3. Pathology of Multiple Myeloma

The standard of care for PC quantification is still morphologic assessment of the bone marrow aspirate and biopsy (Fig. 3.1). Flow cytometry immunophenotyping (FCIP) is not a reliable method for PC quantification as studies have shown that the FCIP tends to underestimate the percentage of PCs. This is due to a number of factors such as exclusion of lipid phase-associated disease component and ex vivo loss of antigens used for PC identification [2, 3]. In addition, the Ficoll separation process used in some laboratories for mononuclear cell enrichment makes FCIP quantification of plasma cells even more problematic. Although FCIP does not supplant morphologic marrow assessment, multiparametric PC analysis by this method is an important part of the diagnostic evaluation, enabling separation of neoplastic (monoclonal) from background (polyclonal) PC population, which is not possible by morphologic assessment [4]. This feature is utilized in characterization of plasma cells for clonality, calculation of proliferation fraction, and minimal residual disease (MRD) analysis (see below). In addition to Wright-Giemsa stain of the bone marrow aspirate, evaluation of the bone marrow core biopsy by morphology (Hematoxylin-Eosin stain) or by immunohistochemistry (IHC: CD138, MUM-1/IRF-4, immunoglobulin light chains) is necessary to exclude sampling error; it is not uncommon to observe aspirates with very few plasma cells and encounter sheets of PCs associated with fibrosis in the biopsy specimen. Normal bone marrow aspirate contains approximately 1–2 % PCs, and the defined threshold for the diagnosis of myeloma is 10 %, as defined by WHO guidelines [1]. Reactive marrow PCs may be increased above this threshold in a number of conditions, however, therefore establishing PC clonality is essential.

Clonality of PCs is inferred by showing of monotypic immunoglobulin light chain expression (kappa or lambda) and/or abnormal patterns of antigen expression. Rarely, the clonal PCs lack detectable immunoglobulin expression (Ig-negative). FCIP, IHC, and in situ hybridization (ISH) are commonly used methods for establishing PC clonality. FCIP has an advantage of multiparametric analysis of plasma cells, up to 8 or 10 antigens in clinical laboratories, enabling a more precise separation of neoplastic PCs (CD19 and CD45-negative, CD56-positive) from normal PCs (CD19 and CD45-positive, CD56-negative) (Fig. 3.2). FCIP collection of large number of events (500,000 per specimen) enables a high sensitivity evaluation (0.01 %) for the presence of clonal PCs in the bone marrow aspirate. This is especially important for the detection of MRD after treatment. If aspirate is of poor quality due to technical difficulties or marrow fibrosis, IHC or ISH can be performed on the bone marrow biopsy specimen. These methodologies may also be helpful in older specimens, as PCs become more difficult to detect by FCIP in BM aspirates after 72 h.

A number of B-cell neoplasms may exhibit plasmacytic differentiation, the quintessential entity being LPL. In LPL the neoplastic cells exhibit a cytologic spectrum of small lymphocytes, plasmacytoid lymphocytes, and plasma cells. LPL commonly involves bone marrow or lymph nodes, and sometimes spleen and other tissues. It is usually associated with secretion of IgM class of immunoglobulin in the blood. Due to high molecular weight of IgM pentamer molecule, blood viscosity can be increased, leading to syndrome of Waldenström macroglobulinemia (WM). Bone marrow examination helps differentiating LPL from MM, which is important for both therapeutic and prognostic purposes. In MM, a monomorphic plasma cell population is usually pure without associated lymphoid component, whereas in LPL typically small lymphocytes and plasmacytoid lymphocytes predominate. LPL may infiltrate the marrow in a nodular or interstitial pattern and the plasma cells and lymphocytes may be intimately admixed, such as with lymphoid nodules rimmed by plasma cells, or physically separate (Fig. 3.3). FCIP in LPL typically reveals monotypic B-cells, which may be CD5 positive. The plasma cells in LPL are variably well detected by FCIP, express the same light chain as the B-cell component, and typically retain expression of CD19 and CD45 (unlike in MM); they are never positive for CD56 and cyclin D1 [5, 6]. Clinical features are also helpful, including the presence of lymphadenopathy, the absence of bone lytic lesions, and the presence of IgM paraprotein. It is important to emphasize that IgM paraprotein can be associated with other B-cell neoplasms, including marginal zone lymphoma (MZL) and chronic lymphocytic leukemia (CLL). MZL can be particularly difficult to distinguish from LPL on bone marrow biopsy, as both entities can have plasmacytic differentiation. Very rarely, MM can also secrete IgM; IgM myeloma may have lymphoplasmacytoid cytology and be CD19 positive; in such cases detection of cyclin D1 overexpression in IgM myeloma and the presence of bony disease is critical in distinguishing it from LPL [7].

In addition to distinguishing MM from LPL or other lymphomas, bone marrow biopsy can help in the diagnosis of POEMS (polyneuropathy, organomegaly, endocrinopathy, M-protein, skin changes) syndrome. POEMS syndrome is usually represented in the bone marrow by a relatively small proportion of monoclonal lambda plasma cells (associated with increased polyclonal plasma cells), reactive lymphoid aggregates surrounded by plasma cells, megakaryocytic hyperplasia, and varying levels of bone sclerosis [8].

The most important laboratory prognostic factors are proliferation rate of neoplastic plasma cells and cytogenetic findings [9, 10]. Additional prognostic factors, including gene expression profiling, have also been described [11]. Older methods for determining PC proliferation rate included BrdU DNA pulse labeling and fluorescent staining of the aspirate with anti-BrdU antibodies [12]. Clinical correlation studies have shown that BrdU incorporation in >3 % of cells is associated with poor prognosis. However, this method is labor-intensive and is difficult to perform. It has been supplanted by recently developed FCIP methods for measuring S-phase of neoplastic PCs by detection of DAPI nuclear staining (Fig. 3.4). This method has several advantages: (1) neoplastic and non-neoplastic PCs can be accurately discriminated and their relative proportions calculated; (2) it enables measuring proliferation rate of neoplastic PCs only (separate from polytypic background); (3) it is highly sensitive and shows great precision in calculating S-phase of PCs; (4) it can detect aneuploid and polyploid populations adding to prognostic factors; and (5) it enables detection of small clones based on their DNA content. Clinical studies validating the S-phase cut-off value for this method are still in progress, but are likely to be between 1.5 and 3 %.

As mentioned in earlier chapters, several cytogenetic findings have been shown to be associated with poor prognosis, including t(4:14), t(14;16), del(13), and del(17p) by fluorescence in situ hybridization (FISH) and hypodiploidy by karyotype analysis [9, 10].

Amyloid is insoluble and enzyme-resistant form of a misfolded protein. Its accumulation in extracellular space leads to multiple organ dysfunction, including heart, peripheral nerves, esophagus, spleen, and kidney. Many proteins can form amyloid, but the most common one is immunoglobulin light chain (primary or AL amyloid). The misfolding of amyloidogenic protein results in antiparallel beta pleated-sheets that give amyloid its chemical and physical features, including resistance to enzymatic digestion and light transmission properties. The latter is used in amyloid detection in the tissue biopsy, including bone marrow and subcutaneous fat aspirate: Congo Red stain of amyloid deposits shows characteristic birefringence (red-apple green) under polarized light (Fig. 3.5). It is important to emphasize that the presence of clonal PCs in the bone marrow, with associated amyloid deposits, does not automatically imply that the amyloid is of AL type, as monoclonal gammopathies are rather prevalent in older patient population. For that reason, after amyloid is detected, it needs to be subtyped to identify its forming protein. The classical methods of IHC staining of amyloid deposits lack sensitivity and specificity and have been replaced by recently developed mass spectrometry proteomic methods [13]. This method shows a remarkable ability to precisely identify protein forming amyloid; more than a hundred different amyloidogenic proteins have been identified so far using proteomic tools.

There is a wide range of pathologic processes that can accompany PCPDs. The most common ones are Large Granular Lymphocyte (LGL) proliferations and therapy-related myeloid neoplasms.

LGL proliferations (LGL leukemias) are monoclonal or oligoclonal lymphoproliferative disorders of cytotoxic lymphocytes (T or NK-cells). These expansions may be associated with cytopenias (anemia, neutropenia, and/or thrombocytopenia) [14]. It can be challenging to establish the diagnosis of an LGL proliferation in the presence of PCPD, as LGL proliferations can be a part of a normal immune response to emerging PC clone. In addition, cytopenias are often a feature of PCPD itself. Therefore, LGL proliferations are usually diagnosed in cases of disproportionate lymphoid infiltrates and cytopenias that are not explained by the extent of PC involvement of the bone marrow or M-protein concentration. However, criteria for establishing the diagnosis of LGL proliferation in the presence of PCPD are not well-defined.

Therapy for MM includes cytotoxic drugs such as melphalan. The well-known side effect of these drugs is DNA damage in normal hematopoietic cells. The accumulation of DNA damage can lead to secondary, therapy-related, myeloid neoplasms such as myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) [15] (Fig. 3.6). Careful examination of bone marrow specimen for early signs of therapy-related changes is necessary in any MM patient on therapy.