A Rare Case of Non-IgM Lymphoplasmacytic Lymphoma with Unusual Lack of Immunoglobulin Light Chain Production

Background

Lymphoplasmacytic lymphoma (LPL) is one type of low-grade small B-cell lymphoproliferative disorder, often with a spectrum of small lymphocytes, plasmacytoid lymphocytes, and plasma cells. Most LPL cases are associated with an immunoglobulin M (IgM) paraprotein, and involving bone marrow, clinically present as Waldenström macroglobulinemia (WM). LPL with IgG or IgA paraprotein (non-IgM LPL) is rare, accounting for less than 5% of LPL. Due to its rarity, non-IgM LPL is poorly characterized. Several recent studies investigating the clinical course, histopathological features, and treatment demonstrated many overlaps between WM and non-IgM LPL [1–4]. However, non-IgM LPL cases appeared to have higher odds of extramedullary disease, such as lymphadenopathy and organomegaly, than did WM controls. The rates of neuropathy and symptomatic hyperviscosity in IgM LPL were higher, which was more likely related to the physicochemical properties of IgM. The morphological features and immunophenotypic profiles for monotypic B cells and monotypic plasma cells in IgM versus non-IgM LPL cases were similar. The time to treatment initiation from diagnosis was significantly shorter for non-IgM patients than WM patients. However, the overall survival rates in these 2 groups did not appear different.

Case Report

A 67-year-old woman with a history of gastro-esophageal re-flux disease, osteoarthritis, hyperlipidemia, and anxiety and without history of hepatitis C and hepatitis B viral infection was referred for anemia and leukopenia evaluation. A review of the patient’s laboratory test results showed a low hemoglobin level, in a range of 8.2 to 8.8 g/dL, bordering on macrocytosis for at least 4 months and borderline leukopenia (absolute neutrophil count: within the reference range at 2200/µL). Serum protein electrophoresis failed to detect a paraprotein, but a faint IgG and kappa against a dense polyclonal background was revealed by immunofixation. Her serum kappa and lambda free light chain, IgA, and IgG levels were all within the reference range, and IgM levels were mildly decreased. She also had early satiety, constipation, and weight loss (5.4 kg over the last 6 months), for which she underwent esophagogastroduodenoscopy and colonoscopy, with unremarkable findings. The patient denied any bleeding event, fever, chills, nights sweats, or fatigue. B12, folate, copper, methylmalonic acid, homocysteine, haptoglobin, and total bilirubin levels were within the reference range and serum iron was borderline low.

She underwent a bone marrow aspirate and biopsy. Flow cyto-metric analysis revealed that most B cells (19% of the CD45+ events) were positive for CD19, CD20, CD22, CD23 (subset), CD25, and CD200 and negative for CD5, CD10, CD34, CD38, CD103, and CD11c (Figure 1). Interestingly, these B cells showed expression of neither surface nor cytoplasmic kappa and lambda immunoglobulin light chains (Figure 1). Additionally, there was an aberrant population of plasma cells positive for CD45, CD19, CD38, and CD138 and negative for CD56 and CD117, without definitive light chain expression (Figure 2). Bone marrow biopsy and aspirate showed the marrow was hypercellular (90% cellularity) with increased small lymphocytes and plasma cells (Figure 3A, 3B). B lymphocytes highlighted by immunohistochemical staining accounted for approximately 10% to 20% of the marrow cellularity that were positive for CD20 (Figure 3C), CD79a, CD22, PAX5, and BCL2 and negative for cyclin D1, BCL6, CD10 (data not shown), and plasma cell markers. The plasma/plasmacytoid cells were 20% to 30% and positive for CD138 (Figure 3D), CD79a, CD38, MUM1, BCL2 (data not shown), and negative for other B cell markers. Staining for kappa and lambda by different modalities, including immunohistochemical stains (Figure 3E, 3F) and in situ hybridization studies (Figure 4A, 4B) confirmed lack of light chain expression among both the B and plasma cells. Interestingly, immunohistochemical staining showed the plasmacytoid cells produced IgA (Figure 4C) but not IgM (Figure 4D) or IgG (data not shown), indicating that IgA heavy chain was produced but not secreted. The specimen was positive for clonal gene rearrangement of immunoglobulin heavy chain (IgH) and light chain (IgK). Molecular studies showed positive MYD88 L265P mutation (c.794T>C) and CXCR4 mutation (c.1013C>G (p.S338)). No other mutations, translocations (including, CCND1 and IgH), and chromosomal abnormalities were identified by next-generation sequencing, myeloma panel, myelodysplastic syndrome fluorescence in situ hybridization panel, and cytogenetic analysis. The findings were consistent with marrow involvement by non-IgM LPL.

Positron emission tomography/computed tomography displayed diffusely increased activity throughout the skeleton in the marrow space, likely due to patient’s known B-cell lymphoproliferative disorder and anemia. No intense abnormal focal lesion was noted above the background. There were small nodes in the head and neck, with mild activity above the background; however, no bulky disease was found. The patient started treatment with rituximab and bendamustine and reported significant fatigue and a low-grade fever after her first cycle of chemotherapy was completed. The bendamustine dose was reduced owing to her symptoms and borderline-low absolute neutrophil count. She completed 6 cycles of treatment eventually, and the repeated bone marrow biopsy 7 months later showed normocellular marrow, with no histological, immunophenotypic, and molecular evidence of residual LPL.

Discussion

Here we present an interesting case of non-IgM LPL, and the clinical presentation of the patient included anemia, mild lymphadenopathy, and nearly no paraprotein secretion. The bone marrow was hypercellular, with moderate involvement by the lymphoma cells (30–50% of the marrow). The immunophenotype of the B lymphocytes and plasma cells were consistent with those reported in the literature, with the exception that there was lack of surface and cytoplasmic immunoglobulin light chain expression among the aberrant B cells. The lymphoma cells were positive for MYD88 L265P and CXCR4 (p.S338) mutations, which are classical for WM and non-IgG LPL. The patient responded well to rituximab and bendamustine treatment, and no marrow disease was detected after 6 cycles of treatment.

Detection of IgM monoclonal protein is instrumental in the differential diagnosis of an underlying lymphoproliferative disorder. Its presence is mostly caused by LPL and rarely by plasma cell neoplasm [5]. While lack of IgM points to a list of potential differential diagnoses, including plasma cell myeloma/monoclonal gammopathy of undetermined significance (MGUS), light chain MGUS/myeloma, monoclonal gammopathy of renal significance, and non-IgM LPL, the patient’s normal renal function and absence of paraproteins, including serum IgM and monoclonal light chains in serum and urine, favored the diagnosis of plasma cell neoplasm or B-cell lymphomas.

Non-IgM LPL is a rare disorder and its diagnosis is challenging. Patients usually have non-IgM paraproteins, and the neo-plastic cells display prominent plasmacytic differentiation, therefore mimicking plasma cell neoplasm. The accurate diagnosis is crucial because of the different treatment management for these 2 diseases. Clinically, patients with non-IgM LPL usually do not show lytic bone lesion but sometimes can have lymphadenopathy and/or splenomegaly [1,3]. Flow cytometry analysis of the B lymphocytes and plasma cells from non-IgM LPL consistently show monotypic concordant immunoglobulin light chain expression between these 2 components. In one series study [6], the neoplastic plasma cells defined by bright CD38 and CD138 were almost always positive for CD45 (100%, 26/26) and CD19 (96%, 25/26), while only partial/weak positive for CD56 (16%, 4/25), and negative for CD117 (100%, 24/24). In contrast, plasma cell neoplasm cells commonly show loss of CD19 and CD45 and gain of aberrant expression of CD56 and CD117 in about 70% and 30% cases, respectively. In our case, the neoplastic plasma cells were positive for CD45 and CD19 and negative for CD56 and CD117, indicating they were part of the LPL.

A number of recurrent somatic mutations have been disclosed by next-generation sequencing in WM, including MYD88, CXCR4, ARID1A, and CD79B [7–9]. MYD88 is the most common mutation (93–97%), resulting in gain of function of MYD88 and constitutive activation of the NF-κB pathway [10]. MYD88 L265P is the predominant recurrent MYD88 mutation in WM, while non-L265P MYD88 mutations are estimated to be only 1% to 2%. The second most common mutation is CXCR4 (30–40%) which almost always occurs in those with MYD88 mutations and includes more than 40 different types. The most frequent CXCR4 mutated region (50%) is the amino acid S338X at position 1013, with nucleotide changes C > G in 54% and C > A in 25% of the cases, both resulting in a stop codon, followed by S338 frameshift mutation in 21% of cases [11]. MYD88 and CXCR4 mutation status can be associated with clinical presentation, response to Bruton’s kinase inhibitor ibrutinib, and overall survival in WM [12,13]. WM patients with wild-type MYD88 have a greater risk of disease transformation, poor response to ibrutinib, and decreased overall survival. Patients with nonsense CXCR4 mutations have increased bone marrow disease, higher serum IgM levels, and/or symptomatic hyperviscosity at presentation. The gene mutation data in non-IgM LPL is relatively scarce, and recent studies suggest MYD88 and CXCR4 mutations are possibly the 2 most recurrent gene mutations, similar to WM. Limited data indicate the prevalence of a MYD88 mutation in non-IgM LPL is about 43% to 80%, appearing to be lower than in WM [1–4,6]. Approximately 25% to 33% of the non-IgM LPL cases harbor a CXCR4 mutation, which frequently co-occurs with MYD88 mutation (67–75% concurrence rate) [3,6]. Both MYD88 and CXCR4 mutations are not entirely specific for WM and non-IgM LPL and can be also detected in many other B-cell lymphomas [14]. MYD88 mutation, nevertheless, is extremely rare in plasma cell neoplasms, including non-IgM monoclonal gammopathy of undetermined significance and plasma cell myeloma (1.2%, 3/246 cases). Translocation (11;14)/CCND1/IGH can often be detected in plasma cell neoplasms, and presence of this translocation in the correct clinicopathologic setting can help diagnose plasma cell neoplasm, while its absence does not exclude that diagnosis. In our case, positive MYD88 and CXCR4 mutations with the concurrently detected abnormal B-cell population, and lack of myeloma-type genetic changes were further against a diagnosis of plasma cell neoplasm.

The other challenge in the diagnosis of non-IgM LPL is differentiating from other low-grade small B-cell lymphomas. Sometimes it is impossible at the initial diagnosis of disease to separate non-IgM LPL, particularly from marginal zone lymphoma (MZL) with plasmacytic differentiation. Careful clinicopathologic correlation aided by ancillary tests, especially the molecular mutational landscape, can be instrumental in this differential diagnosis. Literature shows CXCR4 mutation occurs in approximately 25% to 33% of non-IgM LPL cases [3,6]; its mutation status in the other small B-cell lymphomas, however, has been rarely investigated. Searching through the literature reveals available data only in 3 recent studies for MZL [15–17]. Both MYD88 mutation and CXCR4 mutation were found to be present in the MZLs, which include cases of nodal, extranodal, and splenic subtypes, and their mutation prevalence was 11.6% (14/121) and 5.8% (7/121), respectively. Nevertheless, it appears that only 1 nodal MZL case had a concurrent MYD88 and CXCR4 mutations (0.8%, 1/121). Therefore, for cases like the current one with both gene mutations in the absence of overt lymphadenopathy or clinical features of other MZL, a diagnosis of LPL is strongly supported. However, the differential diagnosis for cases lacking the expected clinical, pathologic, and molecular findings still remains substantially challenging.

A paraprotein other than IgM is often detected in non-IgM LPL, and the presence of IgG paraprotein is the most likely type, followed by IgA. Non-secretory non-IgM LPL is not common, accounting for about 0% to 13% of the studied cases [1–4,6]. Flow cytometry analysis of the non-IgM LPL cases in the literature indicates the B cells mostly demonstrate monotypic surface light chain, with kappa being the more frequently expressed light chain. To the best of our knowledge, this case is the first non-IgM LPL in which B lymphocytes lost both surface and cytoplasmic Ig light chain production and/or expression and thus lost a detectable light chain restriction (LCR), as well as produced IgA but failed to secrete it. Flow cytometric analysis with the finding of B-cell populations showing surface LCR (so-called monotypic) is a quick and efficient way of diagnosing B-cell lymphomas. However, lack of LCR can occur in 7% to 19% of the B-cell non-Hodgkin lymphomas [18–20], among which diffuse large B-cell lymphoma (DLBCL) shows the highest rate (22–30%). Interestingly, clinicopathological analysis shows DLBCL lacking surface LCR demonstrates more histological necrosis and less BCL6 and CD10 expression, as compared with DLBCL with LCR. However, the survival outcomes of patients are comparable [19]. The LCR status of LPL is unknown, and it would be interesting to know how it can affect the biological behavior of non-IgM LPL. It remains largely unclear why lymphoma cells lose surface immunoglobulin expression, and it has been suggested that the defect can occur at all levels of immunoglobulin production, including transcription, translation, posttranslational modification, intracellular transportation, and secretion. In the present case, the lymphoma cells showed clonal rearrangement of immunoglobulin heavy and light chains; however, light chain protein failed to present on the cell membrane and cytoplasm. No light chain RNA was detected by kappa/lambda in situ hybridization, suggesting the lymphoma cells perhaps had defect in transcription/post-transcription processes, rendering them to be a non-producer of light chains. IgA was detected intracytoplasmically; nonetheless, IgA paraprotein was not measurable in the serum protein electrophoresis and immunofixation, suggesting the lymphoma cells were IgH chain producers but non-secretors.

Conclusions

Non-IgM LPL is a rare subtype of LPL that usually expresses IgG or IgA with a light chain restriction; non-IgM LPL which is nonsecretory is even rarer. We report a case of an LPL with unique feature which is a non-producer of light chains and producer of IgA heavy chain but a non-secretor. The correct diagnosis is critical for clinical management, and the marrow findings with analogous MYD88 and CXCR4 mutations confirmed the diagnosis of non-IgM LPL for this patient, who had a good response to the combined treatment. The contribution of this unique feature to the biological behavior of non-IgM LPL as well as the mechanism of loss of light chain production requires further investigation.