2006;155:379-386 Crossref.
  • [25] A.L. Volk, S.A. Vannucci, W. Cook, et al. Composite mycosis fungoides and B-cell chronic lymphocytic leukemia. Ann Diagn Pathol. 2002;6:172-182 Crossref.
  • [26] M.H. Delfau-Larue, L. Laroche, J. Wechsler, et al. Diagnostic value of dominant T-cell clones in peripheral blood in 363 patients presenting consecutively with a clinical suspicion of cutaneous lymphoma. Blood. 2000;96:2987-2992
  • [27] A.L. Goeldel, P. Cornillet-Lefebvre, A. Durlach, et al. T-cell receptor gamma gene rearrangement in cutaneous T-cell lymphoma: comparative study of polymerase chain reaction with denaturing gradient gel electrophoresis and GeneScan analysis. Br J Dermatol. 2010;162:822-829
  • [28] C.B. Hutchinson, M. Stoecker, F.F. Wang, et al. Molecular detection of circulating Sezary cells in patients with mycosis fungoides: could it predict future development of secondary Sezary syndrome? A single-institution experience. Leuk Lymphoma. 2012;53:868-877 Crossref.
  • [29] B.T. Tan, K. Seo, R.A. Warnke, D.A. Arber. The frequency of immunoglobulin heavy chain gene and T-cell receptor gamma-chain gene rearrangements and Epstein-Barr virus in ALK+ and ALK− anaplastic large cell lymphoma and other peripheral T-cell lymphomas. J Mol Diagn. 2008;10:502-512 Crossref.
  • [30] A. Nicolae, S. Pittaluga, G. Venkataraman, et al. Peripheral T-cell lymphomas of follicular T-helper cell derivation with Hodgkin/Reed-Sternberg cells of B-cell lineage: both EBV-positive and EBV-negative variants exist. Am J Surg Pathol. 2013;37:816-826 Crossref.
  • [31] J.P. Higgins, M. van de Rijn, C.D. Jones, J.L. Zehnder, R.A. Warnke. Peripheral T-cell lymphoma complicated by a proliferation of large B cells. Am J Clin Pathol. 2000;114:236-247
  • [32] K.K. Reichard, E.J. Schwartz, J.P. Higgins, B. Narasimhan, R.A. Warnke, Y. Natkunam. CD10 expression in peripheral T-cell lymphomas complicated by a proliferation of large B-cells. Mod Pathol. 2006;19:337-343 Crossref.
  • [33] G. Smuk, A. Illés, K. Keresztes, et al. Pheno- and genotypic features of Epstein-Barr virus associated B-cell lymphoproliferations in peripheral T-cell lymphomas. Pathol Oncol Res. 2010;16:377-383 Crossref.
  • [34] M.A. Zaki, N. Wada, M. Kohara, et al. Presence of B-cell clones in T-cell lymphoma. Eur J Haematol. 2011;86:412-419 Crossref.
  • [35] N.M. Mokhtar. Review article composite lymphoma. J Egypt Natl Canc Inst. 2007;19:171-175


a Department of Pathology, Duke University, Durham, NC 27710

b Department of Pathology, Tulane University, New Orleans, LA 70118

c Department of Pathology and Laboratory Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114

d Department of Pathology, Loma Linda University, Loma Linda, CA 92350

e Department of Pathology, City of Hope National Medical Center, Duarte, CA 91010

f Department of Pathology, University of Southern California, Los Angeles, CA 90089

lowast Corresponding author. Department of Pathology, Duke Hospital South, DUMC Box 3712, M-345 Davison Bldg (Green Zone), Durham, NC 27710.

Competing interests: The authors declare no conflict of interest to disclose.

☆☆ Funding support: None reported.

This work was presented in part at the USCAP Annual Meeting, February 26–March 4, 2011, San Antonio, TX.