OUP user menu

Lymphoplasmacytic Lymphoma and Other Non–Marginal Zone Lymphomas With Plasmacytic Differentiation

Pei Lin MD, Thierry J. Molina MD, PhD, James R. Cook MD, PhD, Steven H. Swerdlow MD
DOI: http://dx.doi.org/10.1309/AJCP8FOIVTB6LBER 195-210 First published online: 1 August 2011

Abstract

Small B-cell lymphomas with plasmacytic differentiation frequently present diagnostic challenges. Session 3 of the 2009 Society for Hematopathology/European Association for Haematopathology Workshop focused on lymphoplasmacytic lymphoma (LPL). The submitted cases illustrated classic examples of bone marrow–based and nodal LPL and cases with atypical features, including unusual phenotypes or involvement of extranodal sites. Several cases showed varying degrees of overlap with marginal zone lymphoma, and, as acknowledged in the 2008 World Health Organization classification, a definitive distinction between these 2 possibilities cannot always be established. Session 6 of the workshop focused on other non–marginal zone lymphomas that may display plasmacytic differentiation. This session highlights the wide variety of neoplasms that enter into the differential diagnosis of small B-cell lymphomas with plasmacytic differentiation and demonstrates the use of clinical features and ancillary studies in establishing an appropriate diagnosis by 2008 World Health Organization criteria.

Key Words:
  • Lymphoplasmacytic lymphoma
  • LPL
  • Plasmacytic differentiation
  • Society for Hematopathology/European Association for Haematopathology Workshop

The differential diagnosis of small B-cell lymphomas with plasmacytic differentiation is frequently challenging because numerous types of B-cell lymphoproliferative disorders (LPDs) may show plasmacytic differentiation. Accordingly, 4 sessions of the 2009 Society for Hematopathology/European Association for Haematopathology Workshop were devoted to this topic. One session (session 3) was devoted to lymphoplasmacytic lymphoma (LPL), whereas another (session 6) included discussions of other B-cell lymphomas with plasmacytic differentiation. This report summarizes the conclusions of the latter 2 topics. The marginal zone lymphomas (MZLs) were discussed in 2 sessions (sessions 4 and 5), as summarized in the article by Molina et al in this issue of the Journal.1

Lymphoplasmacytic Lymphoma

The 2008 World Health Organization (WHO) classification defines LPL as a B-cell lymphoid neoplasm composed of small B lymphocytes, plasmacytoid lymphocytes, and plasma cells, usually involving bone marrow and sometimes lymph nodes and spleen, which does not fulfill the criteria for any of the other small B-cell lymphoid neoplasms.2 It is, therefore, a diagnosis made by exclusion. Although often associated with a paraprotein, usually of the IgM type, it is not required for the diagnosis. Waldenström macroglobulinemia (WM) is now defined as an LPL with any level of an IgM paraprotein. Thus, while LPL and WM are closely related entities, they are not synonymous. Nevertheless, much of what we know about LPL and WM is based on some studies that have focused on LPL and others that have focused on WM. Also, because the criteria for LPL and WM have not been uniform, one must recognize that our current paradigms are based on a somewhat heterogeneous group of cases. The discussion here relies on published data for LPL and WM.

A small B-cell lymphoma with plasmacytic differentiation was recognized in the Lukes/Collins (plasmacytoid lymphocyte lymphoma)3 and the Kiel (malignant lymphoma, lymphoplasmacytoid [immunocytic])4 classifications. Subsequently, the Kiel classification recognized 3 and later 2 variants—a lymphoplasmacytoid variant that had only plasmacytoid cells and could otherwise closely resemble chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL) and a lymphoplasmacytic variant that included a frank plasmacytic component. However, the criteria have, in essence, become much stricter during the last several decades as the diagnosis became an exclusionary one and as the list of other B-cell lymphomas with plasmacytic differentiation has grown. Many of the lymphoplasmacytoid cases are now included in the CLL/SLL category. CLL/SLL cases may even show frank plasmacytic differentiation, and infrequent cases of LPL may only show plasmacytoid cells. Follicular lymphomas (FLs) may show plasmacytic differentiation,5 as may mantle cell lymphomas (MCLs), albeit rarely.

The most problematic differential diagnosis, discussed briefly here and more extensively elsewhere,1 is with MZLs with plasmacytic differentiation. Because it may be impossible to determine if a lymphoma is an LPL or one of these other small B-cell lymphomas, the current WHO criteria state that “Because the distinction between LPL and one of these other lymphomas [with plasmacytic differentiation], especially some MZLs, is not always clear-cut, some cases may need to be diagnosed as a small B-cell lymphoma with plasmacytic differentiation and a differential diagnosis provided.”2 Integration of the clinical findings with the morphologic, phenotypic, and molecular/cytogenetic findings is also very important in arriving at the “best” diagnosis, for example, as was highlighted in 1 workshop case in which a previously diagnosed LPL was present in a conjunctival biopsy sample, a typical site for a mucosa-associated lymphoid tissue (MALT) lymphoma.

Other diagnostic problems include the differential diagnosis with diffuse large B-cell lymphoma (DLBCL) because some LPLs have moderately numerous large transformed cells, DLBCL can have plasmacytic differentiation, and there are no clear-cut criteria for the distinction unless there is a sheet of transformed cells. Plasmacytic neoplasms, which may coexist with lymphoid neoplasms, are sometimes another problematic differential diagnosis because it may be difficult to determine if both components, which might even share the same light chain, represent 1 neoplasm or 2. Furthermore, some plasma cell myelomas, such as those with an IGH/CCND1 translocation, may appear lymphoplasmacytic and closely resemble a lymphoma.6,7 Finally, some clonal or oligoclonal lymphoplasmacytic proliferations associated with hepatitis C virus (HCV) infection and type II mixed cryoglobulinemia do not behave like overt neoplasms,1,810 but on the other hand LPL is also one of the lymphomas that has been associated with HCV in some but not all studies.11 Some HCV-associated lymphoplasmacytic proliferations have also been successfully treated with antiviral agents.

Diagnostic Reproducibility

One of the questions raised at the workshop was “How often can a confident and reproducible diagnosis of LPL be made?” The Non-Hodgkin’s Lymphoma Classification Project study found that the “experts” agreed with the ultimate consensus diagnosis of LPL in a disappointing 56% of the cases, even when they had the histologic, immunophenotypic, and clinical data.12 The workshop findings were more encouraging, with the caveats that they are biased toward the more easily categorized bone marrow–based cases and that LPL was viewed essentially as the same diagnosis as probable LPL. On the other hand, data in some cases were more limited than they would be in a clinical setting. Among the 30 cases included in this section of the workshop, there were 23 in which the consensus panel diagnosis was LPL or probable LPL. Of those cases, 13 were in full agreement with the presenter, in 3 the panel members thought they could be more definitive than the contributor, and in 5 the contributor made a bolder diagnosis than the panel. There were only 2 cases with slightly greater differences of opinion, 1 in which the contributor favored the diagnosis of atypical CLL and 1 in which the panel thought that concern for transformation was more appropriate than an outright diagnosis of DLBCL.

There were 7 additional more problematic cases. In 2 of these 7 cases, the panel favored a descriptive diagnosis of a small B-cell lymphoma with plasmacytic/plasmacytoid differentiation (1 in which CD5 was also mentioned). One was submitted as an example of an “IgG-secreting B-cell lymphoplasmacytoid leukemia,” an indolent disorder with leukemic involvement, very frequent splenomegaly, and an IgG paraprotein, often together with an IgM paraprotein, trisomy 12 in 60%, and 13q14 deletions in 40%.13 Another unusual case with an osteolytic bone lesion, a negative bone marrow examination, and increased large cells raised concern for transformation to large B-cell lymphoma. The other 5 cases are discussed in the following section.

Morphologic Features

Bone marrow samples demonstrate an intertrabecular small lymphocytic interstitial infiltrate with or without nodules and with or without paratrabecular and diffuse infiltrates Image 1A. Variable numbers of plasmacytoid lymphocytes and plasma cells, often with periodic acid–Schiff+ intranuclear pseudoinclusions (Dutcher bodies), also must be present Image 1B and Image 1C. Although most cases usually have predominantly lymphoid cells, in some cases, as seen in the workshop, the plasma cells predominate.14 Furthermore, although one generally expects to see the plasma cells and lymphoid cells intermingled, a recent study discussed at the workshop found that light chain–restricted interstitial plasma cell clusters considered a part of the LPL but distinct from the lymphoplasmacytic infiltrates were identified in 19 of 27 cases of LPL in the marrow and were the only plasmacytoid component detected by immunohistochemical studies in 2 of the 19 cases.14 On the other hand, this observation does not negate the observation that plasma cell neoplasms may coexist with lymphoid neoplasms so that when distinct monotypic plasma cell infiltrates are present in the setting of a lymphoid neoplasm, the possibility of 2 separate neoplasms must be considered. This distinction accounted for 3 of the 7 more problematic cases in this session. In 1 of these cases, only the plasmacytic component remained after therapy. While this heightened concern that there was a distinct plasmacytic neoplasm in this case, elimination of clonal B cells but with residual clonal plasma cells is reportedly frequent following at least fludarabine therapy for LPL/WM.15 In addition to looking for pathologic features typically associated with plasma cell myeloma, this problem also highlights the importance of integrating clinical features in arriving at the best diagnosis for a given patient.

As seen in the workshop cases, other typical features of LPL include the presence of prominent mast cells, as often most easily recognized in the spicules in the aspirate smears or with special stains, such as a Giemsa stain or tryptase immunohistochemical stain in the histologic sections, and hemosiderin (Images 1A and 1C). Mast cells may actually support the growth of the LPL.16 LPL may also be associated with immunoglobulin deposition, amyloid, or crystal-storing histiocytosis.2 Features that would exclude marrow involvement by an MZL with plasmacytic differentiation were not specifically identified at the workshop, but the presence of follicles with follicular dendritic cells and lymphoid cells with abundant pale cytoplasm are 2 pathologic features not in favor of an LPL. Given that marrow-based LPL can be associated with adenopathy and splenomegaly, an incorrect diagnosis of LPL from a marrow examination, however, may never become clinically apparent, something that at present may not be critical.

Image 1

Lymphoplasmacytic lymphoma in the bone marrow from patients with IgM paraproteins. A (Case 16) There is an interstitial lymphoplasmacytic infiltrate (H&E). B (Case 16) At higher magnification, the lymphoid and plasmacytic cells are seen in addition to hemosiderin (arrow head) and at least 1 mast cell (arrow) (H&E). C (Case 138) A lymphoid-rich lymphoplasmacytic infiltrate is also seen in this aspirate smear, together with a mast cell (arrow head), hemosiderin (tarrow), and some maturing hematopoietic elements (Wright-Giemsa).

In contrast with the cases of LPL diagnosed when a bone marrow examination is performed in a patient with an IgM paraprotein, the diagnosis of LPL is often much more problematic when a lymphoplasmacytic neoplasm is found in a biopsy sample from an extramedullary site. In fact, in 2 of 7 more problematic cases in this section, the panel could not agree on whether the lymph node biopsies were most like an LPL or an MZL with plasmacytic differentiation. Three patterns of lymph node involvement by LPL have been described with the most classic being lymph nodes with intact sinuses, a relatively monotonous lymphoplasmacytic proliferation, and, sometimes, small residual follicles but without large germinal centers or prominent follicular colonization. Other reported patterns include a vaguely nodular more polymorphous proliferation and a miscellaneous group of cases that simply do not fulfill the criteria for any of the other small B-cell lymphomas.17 These cases may even include small foci of possible marginal zone cells. The panel accepted one lymph node case as LPL that had monocytoid features Image 2 and another that was more plasmacytoid than plasmacytic. In addition, while the presence of prominent colonized follicles is a feature in favor of an MZL rather than an LPL, the panel accepted 1 case with lymph node and marrow involvement that had numerous follicular structures throughout, as highlighted with a CD21 stain Image 3.

Image 2

(Case 248) Lymphoplasmacytic lymphoma (IgM κ) in lymph node with focal monocytoid areas (unanimous panel diagnosis) in patient with IgM paraprotein (H&E).

The workshop experience also highlights the importance of being able to “officially” use the diagnosis of small B-cell lymphoma with plasmacytic differentiation. The panel chose this diagnosis in the 2 cases in which panel members could not agree on whether to even favor an LPL or an MZL. In one case of a patient with an IgM paraprotein of 1.4 g/dL, the lack of a frankly plasmacytic component and the presence of a partially perifollicular growth pattern was considered unacceptable for an LPL by some but not by others Image 4. In another case with an IgM paraprotein of 5.48 g/dL, the contributor thought the lymphoma present was consistent with an MZL, but once again, the presence of multiple, small, follicular-type structures that included some CD21+ follicular dendritic cells plus other probable lymphoid staining and light chain restriction, mostly by lymphoid and plasmacytoid cells rather than plasma cells, was considered unlike LPL by some panel members but considered very acceptable by others. There was hemosiderin present like in LPL.

Until further positive diagnostic features for LPL are found, distinction from MZLs, particularly of the nodal type, that may lack neoplastic cells with abundant pale cytoplasm, lack a distinctive phenotype, and often lack specific cytogenetic abnormalities will rest in part on one’s philosophical approach to this dilemma rather than on any hard data. Some will put the burden of proof on demonstrating that a case is not an LPL and, therefore, diagnose a higher proportion of cases as LPL; others will use MZL as a default term for lymphomas that might be of marginal zone origin; and others with a greater fear of being “wrong” will use the diagnosis of a small B-cell lymphoma with plasmacytic differentiation more freely.

Another, as yet unanswered, morphologic issue with LPL is what the criteria should be for transformation to a DLBCL. LPLs, like the MZLs, may have increased numbers of transformed cells without being considered to have undergone transformation. These cases may have a worse prognosis.18 Another issue raised by one of the cases in the workshop is whether some cases may transform into blastoid-appearing cells, similar to the type of transformation described in some FLs.19

Immunophenotype

LPL is expected to include light chain–restricted lymphocytes and plasma cells. The lymphoid cells are CD20+, but the staining may be weak and additional B-cell stains such as CD79a and PAX5 stains may be helpful. They are usually IgM+, sometimes IgG+, and rarely IgA+. Although CD5, CD10, and CD23 have traditionally been considered to be negative, 1 recent study reports CD5 expression in 43% of cases and CD23 expression in 52% of cases, both usually with only partial expression (with no CD10+ cases).14 In other words, even the CD5+/CD23+ LPL cases will usually still not have a phenotype typical for CLL/SLL. Others report a typical phenotype in 60% to 80% of cases, with the most frequent exception being CD23 expression that is usually dim or partial but with a very small proportion CD5+ or, less commonly, CD10+.20 Among the definite or probable LPL cases submitted to the workshop, 17% were CD5+, 16% CD10+, and 58% CD23+. In some cases, CD5 and CD10 expression varied over time. One must recognize, however, that the workshop cases are probably enriched for “atypical” findings, and, also, one must consider technical and interpretive interlaboratory variation. None of the CD10+ cases were shown to be BCL6+ with 1 definitely negative, 1 probably negative, and 1 in which artifactual negativity was a concern. LPLs are often CD25+ and CD38+. Among the 7 cases in this session with greater diagnostic uncertainty, except for CD23 expression, there was a higher proportion of cases with a phenotype unlike typical LPL (43% CD5+, 43% CD10+, and 20% CD23+).

Image 3

(Case 226) Lymphoplasmacytic lymphoma in lymph node in a patient with a high serum IgA level. A, Note the diffuse, relatively monomorphic-appearing proliferation with sinus preservation and some paler-appearing vaguely nodular foci (H&E). B, A CD21 stain highlights many small aggregates of follicular dendritic cells, some of which represented germinal centers (immunostain with hematoxylin counterstain). C, The relatively monotonous proliferation of small lymphocytes, variably plasmacytic small cells, and occasional transformed cells is associated with prominent hemosiderin (H&E).

The observation that LPL can have atypical phenotypic findings highlights how, although immunophenotypic studies can help make a precise diagnosis, they should not be totally relied on. This is particularly a danger in situations in which there are limited morphologic correlates, as with fine-needle aspirates or peripheral blood samples. This was illustrated in a workshop case diagnosed as LPL by the submitter and panel, but in which a CD5+ phenotype seemingly led to the diagnosis of CLL at a prestigious outside institution.

Image 4

(Case 35) Small B-cell lymphoma with plasmacytic differentiation in a patient with an IgM paraprotein. A, The lymph node demonstrates a perifollicular and more diffuse lymphoplasmacytic proliferation with extensive diffuse perinodal infiltration (H&E). B, CD10, present on the neoplastic cells, highlights the striking perifollicular growth pattern in some areas. C, BCL6 is expressed only on the germinal centers that were surrounded by the CD10+ neoplastic cells (B and C, immunostain with hematoxylin counterstain).

The plasma cells in LPL are CD138+ and have recently been reported to show abnormal coexpression of CD138 and PAX5 in 74% of cases (mean, 23% of CD138+ cells).21 This finding is in contrast with the CD138+ cells in MZL and most plasma cell neoplasms. In another study, without double staining, the plasmacytic cells were also described as being positive for PAX5 but with “some variability in PAX5 reactivity.”14 It is also consistent with an earlier gene expression study report of PAX5 overexpression in the plasma cells of WM and overexpression of 3 downstream target genes.22 This study also showed decreased expression of the IRF4 and BLIMP1 genes in WM plasma cells. WM is associated with overexpression of IL6.22,23

Another feature reported to help in the distinction between LPL and a plasma cell neoplasm is the frequent expression of both CD45 and CD19 on the plasma cell component in most LPLs, although the staining may be diminished compared with normal B cells.14 A CD19– plasmacytic component, a phenotype like that expected in plasma cell myeloma, was found in only 2 cases, both of which were associated with an IgG paraprotein (1 that also had a minor IgM paraprotein). Four other cases with non-IgM paraproteins had CD19+ plasma cells. Of the definite/probable LPL cases in the workshop, 2 of 3 had CD19+ plasma cells (IgM, IgM + IgG) and the third had CD19– plasma cells (IgM). This latter case highlights the phenotypic heterogeneity of the plasma cells, even in the IgM+ cases. Flow cytometric studies in another case in which there was concern for an independent plasma cell neoplasm also had CD19– plasma cells. Finally, proteomic studies have suggested the presence of novel dysregulated proteins in WM.24

Cytogenetic and Molecular Findings

Like other B-cell neoplasms, the B cells in LPL are expected to have clonally rearranged immunoglobulin genes, usually with somatic hypermutation, and reported biased VH usage.25 Biclonal cases with related or unrelated clones are not rare, and class-switch rearrangements can also occur in a minority of cases.2527 In addition, however, cytotoxic T-cell clones are also frequently present, particularly in patients who have not received nucleoside analog therapy.28

In contrast with the teachings of the 2001 WHO “Bluebook,” PAX5 translocations are rarely, if ever, found in LPL,29 even though this association resurfaces occasionally in publications at least as recent as 2009. It was recently suggested that the translocation is present in nonsecretory-type LPL based on 2 cases, both of which manifested with nodal disease.30 Other B-cell lymphoma–associated translocations, including CCND1, MALT1, and BCL10, are also not found in LPL and BCL2 translocations are only rarely reported.17

Also unlike many MZLs, trisomies of chromosomes 3, 12, and 18 are infrequently found; however, trisomy 4 has been reported in 20% of cases.31 The most frequent recurrent abnormality is the very nonspecific 6q21 deletion, which is reported in up to 63% of bone marrow–based LPL/WM cases but appears to be at least much less common in tissue-based LPL cases.20,32,33 It has also been reported that the minimally deleted region is actually between 6q23 and 6q24.3 but with most cases showing a large deletion in 6q.34 This deletion is reportedly absent in “IgM monoclonal gammopathy of undetermined significance.”34 Gene profiling studies of WM suggest a homogeneous gene expression profile independent of 6q deletions and more like CLL or normal B cells than myeloma.23 Furthermore, 6q deletions are not of prognostic importance, although they have been reported to be associated with “features of adverse prognosis.”35,36

When contemplating the diagnosis of LPL, it is also important not to exclude the diagnosis because of a cytogenetic abnormality that may be more commonly associated with another type of B-cell neoplasm but that is, in fact, not a specific finding. In 1 workshop case with variable CD5 expression, the demonstration of del(13)(q14.3) by cytogenetic fluorescence in situ hybridization (FISH) studies, not at all unreasonably, led to the suggested diagnosis of “atypical CLL with an M component.” However, 13q deletions, including some specifically involving the q14 band, often together with other abnormalities, have been reported in WM and in LPL. A minimally deleted region in 13q14 involving MIRN15a and MIRN16-1 has been reported in 10% of patients with WM.37 Again, complicating the differential diagnosis of LPL from MZL, this high-resolution array–based comparative genomic hybridization study found that among the recurrent chromosome abnormalities identified in WM, most were also features of MZL (deletions of 6q23 and 13q14 and gains of 3q13-q28, 6p, and 18q).37,38 These abnormalities were found, however, in only 10% to 38% of cases. Other recurrent abnormalities found in WM (4q and 8q gains in 12% and 10% of WM cases, respectively), however, are not commonly reported abnormalities in MZL.38

Clinical Features

Patients with LPL usually have symptoms related to anemia and are found to have marrow and sometimes peripheral blood involvement, although some patients have more lymphomatous manifestations, leading to nodal or other extramedullary biopsies.2 Among patients with WM, 15% to 30% will have splenomegaly, hepatomegaly, and/or adenopathy. There is a familial predisposition.39 Many cases of LPL are associated with a paraprotein, often but not always of the IgM type. Cases with an IgM paraprotein, as noted earlier, fulfill the criteria for WM. The paraprotein may lead to the hyperviscosity syndrome (in up to about 30% of cases), mixed cryoglobulinemia (in about 20% of patients with WM), deposits in the skin or gastrointestinal tract, or coagulopathies.2 Central nervous system manifestations, known as the variably defined Bing-Neel syndrome, may occur with or without a lymphoplasmacytic infiltrate.40

It should be recognized that some cases that fulfill the criteria for LPL also would fulfill the criteria for IgM monoclonal gammopathy of underdetermined significance (MGUS) if they have a serum IgM paraprotein of less than 3 g/dL, less than 10% lymphoplasmacytic marrow infiltration, and an absence of constitutional symptoms, anemia, hyperviscosity, lymphadenopathy, and hepatosplenomegaly related to the “lymphoplasmacytic process.”41 Other cases with an IgM paraprotein of more than 3 g/dL and/or more than 10% marrow infiltration but still without constitutional symptoms, symptomatic anemia, and hyperviscosity have been designated as “smoldering WM.”41 Patients with IgM MGUS have a markedly increased risk for the development of WM (262-fold), with some other patients having evolution to other B-cell neoplasms, and smoldering WM has a 55% risk of progression to WM at only 5 years.41 Some molecular differences have been reported between IgM MGUS and WM.25

Among the definite or probable workshop cases of LPL, 82% had an IgM paraprotein (<1 g/dL, 4%; 1–3 g/dL, 35%; and >3g/dL, 43%). Dual IgM and IgG paraproteins were seen in 9% of cases, including 1 with IgM and γ heavy chains only. Isotype switching is rare in LPL/WM but has been reported.25,27,42 An IgA paraprotein was present in 9%, and there were no patients without any paraprotein. Among the 7 cases included in the “LPL” session in which there was a greater degree of diagnostic uncertainty, 3 had an IgM paraprotein, 1 had an IgG paraprotein, 1 had IgG and IgA, 1 had IgG and IgM (IgG > IgM), and 1 had none. Similarly, in a series of extramedullary-based LPLs, we found IgG paraproteins in the cases with atypical morphologic features and a seemingly increased proportion of IgG+ neoplasms in this group.2

LPL is considered one of the “low-grade” B-cell lymphomas; however, as noted, some investigators accept large numbers of transformed cells, and in some centers, the cases are treated with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) or rituximab and CHOP, like one of the more aggressive B-cell lymphomas.2,43 An international prognostic scoring system for WM44 and treatment recommendations for WM45 have recently been published, but, not unexpectedly, therapeutic strategies are undergoing constant reevaluation.46 Transformation to a “high-grade” B-cell lymphoma is reported in about 5% to 8% of patients and is reportedly more frequent in patients who have been treated with nucleoside analogs.47,48

There was limited discussion at the workshop concerning the clinical importance of distinguishing LPL from the other lymphomas with plasmacytic differentiation that would be in the differential diagnosis. Nevertheless, because of very different therapeutic strategies and expected clinical course, it would be universally agreed that LPL, a low-grade B-cell lymphoma, must be distinguished from the “aggressive” B-cell lymphomas that may have plasmacytic differentiation, such as DLBCLs, plasmablastic lymphomas, and infrequent Burkitt lymphomas or MCLs. Distinction from the more indolent small B-cell lymphomas with plasmacytic differentiation may be more difficult to justify based on documented different therapeutic approaches and/or survival differences. However, therapies for LPL may differ from those for other B-cell neoplasms, and, in general, there is a move toward more targeted therapies whose effectiveness will depend on what abnormalities are present in specific oncogenic molecular pathways. In fact, Treon [a workshop participant] et al49 criticized the 2001 WHO classification because it “left open the possibility that patients with WM could have any of the recognized IgM-secreting lymphoid neoplasms as their underlying diagnosis, thereby hampering the conduct or interpretation of clinical trials involving WM patients.” This emphasized the clinical interest that exists in not grouping together lymphomas just because they had a plasmacytic component and secreted IgM. It was the reason that the criteria for WM were altered in the 2008 classification to require a specific diagnosis of LPL.

Summary

LPL remains an exclusionary diagnosis because there are no unique and uniform clinically applicable features that characterize the small lymphoid cells or the plasma cells, although PAX5 coexpression by CD138+ plasma cells may be a characteristic feature. The diagnosis can, however, be made, although often with somewhat less certainty than with many of the other diagnoses we make in hematopathology Table 1.

Although the diagnosis is often made with greater confidence from bone marrow examinations from patients with IgM paraproteins, the unanswered question remains, “How many of these cases would end up with a different or less precise diagnosis had an enlarged lymph node been excised or a splenectomy been performed? In fact, the lymph node/extramedullary tissue biopsy specimens submitted to the workshop highlighted our continuing lack of sufficient criteria to reproducibly diagnose LPL with a high level of confidence at extramedullary sites. Nevertheless, there were cases in which the submitter and panel members could agree on the diagnosis with the most classic cases being relatively monomorphic appearing without prominent follicles and with sinus retention.

View this table:
Table 1

The workshop highlighted the need for the equivocation now allowed by the 2008 WHO classification for some small B-cell lymphomas with plasmacytic differentiation. It also highlighted the basic principles of the WHO classification, namely that disease entities are defined by many parameters, including clinical features. Finally, it must be acknowledged that this discussion reflects the current literature and the current practice of the hematopathology community rather than true evidence-based medicine. Challenges obviously remain for the future.

Other Non–Marginal Zone B-Cell Lymphomas With Plasmacytic Differentiation

A variety of LPDs may have plasmacytic differentiation with or without an associated serum M protein.5061 Session 6 of the workshop was designed to illustrate the morphologic spectrum and diagnostic issues of B-cell or plasma cell neoplasms “other than” LPLs or MZLs. Many cases highlighted the diagnostic challenge when a neoplasm manifests in an unusual clinical setting or lacks morphologic or immunophenotypic features that are characteristic of a specific type of B-cell neoplasm. Recurrent issues related to diagnosis and differential diagnosis of these other LPDs are discussed and summarized in this report.

Distinct B-Cell Lymphomas Associated With an M Protein or Plasmacytic Differentiation

Among 26 cases reviewed in session 6, CLL/SLL and FL represented the largest subgroups. Of the cases, 6 were classified as CLL/SLL, 6 as FL, and 1 as a composite of SLL and FL with large cell transformation. All 6 CLL cases had typical morphologic features and immunophenotypes. Serum IgM levels were elevated to the range of what is observed in LPL (up to 3,200 mg/dL [32,000 mg/L] in case 65). These cases confirmed that serum IgM levels overlap with those seen in LPL. Two cases of CLL had a serum M protein of the IgG type.

All 5 cases of FL (cases 39, 117, 171, 207, and 333) exhibited plasmacytic differentiation, although levels of serum M proteins were unspecified. One patient (case 117) had a monoclonal IgA protein. There were examples of FL in which the plasmacytic component was predominantly intrafollicular and others in which it was mostly interfollicular.57 Two cases (cases 39 and 207) also had a marginal zone growth pattern or increased monocytoid B cells mimicking MZL. The IGH/BCL2 fusion demonstrated by FISH in the B cells and CD138+ plasma cells confirmed the diagnosis of FL and supported that plasma cells were part of the same process in 1 case (case 171) Image 5.

B-Cell Lymphomas With Unusual Morphologic Features and Immunophenotype

The cases lacking classical morphologic and immunophenotypic features of CLL or FL were challenging. The best approach was to consider overall clinicopathologic features and provide the “best-fit” diagnosis. The neoplastic cells in 3 cases (cases 153, 174, and 205) of small B-cell lymphoma associated with elevated serum IgM levels (3,189, 3,980, and 6,678 mg/dL [31,890, 39,800, and 66,780 mg/L], respectively) were characterized by immunophenotypes of CD5+, CD23dim/–, cyclin D1–, respectively. The differential diagnoses of atypical CLL vs MZL or LPL with CD5 expression were considered. The presence of splenomegaly raised the possibility of splenic MZL, but splenomegaly is also a finding in a subset of patients with CLL. Expression of dim CD20 and surface immunoglobulin light chain expression are considered characteristics of CLL, although these features are neither specific nor universally observed in every case of CLL. Deletion of chromosome 13q is a nonspecific finding seen in a variety of neoplasms besides CLL. In the end, the absence of typical morphologic features such as proliferation centers (pseudofollicles) and the presence of high levels of serum IgM were considered to favor a diagnosis of CD5+ LPL over atypical CLL by the review panel.

Only a descriptive diagnosis of “small B-cell lymphoma with plasmacytic differentiation” was possible in certain cases. The neoplastic cells in case 50 were composed of small lymphocytes and plasmacytoid lymphocytes Image 6. The patient also had splenomegaly (14 cm) and autoimmune hemolytic anemia as a result of a monoclonal protein that reacted as a nonspecific cold agglutinin. Immunophenotyping identified 2 populations of monotypic B cells with κ light chain restriction: one was CD45-bright, CD5-moderate, and CD23-dim, and the other was CD45-moderate, CD5-dim, and CD23–. The latter population, presumably corresponding to the plasmacytoid lymphocytes, had higher levels of cytoplasmic μ expression. The review panel opted for a descriptive diagnosis of small B-cell lymphoma with plasmacytic differentiation and provided differential diagnoses of atypical CLL vs LPL vs splenic MZL.

Case 278 was presented as an example of B-cell lymphoma with folliculotropism that resembled human herpesvirus 8 (HHV-8) and Epstein-Barr virus (EBV)-associated germinotropic LPD originally described by Du and colleagues.58 The follicles in the lymph node were infiltrated by monotypic B cells with marked plasmacytic differentiation positive for BCL2 and MUM1 and negative for CD138, CD10, and BCL6. FISH analysis also showed no evidence of BCL2/IGH translocation. HHV-8 and EBV infection or autoimmune reactions were excluded as possible causes. The panel provided only a descriptive diagnosis for this very unusual case.

Image 5

(Case 171) Follicular lymphoma with marked plasmacytic differentiation. A, A neoplastic follicle was surrounded by plasmacytoid/plasmacytic cells (H&E). B, The plasmacytic components were most prominent in the interfollicular region as highlighted by a CD138 immunostain. C, The neoplastic cells in the follicle and the plasma cells are positive for BCL2 (immunostain). D, t(14;18) involving the major breakpoint region (MBR) was demonstrated in the follicular B cells and the interfollicular plasma cells.

A discordant phenotype can be observed between tumors involving bone marrow and extramedullary sites, complicating their classification. In case 198, a serum IgG λ M protein of unspecified quantity was identified. A bone marrow biopsy revealed a paratrabecular and interstitial lymphoid infiltrate of cyclin D1–negative small B cells and 20% cyclin D1–positive plasma cells. A subsequent splenectomy specimen found a CD5– small B-cell lymphoma with plasmacytic differentiation. The B cells and plasma cells in the spleen were positive for cyclin D1. Conventional cytogenetic analysis found 46,XX,9qh+9qh-c (a constitutional finding of uncertain significance) in the bone marrow sample and t(11;14)(q13;q32) in the spleen sample. FISH detected IGH/CCND1 fusion in the lymphocytes and plasma cells of the spleen, supporting the rare diagnosis of MCL with plasmacytic differentiation and the absence of CD5 expression. The discordance between bone marrow and spleen in cyclin D1 expression of B cells and cytogenetics suggested 2 different diseases, but the detection of cyclin D1 in the plasma cells in the bone marrow and spleen indicated that at least the plasmacytic cells most likely reflected a single disease process involving both sites.

Image 6

(Case 50) Low-grade B-cell lymphoma with plasmacytic differentiation. A, The peripheral blood showed small B cells and plasmacytoid cells (Wright stain). B, Flow cytometry illustrated 2 populations of B cells: one was CD45-bright, CD5-moderate, and CD23-dim, and the other was CD45-moderate, CD5-dim, and CD23–. The CD45– population expressed a higher level of cytoplasmic μ. APC, allophycocyanin; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein; SSC, side scatter.

Determining the Clonal Relationship of Plasma Cells and B Cells

A recurrent issue in working up a B-cell neoplasm showing a prominent plasmacytic component is to determine the relationship of the B cells to the plasma cells. Is the plasma cell component part of the same neoplasm or an independent process? The most common setting is the development of a significant component of plasma cells during the course of disease that coincides with bone pain or radiographic evidence of a lytic lesion and a serum M protein of the IgG or IgA type in a patient with CLL or other small B-cell lymphoma. The plasma cell component can be of low-grade Marschalko or small cell type or of high-grade plasmablastic type. To make a distinction of CLL and concurrent plasma cell myeloma vs CLL with plasmacytic differentiation or plasmablastic transformation, molecular analysis for the IgH sequence may be required to determine the clonal relationship between the B cells and plasma cells.

Case 344 involved a patient with CLL in whom bone lesions developed during the course of disease. Bone marrow biopsy demonstrated a new population of cytoplasmic immunoglobulin light chain–restricted plasmacytoid cells (CD138+, IRF4/MUM1+, cyclin D1–, CD20–, CD23–, CD56–, and cytoplasmic immunoglobulin λ+) in addition to the monotypic B cells (CD5+, CD19+, CD20 dim+, CD23+, and surface immunoglobulin λ+ dim), raising the possibility of a concomitant plasma cell neoplasm. Lymphocytes from peripheral blood and plasma cells selected from bone marrow by magnetic bead column were tested for the FR3 region of the IGH and of the VJ region of IGK by polymerase chain reaction (PCR). An identical clone was identified, supporting the diagnosis of CLL with plasmacytic differentiation. A subsequent biopsy of the bone lesion revealed a diffuse large cell infiltrate indicating Richter transformation of the CLL.

Case 340 involved a patient with a CD5+ small B-cell lymphoma in whom a pathologic fracture of the left humerus developed. A biopsy of the left humerus found foci of a plasmablastic infiltrate adjacent to aggregates of small lymphocytes Image 7. The plasmablasts were positive for CD138, VS-38c, and λ and negative for CD56, cyclin D1, EBV, and HHV-8. Cytogenetic analysis showed a complex karyotype including t(8;14)(q24.1;q32) but no t(11;14). PCR found a monoclonal IGH gene rearrangement. Because the small B cells and the plasmablasts had immunoglobulin λ light chain expression and only a single clone was identified by PCR, the case was classified as plasmablastic transformation of a small B-cell lymphoma. Cases 340 and 344 illustrate that the onset of lytic or destructive bone lesions in a small B-cell leukemia/lymphoma may indicate development of large B-cell lymphoma, and MYC rearrangement may be implicated as in other plasmablastic lymphomas also.

Analyzing the clonal relationship between B cells and plasma cells in FL and MCL was more straightforward. Two cases presented at the workshop illustrated how FISH analysis confirmed the diagnosis of MCL and FL with plasmacytic differentiation using the t(11;14) and t(14;18) probes, respectively. This was particularly useful in the case of MCL (case 198) in which the neoplastic cells were CD5–.

Image 7

(Case 340) Plasmablastic transformation of low-grade B-cell lymphoma. The patient had a pathologic fracture of the left humerus. Biopsy of the lesion found small lymphoid cells (A) and cells with marked plasmacytic differentiation (B, H&E) and plasmablasts (C, H&E). MYC gene rearrangement was detected.

In many clinical settings, microdissection or sorting is not readily available. Determining the clonal relationship of B cells and plasma cells relies on comparing immunoglobulin light chain patterns between the 2 components by flow cytometry or immunohistochemical analysis. Cells with an identical immunoglobulin light chain are usually presumed to be derived from a common origin. Absence of cyclin D1 and CD56 expression is also considered evidence not in favor of concurrent myeloma, although it does not completely exclude that possibility because 30% and 60% to 75% of myelomas are positive for cyclin D1 and CD56, respectively.

Plasma Cell Neoplasms Involving Lymph Nodes or MALT Sites

The second largest subgroup of cases in session 6 were neoplasms composed exclusively or predominantly of monotypic plasma cells involving primarily nodal (cases 78 and 338) or extranodal tissues where MALT lymphomas are known to arise (cases 62, 108, and 167). The B cells were shown to be polytypic by immunophenotyping or minimal and confined to the residual follicles. The lymphadenopathy could be localized (6 × 5-cm left hilar mass) and slowly progressing, associated with a low level of serum M protein (IgG κ, 1,410 mg/dL [14.1 g/L]; case 338), or generalized with a high level of serum M protein (up to 92 g/L) accompanied by a low tumor burden in the bone marrow (case 78) Image 8. Imaging studies found no lytic bone lesions, and the overall clinical features did not support a diagnosis of disseminated myeloma. FISH studies demonstrated no MALT1 or IGH translocation.

Do these neoplasms represent nodal/extramedullary plasmacytoma or MZL/MALT lymphoma with marked plasmacytic differentiation? Lack of demonstrable monoclonal B cells argued against a diagnosis of MZL, and, yet, it was unclear whether such cases represented the extreme end of the morphologic spectrum of MZL. Reports of MALT lymphoma recurring as extramedullary plasmacytoma and vice versa suggest that the 2 may be part of a spectrum.59,60 Extramedullary plasmacytomas usually lack t(11;14) and cyclin D1 expression, and CD56 expression is rarely observed, suggesting that they are biologically distinct from myeloma.59,60 Further studies are needed to elucidate the relationship of extramedullary plasmacytoma involving lymph nodes and MZL/MALT lymphoma.

In case 62, a patient with an isolated pulmonary amyloidoma had a small serum IgG λ shown by immunofixation electrophoresis. A lung biopsy revealed small clusters of plasma cells that were monotypic for IgA λ and rare small lymphoid aggregates of polytypic B cells by immunostains in the area of the amyloidoma. Liquid chromatography–tandem mass spectrometry, which has been shown to better characterize amyloid fibers than immunohistochemical analysis,61 was performed on peptides extracted from Congo red+ microdissected areas of the tissue. This approach showed that the amyloid was composed mainly of IgA and λ light chains with minor components of other types of immunoglobulin. The amyloidoma was apparently produced locally by the monoclonal plasma cells without evidence of B-cell lymphoma. This case also raises the possibility that localized plasmacytic proliferation associated with isolated amyloidoma may be an entity distinct from MALT lymphoma and a systemic plasma cell neoplasm.

Summary

LPDs associated with plasmacytic differentiation include CLL, FL, MCL, MZL, large cell lymphoma, and others. Most cases are readily classifiable based on overall morphologic features, immunophenotype, and molecular genetics. However, borderline cases are challenging.

When a B-cell neoplasm develops a significant plasma cell component, the possibility of a coexisting plasma cell neoplasm also needs to be ruled out. Demonstration of the IGH/CCND1 or IGH/BCL2 fusion gene in the lymphoid and the plasma cells provides direct evidence for a common origin of B cells and plasma cells in MCL and FL, respectively. Comparing IGH rearrangement patterns or immunoglobulin light chain expression patterns between the plasma cells and B cells is useful clinically in cases that lack a genetic hallmark, although finding identical light chains is not definitive evidence supporting clonal identity. Monoclonal plasmacytic proliferations involving lymph node and extranodal tissue without demonstrable monoclonal B cells may represent plasmacytomas, although distinction from MZL with marked plasmacytic differentiation is sometimes difficult and, in some cases, becomes purely definitional. Localized monoclonal plasma cell proliferations associated with an amyloidoma may represent an entity distinct from LPL or systemic plasma cell neoplasm, and liquid chromatography–tandem mass spectrometry may help determine the nature of amyloid fibers. Some cases could not be definitively classified using our current classification, and a descriptive diagnosis was necessary.

Image 8

(Case 78) The patient had a history of Sjögren syndrome with parotid swelling. The serum IgG level was 7,500 mg/dL (75 g/L), and no lytic bone lesions were identified. The bone marrow demonstrated 20% to 30% monoclonal plasma cells. A, Biopsy of the cervical lymph node showed a diffuse infiltrate of neoplastic plasma cells (inset, some neoplastic cells show prominent nucleoli, and mitotic figures are present) (H&E). B, A residual germinal center adjacent to monocytoid B cells (H&E). C, The plasma cells were CD138+/CD56–/CD20–/cytoplasmic κ+, and the B cells were polytypic. The case highlighted the difficulty of making the distinction between marginal zone lymphoma with marked plasmacytic differentiation and plasmacytoma involving lymph node. APC, allophycocyanin; Cy, cyanine; Cyto, cytoplasmic; FITC, fluorescein isothiocyanate; PE, phycoerythrin; PerCP, peridinin chlorophyll protein; SSC, side scatter.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
View Abstract