Hodgkin and Reed-Sternberg cells (HRS cells) are hallmarks of classic Hodgkin lymphoma (cHL). However, cells with similar morphology and phenotype are occasionally seen in non-Hodgkin lymphomas (NHL). In contract to HRS cells in cHL, HRS cells associated with NHL are isolated single cells or small clusters among the NHL cells, do not form mass lesions, and are not associated with mixed inflammatory background typically seen in cHL. The HRS cells in NHL share similar or identical morphologic and phenotypic features to those of cHL and often pose diagnostic challenges in distinguishing them from cHL and composite lymphoma.
The HRS cells in conditions other than Hodgkin lymphoma were first observed by Strum et al1 in 1970. Shin and Rappaport then reported a more detailed clinicopathologic study of HRS cells in B-cell NHLs in 1993.2 In their report of 7 low-grade B-cell lymphomas, they observed coexisting large cells with a morphology similar to Hodgkin cells (named as RS-like cells). When analyzed by immunohistochemistry, the phenotypes of the large cells in all the cases, however, were CD15−, CD20+, CD30−, which were more typical for mature B cells rather than Hodgkin cells. Since then, similar cases have been reported in single case reports and small case series. Many of the later reports have shown that the phenotypes of the HRS cells were more closely related to Hodgkin cells rather than mature B cells.3 Microdissection analysis of HRS cells in small lymphocytic lymphoma and follicular lymphoma has shown that these cells were clonally related to the coexisting NHLs.4,5 In the setting of peripheral T-cell lymphomas, the coexisting HRS cells were B-cell origin that were distinct from the T-cell lymphoma. The HRS cells were often infected by Epstein-Barr virus (EBV) and might evolve into large B-cell lymphoma.6 Mantle cell lymphoma (MCL) with coexisting HRS cells are very rare, and the clonal relationship between MCL cells and HRS cells has only been demonstrated in 3 cases from the previous reports.7–9 Here we present a clinicopathologic and genomic study of an additional case of MCL with coexisting HRS cells. We demonstrated the common origin of MCL and HRS cells by cyclin D1 expression and CCND1 rearrangement in both cell types.
The patient was a 63-year-old man with a past history of orbital lymphoma 8 year ago who now presented with facial swelling for 2 months. He had no fever, night sweats, or chills. On examination, there was bilateral parotid gland enlargement and increased left cervical adenopathy extending in a multinodular confluent chain at level 2 to level 5 and measuring 6×8×8 cm. There was bilateral axillary adenopathy and inguinal adenopathy, which remained stable. Computed tomography showed no orbital lesions, but progressive adenopathy above and below the diaphragm and adenopathy in the aortopulmonary window. An enlarged cervical lymph node was biopsied and showed MCL. The patient received 6 cycles of bendamustine and rituximab. The patient had a partial response with a small amount of residual disease in axillary areas.
The past history was notable for hepatitis A, hepatitis B, and bilateral orbital “MALT lymphoma” 8 years ago. The patient received 6 cycles of cyclophosphamide, vincristine, and prednisolone and radiation therapy (30.6 Gy to bilateral orbits) with complete remission of the orbital lymphoma. The specimen from the patient’s orbital mass biopsy was retrieved and reviewed and the diagnosis of the orbital lymphoma was revised as MCL.
MATERIALS AND METHODS
Immunohistochemistry and Chromogenic In Situ Hybridization
Tissue samples were formalin-fixed, paraffin-embedded (FFPE) and processed for routine hematoxylin and eosin stains. Immunohistochemistry was performed on FFPE sections using an automated immunohistochemical stainer according to manufacturer’s specification (Ventana Medical System, Tucson, AZ). Briefly, the immunoperoxidase stains were performed using a panel of antibodies including CD3 (2GV6), CD5 (SP19), CD10 (SP67), CD15 (MMA), CD20 (L26), CD23 (SP23), CD30 (Ber-H2), CD45 (RP2/18), CD79a (SP18), cyclin D1 (SP4-R), Ki-67 (30-9), and PAX5 (SP34). All the antibodies were from Ventana Medical System and were prediluted from the manufacturer. The antigen-antibody reaction was detected using a streptavidin-biotin-peroxidase system with diaminobenzidine as the chromogen. Chromogenic in situ hybridization for EBV-encoded RNA (EBER) (EBER-1 DNP probe; Ventana Medical System) was performed on FFPE sections using a Ventana automated stainer according to the manufacture’s specification (Ventana Medical System). Appropriate positive controls and negative controls were included in every case.
Karyotyping and Fluorescence In Situ Hybridization (FISH)
FISH for cyclin D1 rearrangement was performed on FFPE tissue sections using a IGH/cyclin D1 dual color dual fusion (DCDF) probe using the standard protocols according to the manufacturer’s specification (Vysis Inc., Downers Grove, IL). Briefly, 4 or 5 μm sections were deparaffinized, dehydrated, and pretreated using a VP2000 processor (Vysis Inc.). IGH/cyclin D1 DCDF probe was added to 70% formalin-treated denatured tissue and incubated at 37°C overnight. The sections were counterstained by 4’,6-diamidino-2-phenylindole (125 mg/mL). The signals were analyzed on an Olympus BX40 microscope with a fluorescein isothiocyanate, spectrum orange, and spectrum aqua (Vysis Inc.) filter set.
Polymerase Chain Reaction (PCR)
Immunoglobulin (IG) heavy chain gene rearrangement and T-cell receptor (TCR) gamma gene rearrangement were analyzed by PCR using BioMed2 kits and T-cell Receptor Gamma Gene Rearrangement Assay, respectively, according to the manufacturer’s specification (InVivoScribe Technologies, Carlsbad, CA). DNA was extracted from FFPE tissue sections and quantified using a spectrophotometer. The BioMed IGH kit frameworks 1, 2, and 3 master mix and TCR gamma mix 1 and 2 were combined with patient’s DNA and underwent PCR amplification according the manufacturer’s specification. The amplified products were analyzed on an ABI 310 gene analyzer (Applied Biosystems, Foster City, CA).
The cervical lymph node showed a diffuse infiltration of small lymphocytes which completely effaced nodal architecture. The lymphocytes had round to slightly irregular nuclear contours, stippled chromatin, inconspicuous nucleoli, and scant cytoplasm. Scattered mitotic figures were seen. Among the small lymphocytes, there were scattered large atypical cells. These cells were present in single forms and rarely in small clusters and showed nuclear morphology of HRS cells with large, single or multilobated nuclei and a prominent nucleolus in each nuclear lobe surrounded by a rim of pale cytoplasm. The HRS cells were frequently associated with small clusters of epithelioid histiocytes and were occasionally surrounded by T-cell rosettes. There was no increase of eosinophils, neutrophils or plasma cells (Figs. 1A, B). The bilateral orbital lesions from the previous biopsies were reviewed and showed diffuse lymphoid infiltration of small lymphocytes with similar morphology to the cervical mass. There were no HRS cells in the orbital lesions.
The small cell components in the cervical lymph node and the orbital masses showed a MCL immunophenotype with the expression of CD5, CD20, CD45, CD79a, Pax-5, and cyclin D1. These cells were negative for CD15, CD23, and CD30. The HRS cells in the cervical lymph node were positive for CD15, CD30, CD79a, cyclin D1, Pax-5, partially positive for CD5, and were negative for CD3, CD20, CD45, and EBER (Figs. 1C–J). The proliferation index by Ki-67 was 20% to 30% in the small cell components. There were several small foci of disrupted follicular dendritic cell meshwork in the background.
FISH for CCND1/IGH using DCDF probe revealed a variant pattern of CCND1 rearrangement with 1 CCND1/IGH fusion signal and 1 of each normal CCND1 and IGH signals. This pattern was interpreted as a positive CCND1/IGH rearrangement at derivative chromosome 14 with either a nondetectable or deleted rearrangement at derivative chromosome 11. The HRS cells showed identical CCND1/IGH signal pattern (Figs. 2A–C). Chromosome analysis of the cervical lymph node revealed an abnormal karyotype with t(11;14), deletion 6q, and trisomy 3: 47, XY,+3,del(6) (q15q21),t(11;14) (q13;q32) (Fig. 2D). PCR for IG heavy chain gene rearrangement was performed on both orbital masses and cervical lymph node, which showed identical clonal peaks. PCR for TCR gamma gene rearrangement was negative in both lesions.
We present a case of recurrent MCL containing scattered HRS cells showing morphologic and phenotypic features of cHL. The patient originally presented with orbital MCL and recurred 8 years after treatment. The HRS cells were only present in the recurrent disease but not in the original disease. The HRS cells showed morphology and phenotype indistinguishing from those in cHL but the inflammatory background of typical cHL was absent. We provided direct evidence of a clonal relationship between the MCL and HRS cells by demonstrating an identical cyclin D1 rearrangement in both MCL cells and HRS cells. Both cell types exhibited a variant FISH pattern with 1 CCND1/IGH fusion signal. This translocation was further confirmed by the presence of t(11;14) by chromosome karyotyping although the karyotype could not distinguish whether t(11;14) was present in MCL or HRS cells or both. We further proved that the t(11;14) resulted in overexpression of cyclin D1 in both MCL and in HRS cells. These results highly suggest that the MCL and HRS cells arose from the same origin and were clonally related. Interestingly, the HRS cells, having lost mature B-cell markers and gained Hodgkin lymphoma markers, still retained some antigens of MCL, such as CD5, in some cells.
The majority of the analyzed cells in our case showed only 1 fusion signal when analyzed by CCND1/IGH DCDF FISH probes. This pattern was seen in both MCL and HRS cells. This variant FISH pattern can be seen in a small subset of MCL and is likely caused by a variant breakpoint in the derivative chromosome 11 (likely more upstream) resulting in absence of the second fusion signal using the conventional FISH probes.10 In our case, the karyotype showed classic t(11;14) (q13;q32) with the presence of both derivative chromosome 14 and derivative chromosome 11, a pattern consistent with a reciprocal translocation. As cyclin D1 protein was uniformly expressed in MCL cells and HRS cells, we believe that the single fusion signal observed by FISH was on derivative chromosome 14 that represented a functional fusion protein from a productive IGH/CCND1 fusion gene.
Although uncommon, HRS cells have been reported in both B-cell and T-cell NHLs, and may or may not be clonally related to their NHL components. The HRS cells are usually present in scattered single forms or small clusters among the NHL cells. Typical background microenvironment of cHL is usually absent. The immunophenotype of HRS cells resembles that of cHL, which commonly expresses CD30 and CD15, and is often CD45−. In some cases, the HRS cells express EBV. These cells are more commonly found in small lymphocytic lymphoma,4 angioimmunoblastic T-cell lymphoma,6 and rarely in follicular lymphoma,5,11 marginal zone lymphoma,12 and follicular T-cell lymphoma.13 The HRS cells in these cases are considered as transformed cells rather than a separate cHL. The presence of HRS cells in NHLs may pose diagnostic challenges. These cases must be distinguished from composite Hodgkin and NHL, in which the Hodgkin and non-Hodgkin components have their own distinct morphologic and biological manifestation. The Hodgkin component in the composite lymphoma forms its own mass lesion and has all the features of cHL with HRS cells and the inflammatory background.
MCL with coexisting HRS cells in the same lesion is very rare. A search of English literature found a total of 7 reported cases. The HRS cells in all the reported cases as well as our case had similar morphology to those of cHL with large, unilobated or multilobated nuclei and prominent nucleoli in each nuclear lobe. The nuclei were often surrounded by ample cytoplasm. Phenotypically, the HRS cells invariably expressed CD30 with variable expression of CD15. Some cases were reported as CD20+ and CD45+, but the expression was usually weak and partial. EBV was positive in some cases, whereas negative in others. A spectrum of background was noted in the reported cases ranging from minimally associated inflammatory cells to a cHL type mixed inflammatory background. Nearly all the cases were reported as composite MCL and cHL (Table 1). HRS cells and MCL were clonally related in 4 cases (including our case), clonally unrelated in 2 cases, and unknown from the remaining 2 cases. On the basis of the histologic pattern and clonal relationship of the MCL component and HRS cell component, these cases can be grouped into 2 pathologic variants.
The first pathologic variant included the patient from our report and 2 other reported cases. The HRS cells were present in single forms and small clusters among MCL cells that often retained a vague nodular pattern.7,8 These cells were frequently associated with small clusters of histiocytes and might be surrounded by T-cell rosettes. However, eosinophils, neutrophils, and plasma cells were usually absent. The cases reported by Tinguely et al7 and Schneider et al8 had nodular pattern with scattered HRS cells and lymphocytic and histiocytic cells in a nodular proliferation of MCL, which resembled lymphocyte-rich cHL. Analysis of IG gene V regions by single cell microdissection or analysis of t(11;14) by FISH had proven a clonal relationship between MCL and HRS cells. The clonal relationship with identical t(11;14) rearrangements in HRS cells and MCL was also observed in our case. Interestingly, all 3 cases had a history of recurrent MCL or mantle cell leukemia, and the HRS cells were all reported in the recurrent disease. The pathology of the original diseases in the cases by Tinguely and colleagues and Schneider and colleagues was unknown. We had opportunity to review and compare the initial orbital lesions to the recurrent lesion. In our case, the HRS cells were only present in the recurrent disease but not in the initial orbital biopsy. This supports the notion that HRS cells are not an early event of the MCL but rather a transformed event after a long-standing disease.9,16 Whether the HRS cells are considered as transformed large cells as part of MCL or a new cHL in a composite MCL/cHL lymphoma is debatable. Tinguely and colleagues had shown that, in their study, in addition to t(11;14) in HRS cells, the HRS cells had mutated IG variable region gene rearrangement that was not present in MCL cells. This indicated that the GC microenvironment played role in the transformation of HRS cells. Whether this finding warrants the classification of HRS cells as the distinct component of composite lymphoma is to be determined. Nevertheless, these cases must be distinguished from true lymphocyte-rich cHL in which the HRS cells are located among the expanded, benign mantle zone cells.
The second pathologic variant included cases of true composite lymphoma as well as some cases with borderline features. These cases contained 2 distinct, separate MCL and cHL components, and likely represented true composite lymphomas. Caleo and colleagues reported 2 cases of composite MCL and cHL in which both lymphomas had their own distinct morphology and background. Both cases had nodules of cHL intermixed with diffuse or nodular areas of typical MCL. The cHL had distinct areas from MCL and contained HRS cells with T cell and histiocyte background. The cHL and MCL were proven to be different clones by clonal analysis of the microdissected cells.14 There was no information on treatment and outcome on their patients. Their results suggested that these cases contained separate components of MCL and cHL which might arise from different clones, and were best classified as composite MCL and cHL.
Two additional cases had a borderline morphology in which the HRS cells were more numerous and were associated with histiocytes, small lymphocytes, as well as eosinophils.9,15 One case had a morphology of mixed cellularity cHL and the second case had focal collagen bands suggesting a nodular sclerosis pattern.15 t(11;14) was detected in HRS cells in 1 case but cyclin 1 expression was negative.9 The clonal relationship between HRS cells and MCL in second case was unknown.15 It is unclear whether these cases are true composite lymphomas. The detection of t(11;14) in HRS cells in 1 case indicates that these cells are transformed cells from MCL. The emergence of additional features of cHL, such as eosinophils and collagen fibrosis, suggests that the HRS cells represent an ongoing transformation event which may give rise to true composite lymphoma over time.
It is possible that these pathologic variants merely represent a morphologic spectrum from a same process in different time points of the course. In the early stage of transformation, the HRS cells are mostly single cells without the associated “Hodgkin type” microenvironment. When disease progresses, the HRS cells proliferate, form larger clusters, attract inflammatory cells, and eventually result in their own mass lesion. The finding of clonal relationship between the MCL and HRS cells in a borderline case supports this hypothesis.9 It is speculated that, given time, the HRS cells will eventually progress into true cHL, but more cases will be needed to draw more definitive conclusion on how to properly classify, and hence optimally treat, these patients.
The prognosis of MCL with HRS cells is unknown due to the small numbers of cases reported and lack of long-term follow-up. Two patients received combination chemotherapy for high-grade NHL. One patient had a partial response and expired from infection and the other patient achieved complete remission.9,15 Another patient received an alternating Hodgkin and non-Hodgkin regimen. The patient had a good clinical response but died soon after from complications.8 Our patient received the therapy of lower intensity regimen targeted on MCL with a good response at a follow-up of 6 months. It appeared that, from the limited cases reported, consideration of treatment choices may be placed on not only the pathologic findings but also the clinical behavior. The “early lesions” may be approached by using a NHL regimen. If a diagnosis of bona fide composite MCL and cHL is established, a tailored regimen containing chemotherapy sensitive for both NHL and Hodgkin lymphoma appears to be the best approach.
1. Strum SB, Park JK, Rappaport H. Observation of cells resembling Sternberg-Reed cells in conditions other than Hodgkin’s disease. Cancer. 1970;26:176–190.
2. Shin SS, Ben-Ezra J, Burke JS, et al. Reed-Sternberg-like cells in low-grade lymphomas are transformed neoplastic cells of B-cell lineage. Am J Clin Pathol. 1993;99:658–662.
3. Gomez-Gelvez JC, Smith LB. Reed-Sternberg-like cells in non-Hodgkin lymphomas. Arch Pathol Lab Med. 2015;139:1205–1210.
4. Kanzler H, Kuppers R, Helmes S, et al. Hodgkin and Reed-Sternberg-like cells in B-cell chronic lymphocytic leukemia represent the outgrowth of single germinal-center B-cell-derived clones: potential precursors of Hodgkin and Reed-Sternberg cells
in Hodgkin’s disease. Blood. 2000;95:1023–1031.
5. Bayerl MG, Bentley G, Bellan C, et al. Lacunar and Reed-Sternberg-like cells in follicular lymphomas are clonally related to the centrocytic and centroblastic cells as demonstrated by laser capture microdissection. Am J Clin Pathol. 2004;122:858–864.
6. Quintanilla-Martinez L, Fend F, Moguel LR, et al. Peripheral T-cell lymphoma with Reed-Sternberg-like cells of B-cell phenotype and genotype associated with Epstein-Barr virus infection. Am J Surg Pathol. 1999;23:1233–1240.
7. Tinguely M, Rosenquist R, Sundstrom C, et al. Analysis of a clonally related mantle cell and Hodgkin lymphoma indicates Epstein-Barr virus infection of a Hodgkin/Reed-Sternberg cell precursor in a germinal center. Am J Surg Pathol. 2003;27:1483–1488.
8. Schneider S, Crescenzi B, Schneider M, et al. Subclonal evolution of a classical Hodgkin lymphoma from a germinal center B-cell-derived mantle cell lymphoma
. Int J Cancer. 2014;134:832–843.
9. Murray C, Quinn F, Illyes G, et al. Composite blastoid variant of mantle cell lymphoma
and classical Hodgkin lymphoma. Int J Surg Pathol. 2016;7:1–6.
10. Li JY, Gaillard F, Moreau A, et al. Detection of translocation t(11;14)(q13;q32) in mantle cell lymphoma
by fluorescence in situ hybridization. Am J Pathol. 1999;154:1449–1452.
11. Dilly M, Ben-Rejeb H, Vergier B, et al. Primary cutaneous follicle center lymphoma with Hodgkin and Reed-Sternberg-like cells: a new histopathologic variant. J Cutan Pathol. 2014;41:797–801.
12. Fung EK, Neuhauser TS, Thompson LD. Hodgkin-like transformation of a marginal zone B-cell lymphoma of the larynx. Ann Diagn Pathol. 2002;6:61–66.
13. Moroch J, Copie-Bergman C, de Leval L, et al. Follicular peripheral T-cell lymphoma expands the spectrum of classical Hodgkin lymphoma mimics. Am J Surg Pathol. 2012;36:1636–1646.
14. Caleo A, Sanchez-Aguilera A, Rodriguez S, et al. Composite Hodgkin lymphoma and mantle cell lymphoma
: two clonally unrelated tumors. Am J Surg Pathol. 2003;27:1577–1580.
15. Hayes SJ, Banerjee SS, Cook Y, et al. Composite mantle-cell lymphoma and classical Hodgkin lymphoma. Histopathology. 2006;48:621–623.
16. Papathomas TG, Venizelos I, Dunphy CH, et al. Mantle cell lymphoma
as a component of composite lymphoma: clinicopathologic parameters and biologic implications. Hum Pathol. 2012;43:467–480.