Plasma Cell Neoplasms

Definition

The plasma cell neoplasms and related disorders are clonal proliferations of immunoglobulin (Ig)-producing plasma cells or lymphocytes that make and secrete a single class of Ig or a polypeptide subunit of a single Ig that is usually detectable as a monoclonal protein (M-protein) on serum or urine protein electrophoresis. These immunosecretory disorders may consist exclusively of plasma cells (plasma cell neoplasm) or a mixture of plasma cells and lymphocytes. Those with a mixture of plasma cells and lymphocytes are generally categorized as lymphomas and are discussed elsewhere in this book. The plasma cell neoplasms are the subjects of this chapter. Most of these have their origin as bone marrow tumors but occasionally present in extramedullary sites.

Classification

Box 26-1 lists the categories of plasma cell neoplasms included in the WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues .

Definition

Plasma cell myeloma (PCM) is a bone marrow–based, multifocal plasma cell neoplasm usually associated with an M-protein in serum or urine. The bone marrow is the site of origin of nearly all myelomas and in most cases there is disseminated marrow involvement. Other organs may be secondarily involved. The diagnosis of myeloma is made by a combination of clinical, morphologic, immunologic and radiographic information. The disease spans a clinical spectrum from asymptomatic to highly aggressive. In a minority of myelomas, pathologic manifestations of deposition of abnormal Ig chains in tissues are the major clinical findings.

Diagnostic Criteria

The usual findings in PCM are increased and abnormal bone marrow plasma cells or a plasmacytoma together with an M-protein in serum or urine. Frequently bone lesions are present. The diagnostic criteria for PCM put forth by the International Myeloma Working Group are listed in Box 26-2 .

Epidemiology

PCM (and its variants) is the predominant type of malignant immunosecretory disorder. Myeloma accounts for about 1% of malignant tumors and 10% to 15% of hematopoietic neoplasms. Approximately 26,000 new cases of PCM were diagnosed in the United States in 2015, with about 11,000 deaths from myeloma. It is more common in men than women (1.1 to 1) and occurs twice as frequently in African Americans as in whites. The risk for PCM is 3.7-fold higher for individuals with a first-degree relative with the disease. Myeloma is not found in children and is found rarely in adults younger than 35 years; the incidence increases progressively with age thereafter, with approximately 90% of cases occurring in individuals older than 50 years. The median age at diagnosis is about 68 to 70 years.

Etiology and Pathogenesis

Exposure to toxic substances and radiation has been associated with an increased incidence of PCM. Chronic antigenic stimulation from chronic infection or other disease may also be a predisposing factor. Most patients with myeloma, however, have no identifiable exposure history or known chronic antigenic stimulation.

There is evidence that PCM results from a disorder of an early hematopoietic cell that is manifested at a mature stage of B-cell development. Part of the evidence supporting this view is the presence, in nearly all cases of myeloma, of monoclonal blood lymphocytes immunophenotypically and genetically related to the neoplastic bone marrow plasma cells.

Recent information on the molecular genetics of PCM has greatly enhanced the understanding of its pathogenesis (see genetics section later). The bone marrow microenvironment is also important in the pathogenesis and progression of myeloma. Cytokines, growth factors, and the functional consequences of direct interaction of bone marrow stromal cells with neoplastic plasma cells are major constituents that influence the pathophysiology of myeloma. Several pieces of evidence point to the involvement of Interleukin-6 (IL-6) as a factor in the pathogenesis of PCM. IL-6 appears to support the survival and expansion of myeloma cells by stimulating cell division and preventing programmed cell death. IL-6, along with IL-1b and tumor necrosis factor α (TNF-α) and other cytokines, has osteoclastic-activating properties that lead to lytic lesions through a complex mechanism involving the RANKL pathway. These cytokines may also affect hematopoiesis.

Clinical Features

The most frequent symptom at presentation is bone pain in the back or extremities due to lytic lesions or osteoporosis. In advanced cases, vertebral collapse may cause loss of height. Weakness and tiredness, often related to anemia, are common complaints. Some patients are seen with infections, bleeding, or symptoms related to renal failure or hypercalcemia. Rarely, neurologic manifestations due to spinal cord compression or peripheral neuropathy are the reason for seeking medical attention. Occasionally in asymptomatic individuals, the diagnosis of PCM follows discovery of a serum M-protein on protein electrophoresis. Physical findings are often non-specific or lacking. Pallor is most common, followed by organomegaly. Palpable plasmacytomas are rare, but tenderness and swelling over the site of a pathologic fracture or plasmacytoma may be encountered. Tissue masses and organomegaly due to plasma cell infiltration or amyloidosis are found in a few patients. Skin lesions due to plasma cell infiltrates or purpura are observed in rare cases.

Laboratory Findings

Box 26-3 lists the diagnostic studies recommended by the International Myeloma Working Group for the assessment of patients suspected of having PCM. The data obtained from these studies form the basis for clinical-pathologic criteria for diagnosis of PCM and provide important prognostic information.

Assessment of serum and urine for M-protein is an essential component of the evaluation of patients suspected to have a PCM. Agarose gel electrophoresis is the preferred method to screen for M-proteins. A M-protein is found on serum protein electrophoresis (SPE) in most patients with myeloma ( Fig. 26-1 ). The total immunoglobulin is usually increased due to the M-protein, but normal polyclonal immunoglobulins are commonly decreased. A SPE M-protein may be undetectable in cases with low levels of monoclonal Ig, as commonly seen in IgD, IgE, and light chain myeloma; hypogammaglobulinemia due to decreased normal polyclonal immunoglobulins may be the only abnormal SPE finding. Urine protein electrophoresis (UPE) on a concentrated urine specimen, and Ig quantification on a 24-hour urine collection should be performed in all cases of suspected myeloma ( Fig. 26-2 ). Monoclonal light chains (Bence-Jones protein) are found in some patients without a serum M-protein. Serum and urine immunofixation electrophoresis is the gold standard for characterizing heavy chains and light chains and for detecting small quantity M-protein, as may be seen in patients with light chain amyloidosis, plasmacytoma, heavy chain disease, and light chain deposition disease and following treatment for myeloma (see Figs. 26-1 and 26-2 ). Immunofixation is capable of detecting an M-protein of 0.02 g/dL in serum and 0.004 g/dL in urine. With immunofixation electrophoresis, an M-protein is identified in the serum or urine in about 97% of myeloma cases. Monoclonal light chains are found in the urine in approximately 75% of cases; in nearly two thirds of them, the light chains are kappa. A patient may have a negative urine electrophoresis when immunofixation of a concentrated urine specimen identifies a monoclonal light chain. Light chains are reabsorbed by proximal renal tubules. Therefore, renal function plays a role in determining whether light chain is detectable in urine.

The serum free light chain immunoassay provides a highly sensitive method for detecting very small quantities of monoclonal light chains; it is even more sensitive than IFE. The quantity and the serum free light chain kappa/lambda ratio are powerful determinants of disease activity. Serum free light chain analysis is important in screening and monitoring patients with plasma cell neoplasms, especially oligosecretory ones such as some light-chain-only myelomas, amyloidosis, solitary plasmacytoma, and a majority of those previously considered non-secretory myeloma. The baseline serum free light chain quantity and ratio are important indicators of prognosis for every category of plasma cell neoplasm including MGUS. A normal serum free light chain ratio is a criteria of stringent complete response for treated plasma cell neoplasms together with absence of an M-protein by IFE and absence of clonal plasma cells in the bone marrow.

An IgG M-protein is found in slightly more than half of patients with myeloma, and IgA and monoclonal light chains only are found in approximately 20%. IgD, IgE, IgM, and biclonal myeloma, all of which are rare, compose the remainder of M-proteins. Less than 3% of patients have a non-secretory myeloma by IFE, but low quantities of monoclonal light chain are detectable in a majority of these by serum free light chain analysis. An average of the frequencies of the various M-proteins in several large series of myeloma patients is shown in Table 26-1 . Kappa light chain is more common than lambda light chain for all immunoglobulin types of myeloma except IgD. The quantity of serum M-protein varies from undetectable to more than 10 g/dL. The median is approximately 5 g/dL for IgG myeloma and 3.5 g/dL for IgA. Approximately 40% of patients with symptomatic myeloma have an M-protein less than 3 g/dL. In cases of light-chain-only myeloma, the serum M-protein may be very low or undetectable; the 24-hour urine protein is usually mildly to markedly elevated.

Anemia is present in about two thirds of patients at diagnosis. Red blood cell indices are usually normocytic and normochromic. Leukopenia and thrombocytopenia are found in less than 20% of patients but frequently evolve as the disease progresses. Patients occasionally have leukocytosis or thrombocytosis. The erythrocyte sedimentation rate is variably increased and roughly related to the level of the M-protein.

Hypercalcemia is present in nearly one fifth of patients, and creatinine is elevated in one fifth to one third. Hyperuricemia is found in more than half of patients. Hypoalbuminemia is observed in patients with advanced disease.

Radiographic Studies

Radiographic skeletal surveys reveal lytic lesions, osteoporosis, or fractures in 70% to 85% of cases of myeloma at diagnosis. In some cases, all of the changes are observed. The vertebrae, pelvis, skull, ribs, femur, and proximal humerus are most often affected.

Computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography–computed tomography (PET-CT) have evolved to play an important role in the diagnosis and management of patients with a plasma cell neoplasm. CT and MRI are more sensitive than the conventional skeletal survey and are capable of detecting small osteolytic lesions in areas not well visualized by conventional techniques. MRI imaging has prognostic significance in patients with symptomatic myeloma. Patients with normal or only focal abnormalities on thoracolumbar MRI have a better treatment response and survival. MRI findings also have significance in diagnosis of solitary plasmacytoma and smoldering myeloma (see the sections on smoldering plasma cell myeloma [asymptomatic myeloma] and solitary plasmacytoma). PET-CT is superior in detection of extent of disease, including soft tissue disease, and may be the best technique for assessment of active or inactive disease following therapy.

Blood Smear and Bone Marrow Findings

Rouleaux formation is usually the most striking feature on blood smears and is related to the quantity and type of M-protein ( Fig. 26-3 ). The blood smear may show a faint purple background when the level of M-protein is markedly elevated. Circulating nucleated red cells or a leukoerythroblastic reaction may be observed in some cases. Plasma cells are found on blood smears in approximately 15% of cases, usually in small numbers. They are more commonly observed in the advanced stages of disease. Marked plasmacytosis is present in plasma cell leukemia, which is discussed in the section on plasma cell leukemia.

The bone marrow examination is the most important element of the diagnosis of PCM. A bone marrow examination is nearly always required to confirm the diagnosis, even when there is substantial clinical, laboratory, and radiographic evidence. The bone marrow study also provides prognostic information and is useful in following patient response to therapy and identifying recurrent disease. The bone marrow is the major source of tissue for immunophenotyping, cytogenetics, and molecular studies. In many cases, the diagnosis can be made from the bone marrow examination alone Criteria for morphologic diagnosis of myeloma are shown in Box 26-4 .

Aspirate smears and trephine biopsy sections are both required for optimal evaluation. They are independently diagnostic in many cases, but in some patients it is a combination of findings in the two preparations that leads to the diagnosis. The average number of plasma cells in the aspirate smears is 20% to 36% ( Fig. 26-4 ). In about 5% of cases of symptomatic myeloma, the plasma cells number less than 10%. This may be due to a suboptimal marrow aspirate or because of the frequent focal nature of lesions and uneven distribution of myeloma in the marrow. The neoplastic plasma cells vary from normal appearing with mature features to blastlike cells barely recognizable as plasma cells. The atypical features that characterize many cases of myeloma encompass changes in both the nucleus and cytoplasm. The myeloma cells are often larger than normal plasma cells but may be normal size or small. Moderate to abundant basophilic cytoplasm is usual. An array of cytoplasmic changes is observed. These include fraying and shedding of the cytoplasmic edges, vacuoles, granules, and cytoplasmic inclusions. The nucleus is larger than normal in most cases, and the chromatin less condensed; nucleoli are variably prominent.

Various types of cytoplasmic and nuclear inclusions are observed in myeloma cells and may distort the cytoplasm. Cytoplasmic crystals are found occasionally in myeloma and are a common finding in adult Fanconi's syndrome ( Figs. 26-5 and 26-6 ). Except in adult Fanconi's syndrome, in which the light chain type is invariably kappa, crystals have no obvious relationship to immunologic type of myeloma.

Multiple dark staining cytoplasmic inclusions are observed in rare cases of myeloma ( Fig. 26-7 ). These are often associated with large pleomorphic plasma cells. Multiple small Russell body–type hyaline intracytoplasmic and intranuclear inclusions are relatively common ( Fig. 26-8 ). In contrast to hyaline intranuclear inclusions, Dutcher-type nuclear inclusions are pale staining, single, and generally large ( Fig. 26-9 ). In some cases, cytoplasmic inclusions resemble the Buhot plasma cell structures found in patients with mucopolysaccharidosis. Phagocytic plasma cells are observed in a minority of cases of myeloma; rarely, erythrophagocytosis is striking.

Approximately 2% of myelomas are distinguished by marked nuclear lobation and convolution. In some cases, these cells are mixed with other easily recognizable plasma cells, but in others they comprise a relatively uniform population and may be difficult to recognize as myeloma cells ( Fig. 26-10 ). Small plasma cells predominate in some myelomas, and in approximately 5% of cases the plasma cells have a distinctly lymphoid appearance ( Fig. 26-11 ). In one study, 20% of the cases with lymphoid morphology were IgD myelomas. Lymphoplasmacytic morphology has been associated with CD20 expression on the plasma cells, cyclin D1 positivity, and a t(11;14) chromosomal rearrangement. Overall, attempts to relate morphologic characteristics to monoclonal immunoglobulin type have failed, except for a small number of cases of IgA myeloma with markedly pleomorphic, large multinucleate plasma cells, flaming plasma cells, and cells with pale, frayed, and fragmented cytoplasm ( Fig. 26-12 ). Intranuclear inclusions are found in about 20% of cases of IgA myeloma, much more frequently than for other immunologic types (see Fig. 26-9 ).

On the basis of their cytologic features, myelomas have been classified into mature, intermediate, immature, and plasmablastic cytologic types ( Figs. 26-13 to 26-16 ). Patients with plasmablastic myeloma have a significantly shorter median survival than the other types. There appears to be no significant difference in survival among the other three types. Other classifications include three to six cytologic types.

Histopathology

The diagnostic yield of trephine biopsies is often directly related to the size and number of specimens. Focal lesions may be irregularly distributed and widely spaced. Occasionally only one or two small myeloma lesions are found in a trephine biopsy with no evidence of a plasma cell infiltrate in the remainder of the section or in specimens from the contralateral posterior iliac crest. The pattern of the plasma cell infiltrate may be interstitial, focal, or diffuse ( Figs. 26-17 to 26-19 ). The extent of bone marrow involvement varies from an apparently small increase in plasma cells to complete replacement. The pattern of marrow involvement is directly related to the extent of disease. With interstitial and focal patterns, there is generally considerable marrow sparing and preservation of normal hematopoiesis. With diffuse involvement, expansive areas of the marrow are replaced, and hematopoiesis may be markedly suppressed. There is typically progression from interstitial and focal disease in early myeloma to diffuse involvement in advanced stages of the disease.

A staging system has been proposed based on percentage of marrow space replaced by myeloma in bone marrow trephine biopsies. Less than 20% of the marrow is replaced in stage I, 20 to 50% is replaced in stage II, and more than 50% is replaced in stage III. The extent of involvement in biopsy sections usually reflects the overall tumor burden. There appears to be good correlation between histologic stage, clinical stage, and prognosis.

Myelomas with atypical plasma cell morphology may be difficult to recognize in trephine biopsies ( Figs. 26-20 to 26-22 ). Plasmablastic myeloma, cases with lymphoid-appearing plasma cells, plasma cells with lobulated nuclei, or markedly pleomorphic plasma cells are particularly problematic. Cytologic examination of the cells in aspirate smears is often essential for diagnosis in these cases. Occasionally, cytoplasmic inclusions in the myeloma cells are the most striking feature on the bone marrow section. The inclusions are often found in large plasma cells that are distorted by crystalline or globular material. The globular inclusions may be strongly positive with the PAS stain.

In approximately 10% of myeloma cases, the bone marrow lesions show reticulin or collagen fibrosis. In many of these, the fibrosis is extensive. A disproportionate number of fibrotic myelomas produce monoclonal light chains only. Coarse fibrosis has been correlated with diffuse bone marrow involvement and aggressive disease.

Plasma Cell Myeloma Clinical Variants

The three variants of PCM recognized in the WHO classification have clinical and/or pathobiologic characteristics that differ from typical PCM. These are non-secretory myeloma, smoldering (asymptomatic) myeloma, and plasma cell leukemia.

Non-Secretory Plasma Cell Myeloma

Non-secretory myeloma accounts for about 3% of PCMs. In these rare cases, the neoplastic plasma cells appear to lack the capacity to secrete immunoglobulin, and there is no M-protein in either the serum or urine by immunofixation analysis. In about two thirds of these patients, however, elevated serum free light chains or an abnormal free light chain ratio is detectable. Monoclonal light chains are demonstrated in the cytoplasm of the myeloma cells in about 85% of cases by immunohistochemical staining. In 15% of cases, no staining is detected, suggesting that Ig is not produced (non-producer myeloma). Acquired mutations of the Ig light chain variable genes or alteration of the constant region are involved in the pathogenesis of the non-secretory state. Patients with secretory myeloma at the time of diagnosis may occasionally become non-secretory or oligosecretory at relapse. Non-secretory myeloma must be distinguished from the rare IgD and IgE myelomas that generally have low serum M-protein and may not be routinely screened for by immunofixation. The cytologic and histologic features, immunophenotype, and genetics of non-secretory myeloma appear to be similar to other myelomas.

The clinical features of non-secretory myeloma are also generally similar to other PCMs, except for a lower incidence of renal insufficiency and hypercalcemia and less depression of normal polyclonal Ig. Treatment of non-secretory myeloma is the same as for other PCMs. Prognosis has improved significantly in the past decade similar to other myelomas. In one large study, patients with non-secretory myeloma had a more favorable survival rate than those with secretory myeloma. Survival is better for patients with a normal baseline serum free light chain ratio than for those with an abnormal ratio.

Smoldering Plasma Cell Myeloma (Asymptomatic Myeloma)

About 8% to 14% of patients with PCM are asymptomatic at the time of diagnosis. These patients have 10% or more bone marrow plasma cells and an M-protein at myeloma levels but lack related end-organ impairment ( Box 26-5 ).

The median level of serum M-protein in patients with smoldering myeloma is 30g/L, and a majority has between 10% and 20% bone marrow plasma cells. Approximately 70% of patients have monoclonal light chains in urine, and polyclonal immunoglobulins are decreased in more than 80%. Plasma cells are cytologically atypical in bone marrow aspirate smears, and focal aggregates of plasma cells, interstitial infiltration, or both are found in trephine biopsy sections. The immunophenotype and genetics appear to be similar to other myelomas.

Similar to monoclonal gammopathy of undetermined significance (MGUS), patients with smoldering myeloma may remain clinically stable for a long time, but they are much more likely to eventually progress to symptomatic myeloma. In one report of a large series of patients, the cumulative probability of progression to symptomatic myeloma or amyloidosis was 51% at 5 years, 66% at 10 years, and 73% at 15 years; the median time to progression was 4.8 years.

The serum M-protein level and percentage of plasma cells in the bone marrow have been combined to create a risk-stratification model for smoldering myeloma with three prognostic groups. Patients with both an M-protein of greater than 3g/dL and greater than 10% bone marrow plasma cells form the highest-risk group for progression. Other important risk factors for progression include high percentage of bone marrow plasma cells with an aberrant immunophenotype, detection of bone lesions only by MRI, abnormal serum free light chain ratio, high-risk gene-expression profile (GEP), presence of circulating plasma cells, IgA isotype, high plasma cell proliferative rate, and low polyclonal immunogloblulins.

In the past, patients with smoldering myeloma were rarely treated until they had symptoms related to the myeloma. This approach was based on a lack of clinical evidence that therapy before transformation to symptomatic myeloma was of any benefit. Recently, however, studies have shown that treatment for the highest risk smoldering myeloma patients may delay progression to symptomatic disease and improve overall survival. In one report, the highest-risk patients were defined as those with extreme bone marrow plasmacytosis (>60%), extremely abnormal serum immunoglobulin free light chain ratio (>100), and multiple bone lesions detected only by modern imaging. The therapeutic benefit reported for these highest-risk patients clearly indicates that asymptomatic patients having any of these biomarkers of malignancy should be considered comparable to symptomatic plasma cell myeloma.

Plasma Cell Leukemia

Plasma cell leukemia (PCL) is a myeloma in which the number of neoplastic plasma cells in the blood is greater than 20% of the total leukocytes or the absolute plasma cell count exceeds 2 × 10 ⁹ /L. The neoplastic plasma cells are also commonly found in other extramedullary sites including liver and spleen, body cavity effusions, and spinal fluid. PCL may be primary, present at the time of initial diagnosis, or secondary, evolving during the course of disease in a patient with previously diagnosed myeloma; approximately 60% to 70% of cases are primary. Primary PCL is a distinctive type of PCM with characteristic cytogenetic and molecular findings, and an aggressive clinical course with short remissions and survivals. Primary PCL is found in approximately 2% to 4% of cases of myeloma. Secondary PCL is a leukemic transformation that occurs in approximately 1% of previously diagnosed PCM.

Most of the usual clinical and laboratory abnormalities associated with other myeloma are found in patients with PCL, but there are several features that distinguish it. The median age at diagnosis is younger; lymphadenopathy, organomegaly, and renal failure are significantly more common; and lytic bone lesions and bone pain less common. Anemia is present in 80% of cases of PCL and thrombocytopenia in 50%. Nucleated red cells are frequently observed in blood smears. The total leukocyte count may be in the normal range but is usually elevated and may be as high as 100 × 10 ⁹ /L. All types of M-proteins have been reported in PCL, but a higher proportion of cases of light chain only and IgD myeloma present with PCL than IgG or IgA, and PCL has been reported in approximately 20% of cases of the rare IgE myeloma.

The cytologic characteristics of the leukemic plasma cells span most of the morphologic spectrum found in other myelomas, but large and pleomorphic plasma cells are uncommon. The leukemic cells vary from normal appearing to some that are barely recognizable as plasma cells. Often, many of the plasma cells are smaller than usual with relatively little cytoplasm and may resemble plasmacytoid lymphocytes ( Fig. 26-23 ). Cases with these features may be difficult to distinguish from a lymphoplasmacytic lymphoma on blood smear examination.

The immunophenotype of the neoplastic plasma cells is generally similar to other myelomas, except for more frequent expression of CD20 and less frequent expression of CD56, which is lacking in approximately 80% of cases ( Fig. 26-24 ). CD117 and HLA-DR are also less commonly expressed in PCL (see the section on immunophenotype).

An abnormal karyotype is more frequently found in PCL than in other myelomas, and there is a higher incidence of high-risk genetics in both primary and secondary PCL. These include hypodiploidy, del(13q), del(17p) t(14;16), 1q amplification, and 1p losses. The t(11;14), usually associated with a favorable prognosis in PCM, is also more frequent in primary PCL.

Treatment is similar to that for other advanced myelomas. Patients with PCL have more aggressive disease, poor response to therapy, and a significantly shorter survival than patients with more typical myeloma. Patients with secondary PCL have a shorter survival than those with primary PCL: 1.3 months versus 11.2 months. The high frequency of unfavorable genetic abnormalities only partially explains the poor prognosis of PCL.

Immunophenotype (Flow Cytometry)

Recently a number of investigators have provided data supporting an important role for flow cytometry both at diagnosis and in posttreatment management of plasma cell neoplasms.

Immunophenotypic Features of Normal and Neoplastic Plasma Cells

Normal Plasma Cells

Plasma cells are generally defined immunophenotypically by bright CD38 expression. CD38 expression is not specific for plasma cells, as it is seen at various levels on virtually all other nucleated marrow subsets, but normal plasma cells express higher levels of CD38 than any other normal hematolymphoid cell population ( Fig. 26-25 ). Plasma cells also express CD138, and this antigen is essentially specific for plasma cells among hematolymphoid cells Normal bone marrow plasma cells are considered to be positive for CD19 and CD45, and negative for CD20 and CD56 (see Fig. 26-25 ). However, it is apparent that minority subsets of normal bone marrow plasma cells deviate from each of these prototypic features. Notably, some antigens are modulated on the basis of the maturational stage; as they mature, plasma cells show decreasing intensity of CD45 and CD19 and increasing CD138. A normal CD19-negative, CD56-positive plasma cell population has been detected in bone marrow, and it is postulated to represent a terminally differentiated, long-lived subset. Additional immunophenotypic findings in normal plasma cells include bright expression of CD27 and CD81 and lack of CD28, CD117, and CD200. Normal plasma cell populations express polytypic cytoplasmic immunoglobulin, with kappa : lambda ratios in the range of 1-2 : 1, but occasionally as high as 4 : 1 in reactive plasma cell proliferations.

Neoplastic Plasma Cells

The neoplastic cells in PCM deviate immunophenotypically from normal plasma cells in virtually all cases ( Fig. 26-26 ). Like their normal counterparts, PCM cells express CD38 and CD138, but CD138 expression tends to be brighter and CD38 dimmer ( Fig. 26-27 ) than in normal plasma cells. CD38 intensity in PCM usually exceeds that of other marrow populations, but occasionally there is significant overlap with other cell types ( Fig. 26-28 ). CD19 is absent in about 95% of PCM cases, whereas CD56 is expressed in 60% to 80%. * The reported percentage of PCM cases that express CD45 varies widely, ranging from 18% to 75%. ^† These differences likely result from both technical issues (see later) and biological issues. Regarding the latter, as described earlier, CD45 decreases with maturation of plasma cells, and variability of CD45 expression is a common feature in myeloma; the plasma cells with the brightest CD45 represent the proliferative compartment. Therefore, it would not be surprising to find variation in CD45 expression depending on disease stage or as a consequence of therapy. Reported variability of CD45 in PCM cases during the course of disease supports this notion.

* References .

† References .

CD20 is expressed in approximately 10% to 20% of PCMs, CD117 in 30%, and CD200 in 60% to 75%. * The reported frequency of CD28 expression varies from 16% to 48%. ^† Some of this variability may be explained by the increasing expression of CD28 with more advanced disease. CD81 has been reported to be underexpressed relative to normal plasma cells (dim or negative) in 95% of PCMs. CD27 is negative in up to half of cases; loss of CD27 is more frequent in advanced disease.

* References .

† References .

Technical Issues

Technical Issues Related to Minimal Residual Disease Analysis

Minimal residual disease (MRD) analysis is becoming increasingly important in following patients with PCM (discussed later in this section), and flow cytometry has emerged as the method of choice for MRD detection. Because of the wide variability in CD45 expression and light scatter characteristics in myeloma, gating requires the use of specific antigens. Gating on bright CD38-positive events is the most widely used approach to myeloma, and this generally suffices at diagnosis. However, in the setting of MRD analysis, bright CD38 gating alone is insufficiently sensitive or specific because of the dimmer CD38 expression on PCM and the potential co-occurrence of non–plasma cell events, aggregates, and debris in the bright CD38 region. Consequently, MRD analysis in myeloma requires gating on more than one marker. Because of its 100% specificity and sensitivity for plasma cells, CD138 has emerged as a favored marker for MRD gating. However, optimization of CD138 assessment may be hampered by technical issues, including clone choice, lyse reagent, and refrigeration. CD38 and CD138 in tandem appear to be an effective gating strategy, capturing the vast majority of cases. If feasible, a three-parameter gate with CD38, CD138, and CD45 appears to be maximally sensitive.

The detection of MRD in PCM depends on the identification of immunophenotypic aberrancy on the plasma cells. Simply incorporating CD19 and CD56 has been suggested to capture more than 90% of PCM MRD. However, the detection of normal CD19-negative, CD56-positive plasma cell populations raises some concerns about the use of these as the sole criteria for MRD. It appears that assessment of combinations of multiple antigens with aberrant expression patterns (discussed earlier) is required for optimal MRD assessment ( Fig. 26-29 ). The Euroflow group in 2012 recommended CD19 and CD56 as first-tier makers, followed by assessment of CD27, CD28, CD81, and CD117 as follow-up markers if necessary. Assessment of CD200 also appears very promising, but bears further investigation. Because of the need for multiparameter gating and the need to assess multiple aberrantly expressed antigens, high-color flow cytometry (≥6 colors) seems to be optimal for MRD analysis of PCM. It is worth noting that minor immunophenotypic modulations can occur over time in patients treated for PCM, but these are unlikely to compromise a robust analysis that is based on assessment of multiple antigens.

In general, a sensitivity of 10 ⁻⁴ is considered to be a minimal requirement for an MRD detection method. The European Myeloma Network Report recommends a minimum of 100 aberrant plasma cell events to make a diagnosis of MRD. Thus, to achieve a sensitivity of 10 ⁻⁴ , a total of 1 million events needs to be acquired in a single-tube analysis. Note that this group does not require all 100 events to be present in the same tube, just that the aberrant plasma cells events across tubes totals at least 100. If one is using a multitube analysis, fewer events need to be acquired per tube to satisfy the European Myeloma Network recommendation.

Diagnostic Issues

Flow cytometry analysis contributes to the diagnosis of PCM by identification of clonal and aberrant plasma cells. Although the diagnosis of PCM is generally made independent of flow cytometry analysis, it may play a decisive role in the differential diagnosis in some cases, as discussed later.

Unusual Morphologic Variants of Myeloma

Occasionally myelomas are encountered that are difficult to recognize by morphologic evaluation, especially anaplastic myelomas and those with strikingly lymphoid or lymphoplasmacytoid cytologic features. Detection of a characteristic immunophenotype by flow cytometry will help to discriminate these from other neoplasms. It is worth mentioning that expression of CD20 in PCM is often associated with lymphoplasmacytoid morphology, creating an additional diagnostic challenge. Notably, however, co-expression of CD19 and CD20 in PCM is extremely rare ( Fig. 26-30 ).

Florid Reactive Plasmacytosis

Although uncommon, reactive bone marrow plasma cell proliferations can reach proportions at which there is a possibility for confusion with PCM. Associations with florid reactive bone marrow plasmacytosis include autoimmune disorders, carcinomas, Hodgkin's lymphoma, drug-induced agranulocytosis, HHV-8–associated mutlicentric Castleman's disease, and HIV. Demonstration of a normal plasma cell immunophenotype and polytypic cytoplasmic light chain expression can serve to discriminate florid reactive plasmacytosis from PCM. Correlation with other clinical and laboratory features and the application of immunohistochemistry or in situ hybridization for light chain can also serve to make this distinction.

Non-Hodgkin's Lymphomas With Extreme Plasma Cell Differentiation

Various non-Hodgkin's lymphomas (NHLs) may show plasmacytic differentiation of the neoplastic cells, most commonly marginal-zone lymphomas and lymphoplasmacytic lymphoma. Occasionally, the plasmacytic differentiation may be so prominent as to be confused for plasmacytoma or myeloma. The differential diagnosis depends on the detection of an abnormal, clonal B-cell population associated with the clonal plasma cells. When this population is very small, its recognition may be difficult or impossible with light microscopy and immunohistochemistry. Flow cytometry, with its enhanced sensitivity for detecting minor abnormal B cell populations, is well suited to make this distinction. Additionally, immunophenotypic differences have been described between the clonal plasma cells in non-Hodgkin's lymphoma and those of myeloma. The most useful discriminating feature appears to be CD19 positivity in the clonal plasma cells seen in >90% of lymphomas with plasmacytic differentiation, versus only rarely in PCM. The clonal plasma cells in NHL are also more likely to express CD45 and surface immunoglobulin, and less likely to express CD56, than those of PCM.

Prediction of Genetic Abnormalities

Various immunophenotypic features have been associated with genetic subgroups of myelomas, including CD19, CD20, and CD23 expression with the t(11;14); CD28 expression with the 17p deletion and t(4;14); lack of CD117 with the 13q deletion, non-hyperdiploidy, and IGH translocation; CD117 and CD56 expression with hyperdiploidy; and lack of CD27 with t(4;14) and t(14;16). * These associations lack sufficient sensitivity and/or specificity to be clinically useful.

* References .

Prognostic Issues

Quantitative Issues at Diagnosis

Percentage of myeloma cells in bone marrow aspirates has long been recognized as a prognostic factor in myeloma, although it usually does not maintain its significance in multivariate analysis. Paiva and colleagues demonstrated that the number of plasma cells enumerated by flow cytometry was a significant predictor of overall survival in a multivariate analysis, along with patient age and high-risk cytogenetics, whereas morphologic plasma cell count was not significant.

As detailed earlier, the bone marrow from patients with PCM at initial diagnosis typically contains few or no normal plasma cells, but in a minority of cases the marrow contains greater than 3% or greater than 5% normal plasma cells/total plasma cells (different cutoffs have been used in different studies). Paiva and associates recently demonstrated that greater than 5% normal plasma cells as a percentage of total plasma cells in diagnostic PCM marrows (14% of cases in their series) was associated with significantly better progression-free and overall survival, although this was not significant in a multivariate analysis that incorporated cytogenetics.

Minimal Residual Disease

The prognostic value of qualitative MRD determination (positive or negative) by flow cytometry after various therapies has now been documented in various studies, and flow cytometric remission appears to be a more powerful predictor of outcome than either complete or stringent remission based on immunofixation and free light chain analysis, respectively. More recently, quantitative flow cytometry of log reduction in MRD over time has been demonstrated to be an independent prognostic indicator.

Circulating Plasma Cells

Presence and/or number of circulating plasma cells in the blood have been investigated as a risk factor in PCM. Nowakowski and colleagues found that the presence of greater than 10 clonal plasma cells per 50,000 mononuclear cells in the blood at diagnosis of PCM was an independent predictor of poorer overall survival. Similarly, Dingli and associates showed that the presence of detectable clonal plasma cells in the peripheral blood at the time of autologous stem cell transplantation for PCM was an independent predictor of poorer outcome.

Immunophenotype (Immunohistochemistry)

Immunohistochemistry can supplement flow cytometry or provide the primary immunophenotypic assessment for plasma cell neoplasms when a specimen is not obtained for flow cytometry or contains an inadequate number of plasma cells for analysis. The following are indications for immunohistochemical stains on bone marrow biopsies or other tissues in the assessment of plasma cell neoplasms.

• Assessment of quantity of plasma cells in bone marrow biopsies

• Identification of a monoclonal (vs. polyclonal) plasma cell proliferation

• Identification of unusual morphologic variants of myeloma

• Distinction of myeloma from other neoplasms

Plasma cells may be difficult to recognize and quantify in suboptimally prepared sections and when distributed interstitially in the marrow. Stains for plasma cell associated antigens (e.g., CD138, CD38, and kappa and lambda) will usually stain plasma cells brilliantly on biopsy sections, allowing easy quantification.

Immunohistochemical stains and in situ hybridization for kappa and lambda light chains are useful in characterizing malignant plasma cell proliferations and differentiating them from reactive causes of increased plasma cells such as connective tissue disorders, chronic liver disease, chronic infections, and metastatic tumors. Normal/reactive plasma cells and myeloma plasma cells are both rich in cytoplasmic immunoglobulin and generally react strongly with antibodies to kappa or lambda light chains. In normal marrow and in reactive plasma cell proliferations, there is a polyclonal pattern of kappa and lambda staining plasma cells, usually with a slight to moderate kappa predominance ( Fig. 26-31 ). In cases of myeloma, the plasma cells express a monoclonal pattern of reactivity. Neither the number of marrow plasma cells nor the quantity of M-protein correlate well with the light chain ratio. Kappa and lambda stains are particularly useful in cases with a relatively low percentage of marrow plasma cells. Stains for known aberrantly expressed antigens may also be used to detect populations of neoplastic plasma cells, especially CD56 and CD117.

Immunohistochemistry is often important in distinguishing a poorly differentiated myeloma from a lymphoma or metastatic solid tissue tumors. Stains for kappa and lambda light chains and CD138, along with stains for antigens associated with other neoplasms considered in the differential diagnosis, are usually diagnostic. Of note, interpretation of CD138 expression by a poorly differentiated neoplasm must be done with caution. Although CD138 is plasma cell specific among normal hematopoietic cells, it is expressed by some B-cell lymphomas and is positive in a number of metastatic carcinomas. In the differential diagnosis of a suspected poorly differentiated PCM, a positive CD138 stain should be supported by other plasma cell markers, especially Ig light chain stains.

Genetics

Cytogenetics and molecular genetics of plasma cell neoplasms have been extensively studied in the past 2 decades. The information derived from these studies has profoundly expanded knowledge of the pathogenesis of these diseases and had major impact on diagnosis and management of patients. Genetic findings are the most important indicator of risk at the time of diagnosis and the major factor in risk-stratification protocols. This discussion will provide an overview of the genetics of plasma cell neoplasms, recommendations for genetic testing, and a molecular genetic classification recommended by the IMWG.

Both numerical and structural chromosome abnormalities occur in PCM and include trisomies, translocations, whole or partial chromosome deletions, and partial duplications; complex cytogenetic abnormalities are frequent. Abnormalities of every chromosome have been reported. There are two major groups of genetic abnormalities in PCM: hyperdiploid (~60% of cases) and non-hyperdiploid (~40% of cases). Non-hyperdiploid cases have structural chromosomal abnormalities. The most frequent structural change in this group is translocations involving the heavy chain locus (IGH) on chromosome 14q32. Recurrent partners in these IGH translocations include the following oncogenes: cyclin D translocations—11q13 CCND1 (15%), 12p13 CCND2 (<1%), and 6p21 CCND3 (2%); FGF-R3/MMSET translocation—4p16.3 (15%); MAF translocations—16q23 CMAF, (5%), 20q11 MAFB (2%), and 8q24 MAFA (<1%). Hyperdiploid myelomas lack recurrent translocations and manifest trisomies of odd-numbered chromosomes: 3, 5, 7, 9, 11, 15, 19, 21. There is minimal overlap between hyperdiploid and non-hyperdiploid groups, but some hyperdiploid cases have secondary non-recurrent translocations involving 14q32.

IGH translocations and hyperdiploidy are early, and probable initiating, events in the genesis of plasma cell neoplasms. Dysregulation of one of the cyclin D genes (D1, D2, D3) is the unifying feature of the two genetic groups. Overexpression of one or more cyclin is found by gene expression profiling (GEP) in nearly all myelomas. Direct or indirect dysregulation occurs in cases with a CCND D or MAF translocation. The mechanism is not understood in MMSET/FGFR3 myelomas, which have a high level of CCND2 expression. Hyperdiploid myelomas with trisomy 11 overexpress CCND1 or CCND1 and CCND2 . The mechanism involved in hyperdiploid tumors without trisomy 11, which mostly overexpress CCND2 , is not fully understood.

Investigators have classified plasma cell myeloma using patterns of translocations (T) and cyclin (C) D expression into groups based mostly on early pathogenic events (TC groups). These TC groups seem to represent distinct biologic entities that may have prognostic significance. A molecular classification that is similar to the TC groups, but not identical, consists of seven distinct groups of myeloma that are based on unsupervised clustering of tumors by GEP. Other investigators have identified a 10-subgroup classification of myeloma by GEP.

Genetic events are essential initiating factors in plasma cell neoplasms, but it is less clear why some patients progress from MGUS to symptomatic myeloma and others with identical or similar genetic abnormalities do not (see the discussion of genetics in the section on MGUS). It seems that additional but less well-studied pathogenic events are necessary for progression. These events most likely involve secondary genetic aberrations. Several genetic abnormalities are found more frequently in symptomatic myeloma than in MGUS, and these may represent markers of disease progression. Deletion or mutation of TP53 (17p13), IGH or IGL translocations, MYC or MYCN translocations , losses of chromosome 1p and gains of 1q, mutations of genes resulting in activation of the NF-κB pathway, inactivation of CDKN2C or RB1, activating mutations of KRAS or NRAS , and mutations of FGFR3 in myeloma with t(4;14) are all found significantly more frequently in myeloma than in MGUS. *

* References .

Factors involving the bone marrow microenvironment may also play a key role in disease progression. Extracellular matrix proteins, cytokines, and growth factors as well as the functional consequences of interaction of bone marrow stromal cells with the neoplastic plasma cells all seem to influence the pathophysiology of myeloma.

Genetic Testing for Plasma Cell Myeloma

For many years, the standard for detecting genomic abnormalities and outcome discrimination in PCM was conventional karyotype cytogenetic studies. This technique remains an important component of genetic assessment but has a relatively low sensitivity. Only 30% to 40% of PCMs have identifiable abnormalities by karyotype analysis. The low rate of detection is attributable to low in vitro proliferation of many myelomas and the fact that a number of the important structural changes in myeloma are cryptic. Despite its relatively low sensitivity, conventional cytogenetic analysis should still be performed at diagnosis of PCM. Positive numerous abnormalities including complex changes can be appreciated by karyotype cytogenetics, some of which would not be detected by FISH studies. For example, patients with important prognostic karyotype changes such as deletion of chromosome 13 and hypodiploidy may not have FISH-defined risk abnormalities.

Directed fluorescent in situ hybridization (FISH) should be performed in all cases of myeloma at the time of diagnosis. Interphase FISH does not require in vitro mitoses of the neoplastic plasma cells providing a far more sensitive method of detection of genetic aberrancies. More than 90% of PCMs have detectable abnormalities by FISH analysis. FISH also detects important cryptic genetic abnormalities like the t(4:14)(p16;q32) FGFR3/ IGH that would be missed by conventional karyotyping. FISH studies have become the major methodology for establishing risk-based stratification of patients with myeloma.

An important technical caveat on FISH analysis is that often the clonal plasma cell percentage in a bone marrow specimen is below the necessary limit for a successful FISH study. Concentrating the plasma cells by cell sorting or cytoplasmic immunoglobulin-enhanced FISH are recommended as they significantly improve the yield of positive results.

GEP is a powerful technique for patient risk stratification. The GEP distinguishes high-risk and low-risk myelomas and is the most sensitive and specific technique for identification of high-risk PCM. Overlapping but different gene signatures for risk stratification have been developed by different investigators but with similar results. A well-validated risk-stratification model uses 70 genes (GEP70) linked to high-risk. Other investigators have reported a 15-gene model of risk that also seems to be effective. Although important in clinical trials, the use of GEP in clinical routine depends on some technical and logistical resolutions.

An example of a risk-adapted therapy scheme that uses genetics as the primary criteria is the Mayo Stratification of Myeloma and Risk-Adapted Therapy (mSMART), shown in Table 26-2 . Data from each of the three genetic techniques discussed earlier are used in this system, but the major component is based on FISH analysis. The three risk-groups have significantly different overall survival rates.

The International Myeloma Working Group molecular cytogenetic classification of PCM and recommendations on genetic testing are shown in Boxes 26-6 and 26-7 .

Differential Diagnosis

The most common differential diagnosis among the plasma cell neoplasms is that of early myeloma versus MGUS or a reactive bone marrow plasmacytosis. In most cases, this is not difficult because the composite clinical and pathologic findings required for a diagnosis of myeloma are lacking in MGUS and in reactive plasma cell proliferations. Only when the M-protein or percentage of bone marrow plasma cells are at the high extreme for MGUS is the distinction from asymptomatic myeloma problematic. In some patients, differentiation of early myeloma and MGUS is not possible at the time of initial evaluation. Close observation and monitoring for evidence of progression to overt malignancy must be continued indefinitely.

Reactive bone marrow plasmacytosis of 10% or more may occur in several conditions including viral infections, immune reactions to drugs, autoimmune disorders such as rheumatoid arthritis and lupus, and AIDS. Reactive plasmacytosis is distinguished from myeloma by the lack of an M-protein in the serum or urine in most instances. The plasma cells are generally mature appearing, and stains for kappa and lambda light chains on bone marrow sections show a polyclonal plasma cell staining pattern (see Fig. 26-31 ). The rare systemic polyclonal immunoblastic proliferations are among the most difficult reactive plasma cell proliferations to differentiate from myeloma. The disorder is uncommon and usually presents as an acute systemic illness with fever, lymphadenopathy, and hepatosplenomegaly; anemia and thrombocytopenia are present in most patients. Autoimmune manifestations are often present. The leukocyte count is usually elevated with large numbers of plasma cells, immunoblasts, and reactive lymphocytes, and there is eosinophilia and neutrophilia in some cases ( Fig. 26-32 ). The bone marrow is heavily infiltrated by immunoblasts, plasma cells, and reactive lymphocytes (see Fig. 26-32 ). Lymph nodes and other organs may also be involved. Usually marked polyclonal hypergammaglobulinemia is present, but there are no M-protein or bone lesions. Patients usually respond to steroid therapy alone or to chemotherapy with complete resolution of the polyclonal immunoblastic proliferation.

Occasionally myeloma must be distinguished from a lymphoma with extreme plasma cell differentiation such as lymphoplasmacytic lymphoma, marginal-zone lymphoma, immunoblastic large-cell lymphoma, or plasmablastic lymphoma ( Fig. 26-33 ). Any of these may show morphologic similarities to myeloma and be associated with an M-protein. In most cases, lymphomas with plasma cell differentiation present with extramedullary disease, and at least some of the diagnostic criteria for myeloma are lacking. Careful morphologic study will usually distinguish these tumors by identification of areas with features of lymphoma, and a clonally related lymphocyte population may be identified by immunophenotyping. The distinguishing immunophenotypic features are discussed in detail in the sections on flow cytometry—diagnostic issues, earlier in the chapter and later in the section on differential diagnosis of extraosseous plasmacytoma. Genetic and molecular studies are also useful in differentiating these disorders.

Myelomas composed of small lymphoid-appearing plasma cells often express CD20 and may mimic a lymphoplasmacytic or marginal-zone lymphoma with extreme plasma cell differentiation. They may be distinguished from lymphoma by cyclin D1 positivity and a t(11:14) chromosome rearrangement. Plasmablastic lymphoma usually differs from plasmablastic myeloma in clinical presentation and its frequent association with HIV and EBV. In clinically atypical cases, especially when the bone marrow is involved at presentation, there may be no defining features that distinguish the two disorders.

Plasma cell leukemia in which the plasma cells are small with lymphoid features may be especially difficult to differentiate from a peripheralized lymphoplasmacytic lymphoma. The combination of clinical findings, type of M-protein, bone marrow examination, immunophenotype, and genetics usually lead to the correct diagnosis.

Several metastatic tumors may present with lytic bone lesions and bear morphologic resemblance to myeloma. Immunohistochemical staining with an appropriately selected panel of antibodies usually resolves the issue.