Chapter 105: Plasma Cell Neoplasms: General Considerations

Plasma cell neoplasms (PCNs) are clonal B-cell tumors that range from stable disease without functional abnormalities (monoclonal gammopathy, synonym monoclonal gammopathy of unknown significance) to one of slowly proliferating plasma cells (smoldering myeloma [SMM]), to one resulting in end-organ compromise (myeloma). PCNs are accompanied by the synthesis and release into the plasma of a monoclonal (M) protein, and, in the case of myeloma, either diffuse osteoporosis or osteolytic lesions. Myeloma accounts for approximately 1 percent of all malignant diseases and 10 percent of hematologic malignancies.

Approximately two-thirds of patients presenting with an M protein have (1) monoclonal gammopathy (Chap. 106), whereas approximately 15 percent have (2) myeloma (Chap. 107). Other diseases associated with M-protein productions are (3) immunoglobulin light-chain amyloidosis (AL) (10 percent; Chap. 108) resulting from the deposition of immunoglobulin (Ig) fragments in visceral organs a consequence of extensive misfolding of these Ig fragments, (4) SMM (3 percent; Chap. 107), (5) Waldenström macroglobulinemia (3 percent; Chap. 109), (6) lymphoproliferative disorders (2 percent; Chap. 90), (7) solitary or extramedullary plasmacytomas (1 percent; Chap. 107) and (8) miscellaneous other diseases (2 percent).³

B-cell development is discussed in detail in Chaps. 74 and 75. In brief, B-cell lymphopoiesis occurs initially in the marrow and in lymphoid tissues. In the marrow, the pro–B-cell, undergoes rearrangement of immunoglobulin heavy chain (IGH) genes and, then, is designated a pre–B-cell, which is characterized by the presence of cytoplasmic μ chains. Subsequent rearrangement of the light chain enables the cell to express surface IgM, the immature B lymphocyte phase of development. These cells leave the marrow and upon entering the blood express surface IgD, which then defines them as virgin B cells, also characterized by G₀ cell-cycle arrest. Virgin B cells enter the lymphoid tissue, where they are exposed to antigen-presenting cells, become activated when in contact with the corresponding antigen and differentiate into short-lived, low-affinity plasma cells or memory B-cells. These memory B-cells travel from the extra-follicular area of the lymph node to the primary follicles, where if confronted with an antigen, presented by follicular dendritic cells, a secondary response is induced. At this stage, primary follicles change into secondary follicles containing germinal centers. Through activation by an antigen, the memory B cells differentiate into centroblasts, resulting in Ig isotype switching and somatic mutations in the variable region of the immunoglobulin gene with the generation of high-affinity antibodies. Centroblasts progress to the centrocyte stage and reexpress surface Ig. The centrocytes with high-affinity antibodies differentiate into either memory B cells or plasmablasts, which then move to the marrow and terminally differentiate to plasma cells. Marrow plasma cells produce most of the plasma immunoglobulins and have a life span of approximately 3 weeks.

Three distinct gene segments, the variable (V_H), diversity (D), and joining region (J_H) genes, encode the variable region of the heavy chain, whereas two segments, variable (Vκ or V) and joining (Jκ or J) region genes, encode the variable fraction of the light chain. The IGH locus on chromosome 14q32 contains an estimated 100 to 150 V_H genes, 30 D, and six J_H gene segments. Because some of the V_H genes are nearly identical, it is likely that 60 to 70 V_H genes are available for rearrangement. These 60 to 70 genes belong to seven families (V_H 1 to 7) whose members have more than 80 percent sequence homology. Of the 75 known Vκ sequences, only 36 are potentially functional and of the 36 known V sequences, only 24 are functional. Rearrangement of the V gene segments is dependent on the protein products of the recombinase-activating genes RAG-1 and RAG-2. Recombination of V genes starts in lymphoid progenitors within the IGH locus of either the maternal or paternal chromosome 14. If the initial V_H-D-J_H rearrangement yields a sequence that cannot be translated, then rearrangement of the IGH locus proceeds on the other allele. The presence on the B-cell surface of a fully assembled μ heavy chain rearrangement begins when one of the Vκ genes rearranges to one of the Jκ genes. If κ light-chain rearrangement is unsuccessful on both alleles, by default light chains will subsequently rearrange. The Ig heavy and light chains each contain three hypervariable complementarity determining regions segments, which are the areas of the immunoglobulin in direct contact with the antigen. In a process of trial and error, immunoglobulins increase their affinity for an antigen by a series of somatic mutations. The complementarity determining regions of the IGH (CDR) 3 is the most variable portion of the Ig molecule, because it not only contains somatic mutations as is the case for CDR 1 and CDR 2, but it also encompasses the 3′ end of V_H, all of D, and the 5′ end of J_H. It is, therefore, an ideal marker to detect a very small population of the malignant myeloma clone within a larger population of normal cells.

ETIOLOGY OF PLASMA CELL NEOPLASM

Monoclonal Gammopathy

Although monoclonal gammopathy (Chap. 106) shares the same constellation of risk factors and cytogenetic abnormalities with myeloma, it is an antecedent neoplasm that may undergo clonal evolution to any one of the PCNs or to a B-cell lymphoma.¹ Two studies have reported that monoclonal gammopathy is a precursor to myeloma in virtually all cases.²³

Retrospective population-based cohort studies have established that nearly 80 percent of cases of myeloma develop from IGH monoclonal gammopathy. The remaining 20 percent have serum free light chain (FLC) monoclonal gammopathy. The prevalence of FLC monoclonal gammopathy is 0.8 percent in the general population. It progresses to myeloma in a minority of patients at the rate of 0.3 percent per year,³ much lower than the conventional monoclonal gammopathy progression rate of approximately 1 percent.

Factors such as chronic antigen stimulation and chemical exposure have been suspected in the development of monoclonal gammopathy and other PCNs. Some studies have found a positive association,⁴⁵ but the results have not been consistent and given our current understanding of the genetic precedents of myeloma, one would have to show a direct effect of such agents on the causal mutations involved.

A familial history of monoclonal gammopathy and myeloma has been reported to be a risk factor for developing the disease in first-degree relatives, including a population-based study from the Mayo Clinic. A twofold increased relative risk was noted for the development of monoclonal gammopathy among the first-degree relatives of myeloma and monoclonal gammopathy patients. In a large Swedish population study, among first-degree relatives of patients with monoclonal gammopathy, a threefold increased risk for both monoclonal gammopathy and myeloma, a fourfold risk of developing Waldenström macroglobulinemia, and a twofold risk of developing B-cell chronic lymphocytic leukemia was observed.⁶ These observations support the role of both germline susceptibility genes and possibly immune-related phenomena in the causation.⁷ Because of the extremely low lifetime-risk of developing myeloma in the general population (0.2 percent), it would be inefficient to screen the first-degree relatives of persons with myeloma or monoclonal gammopathy.

Hyperphosphorylated paratarg-7, a frequent autoantigenic target of human paraproteins, is linked to both familial and nonfamilial forms of monoclonal gammopathy and myeloma.⁷ Paratarg-7 is the target of 15 percent of the monoclonal proteins of the IgA and IgG type, and 11 percent of IgM-monoclonal gammopathy or macroglobulinemia patients. All patients with paratarg-7–specific paraproteins were carriers of a hyperphosphorylated protein (pP-7); this hyperphosphorylation is inherited in a dominant fashion. Hyperphosphorylation is a result of inactivation of phosphatase 2A.⁸ Hyperphosphorylated paratarg-7 carriers are most prevalent among Americans of African wdefined single risk factor for monoclonal gammopathy or myeloma in all ethnic groups⁹ and is associated with a sixfold increased risk of IgM monoclonal gammopathy or macroglobulinemia.¹⁰

Nearly 50 percent of patients with monoclonal gammopathy have plasma cells with translocations involving the IGH locus on chromosome 14q32 and one of the five partner chromosomes: 11q13 (cyclin D1 gene), 4q16,3 (FGFR-2 and MMSET), 6q21 (CCND3), 16q23 (c-maf), and 20q11 (maf-B).^11,12,13,14

Unfortunately, none of the molecular or chromosomal abnormalities associated with myeloma predict the evolution of monoclonal gammopathy to myeloma. Two clinical risk stratification models propose high-risk features that can predict progression from monoclonal gammopathy and SMM to myeloma.^15,16 While one model uses the type of immunoglobulin, quantity of M-protein, and the serum FLC ratio to determine the risk of progression, the other model is based on flow cytometry findings of aberrant plasma cells, marrow plasma cell percentage, DNA aneuploidy, and immune paresis (a decrease in noninvolved immunoglobulins).

Smoldering Myeloma

SMM is discussed in Chap. 107. In addition to the marrow plasma cell burden and quantitative M-protein (>3 g/dL), presence of light-chain proteinuria and IgA M-heavy chain were identified as separate risk factors predicting progression to active myeloma.^17,18,19,20 The median time of progression to myeloma has been reported to range between 3 and 8 years in low-risk groups, and between 1 and 2 years in high-risk groups.^{17,18,19,20,21,22} Some studies have also investigated the use of magnetic resonance imaging (MRI) in detecting skeletal abnormalities not seen on a skeletal survey; time to progression to myeloma was much shorter in patients with focal lesions on the MRI²³ and these patients should be treated.

A number of models estimating the risk of progression to myeloma have been proposed. The presence of serum M-protein of greater than 3 g/dL, an FLC ratio outside the reference range of 0.125 to 8, and greater than 10 percent plasma cells in the marrow represents SMM with a high-risk of progression. Patients with these three risk factors had a cumulative risk of 76 percent of progression to myeloma within 5 years.^15,24 The risk of progression decreased to 51 percent in patients with two of the risk factors and to 25 percent in patients with a single risk factor.^24,25 Another clinical risk stratification model uses flow-cytometry of marrow aspirates. Using as risk factors (1) greater than 95 percent of all plasma cells being aberrant at diagnosis, (2) DNA aneuploidy, and (3) immune paresis, the presence of one, two, or three risk factors translated to 4, 46, and 72 percent risk of progression to myeloma with 5 years of observation, respectively.¹⁶ Other researchers have found that in addition to the intrinsic, molecular, and cytogenetic abnormalities of the plasma cells, an angiogenic switch and immunologic factors play a key role in the transformation of SMM to established myeloma.²⁶

Myeloma

The development of myeloma is a complex multistep process involving karyotypic instability, Ig translocations, cell-cycle abnormalities (cyclin Ds), and multiple other mutations²⁷ (Chap. 107). No molecular or chromosomal abnormalities can distinguish among monoclonal gammopathy, SMM, or myeloma at the time of diagnosis. Certain mutations occur in much higher frequencies in myeloma, such as p53 deletions, especially in refractory and extramedullary presentations,^12,28 N-RAS and K-RAS mutations, chromosome 1p deletion and gain of 1q21, and translocations involving MYC (8q24).^{29,30,31,32,33} In whole-exome sequencing studies, intraclonal heterogeneity has been shown to be present at all stages of development—from monoclonal gammopathy to SMM to progressive myeloma.³⁹ Genetic complexity increases as the disease progresses to myeloma.

Effect of Endogenous Factors There is an increased relative risk of monoclonal gammopathy and myeloma in overweight and obese patients based on their body mass index (BMI). With the exception of elite athletes, a BMI of 25.0 to 29.9 kg/m² and 30 kg/m² or greater define overweight and obese individuals, respectively.^{35,36,37,38,39,40,41} Fat tissue is a dynamic endocrine organ, secreting adipokines, hormones that play an important role in energy homeostasis and inflammation. Several adipokines, such as leptin and adiponectin, have been implicated in the development of cancer.⁴⁰ Adiponectin levels are inversely correlated with obesity. Adiponectin serum concentrations were lower in monoclonal gammopathy patients who subsequently developed myeloma. In the KaLwRij strain of C57 black mice, which is permissive to the growth of 5T myeloma, compared to the parental strain of C57B16, which is not, adiponectin gene expression is significantly lower than in the parental strain. An increased myeloma burden was found in adiponectin-deficient mice, while pharmacologic enhancement of circulating adiponectin resulted in apoptosis of myeloma cells and also prevented bone disease. Fat tissue is a principal source of interleukin (IL)-6, one of the principal growth and antiapoptotic cytokines acting on myeloma cells. Obese individuals have been shown to have shorter telomeres than nonobese individuals. Because telomeres protect chromosomes from injury, including undesirable translocations, this effect may also contribute the relationship of body mass with PCN.

Effect of Exogenous Factors Aspirin has been shown not only to reduce cancer incidence, but to also dramatically decrease cancer mortality, especially in colorectal cancer, esophageal, gastric cancer, breast cancer, prostate cancer, and lung cancer. Aspirin inhibits several pathways that are important in myeloma, including nuclear factor κB (NF-κB), AKT activation, and the BCL-2 family of proteins. Aspirin is used frequently as thromboprophylaxis in myeloma patients receiving immunomodulatory therapy. In a prospective study designed to examine whether regular aspirin use influences the risk of myeloma, participants taking 5 or more tablets of 325 mg per week had a 39 percent lower myeloma incidence than nonusers. The association appeared stronger in men than in women.⁴³ Aspirin inhibits proliferation and induces apoptosis of myeloma cell lines in vitro through regulation of BCL-2 and BAX and suppression of vascular endothelial growth factor (VEGF). In addition, in vivo studies in mice showed that aspirin administration resulted in retardation of tumor growth and in increased survival.⁴⁴ A number of case-control and cohort studies have established that smoking has no association with the incidence of myeloma. Convincing evidence has not been found linking alcohol consumption to myeloma development.⁴⁵

Occupation Many studies evaluated the potential role of exposure to certain occupations and/or toxin and the subsequent risk of myeloma development. Agricultural workers and farmers were studied in the United States and Europe. The majority of studies report an increase frequency of myeloma in agricultural workers, whereas other reports fail to find such a correlation.⁴⁶ Exposure to toxins such as organic solvents (e.g., toluene, benzene), pesticides, paints, and others products with trace benzene content have been investigated for an association with the incidence of myeloma, but the findings are inconsistent.⁴⁷

Radiation Studies on the survivors of the atomic bombing in Japan have failed to establish a cause-and-effect relationship between high-dose radiation exposure and an increased incidence of myeloma.⁴⁸ Studies from the United Kingdom have reported no increased frequency of myeloma in workers exposed to ionizing radiation, nuclear plants, and/or plutonium.⁴⁵ In a large study in China of x-ray technicians, there were no reported cases of myeloma or plasma cells disorders.⁴⁹ Thus, radiation exposure is not linked to the risk of myeloma.

Chronic immune stimulation has not been shown to play a causative role in the etiology of myeloma. No link between infections, allergic conditions, or immunizations and the development of myeloma has been established. Patients with autoimmune disorders, in general, have not been found to have an excess risk of myeloma. Some studies report an increased risk of myeloma in HIV^50,51 and hepatitis C patients,⁵² although more convincing data are needed to establish a cause-and-effect relationship.

Waldenström Macroglobulinemia

Etiology In a small study of 65 patients and 213 controls, a preceding autoimmune condition did not predict the development of Waldenström macroglobulinemia.⁵³ In contrast, many other studies have found such a link. In a large population-based study, that included 146,394 hepatitis C patients and 572,293 controls, a threefold increased risk of macroglobulinemia was observed along with a 20 to 30 percent increased risk of lymphoma.⁵⁴ In a study of 361 U.S. veterans with Waldenström macroglobulinemia after a 27-year followup, a two- to threefold increased risk of developing the disease was found in patients with autoimmune-related conditions, hepatitis, HIV infection, and rickettsiosis.⁵⁵ In two Swedish population-based studies, a personal history of autoimmune conditions and infections was associated with an increased risk of macroglobulinemia.⁵⁶

Case-control studies and a large population-based study in Sweden have established the role of familial clustering of macroglobulinemia, thereby raising the concept of common susceptibility genes that could predispose to the disease.^{53,57,58,59,60,61} In study of macroglobulinemia, 19 percent of patients had at least one identified first-degree relative with macroglobulinemia or a B-cell disorder.⁶⁰ In genome-wide linkage analyses of high-risk families with macroglobulinemia and IgM monoclonal gammopathy, the strongest linkage was found to be chromosomes 1q, 3q, 4q, and 6q.^{58,60,62,63,64} In a gene-sequencing study performed on marrow cells of macroglobulinemia patients, a recurrent somatic mutation of the gene MYD88L265P on chromosome 3 that encodes signal transduction and innate immunity was found to be in approximately 91 percent of the patients tested.⁶⁰

In a novel study using expression cloning, several common self-antigens designated Paratarg-7 were discovered.¹⁰ When carriers of the phosphorylated forms of Paratarg-7 were studied, they were found to have a 6.5-fold higher risk of developing IgM monoclonal gammopathy or macroglobulinemia. The antigen causes continuous autostimulation of cognate T-helper cells, which, in turn, specifically activate B cells with high affinity to Paratarg-7.

MYELOMA

Plasma cell differentiation begins in lymph nodes and in the spleen where cells undergo changes in gene expression and cell-surface molecules followed by migration to the marrow or mucosal lamina proper.⁶⁵ The development of plasma cells alters the cellular receptor landscape of the cell with the deletion of important B-cell receptors and the addition of receptors necessary for plasma cell function and antibody production. Changes to cellular receptors include downregulation of major histocompatibility complex (MHC) class II, CD19, CD21, and CD22.⁶⁶ Perhaps, the most important alterations during myeloma development are decreases of the B-cell receptor (BCR), CXCR5, and CCR7. In contrast, plasma cells upregulate CXCR4, CD138, and CD38 (Fig. 105–1).^66,67,68 Plasma cells also undergo changes to transcription factors highlighted by a decrease in PAX5, CIITA, and EBF.^{66,67,68,69,70} Furthermore, plasma cells express genes that are present in B cells at low levels or not at all and are highlighted by increased expression of Blimp-1, IRF4, and XBP1, the only transcription factor exclusively required for plasma cell development.⁷¹ For more information on plasma cell development (Chap. 74).

Early Genetic Events in Myeloma Genesis

Early genetic events in myeloma genesis include the accumulation of sequential genetic changes; however, the full mechanism remains elusive.⁷² Protein kinases provide selective growth advantages to cells and, therefore, act as driver mutations, or inducers of early neoplastic events.⁷³ The dysregulation of the cyclin D genes exposes cells to additional proliferative stimuli and commonly occurs as a result of a translocation of the cyclin D gene to the Ig loci.⁷⁴ Activation-induced deaminase contributes to genetic instability through its involvement in induction of somatic mutations in the immunoglobulin genes and immunoglobulin translocations.⁷⁵ Although important, gene alterations are not the only driving force in the development of PCN. Although ethnicity and genealogy affect the prevalence of monoclonal gammopathy,⁷⁶ they do not impact the rate of progression from monoclonal gammopathy to myeloma. Furthermore, there is a higher incidence in monoclonal gammopathy in relatives of myeloma patients than seen in the general public.^77,78 These results demonstrate that not only are there somatic mutations important to the development of PCNs, but genetic background also contributes to the likelihood of the development of plasma cell diseases.

Late Events in Myeloma Progression

Much focus has been placed on risk factors and early genetic events that can be used as diagnostic and prognostic markers. However, we are learning more about the genetic alterations that occur throughout disease progression. One of the more prominent discoveries focuses on the activation of NF-κB. Approximately 50 percent of myeloma patients exhibit canonical NF-κB activation.^79,80 NF-κB functions, in part, by regulating growth and survival within the myeloma cells; overexpression of positive regulators, such as NIK, NFKB1, NFKB2, and CD40, and inactivation of negative regulators, such as CLYD, TRAF2, TRAF3, and cIAP1/cIAP2, contribute to the constitutive activation of NF-κB within these cells.⁸⁰

RAS mutations are also important contributors to the development of myeloma from monoclonal gammopathy. Oncogenic activation of RAS occurs through mutation of one of three different codons with mutations resulting in constitutively activated RAS. Less than 5 percent of cases of monoclonal gammopathy display mutations of RAS, whereas in newly diagnosed myeloma patients mutated RAS is found in nearly 40 percent of patient samples, suggesting that RAS may be associated with the conversion of monoclonal gammopathy to myeloma.^81,82,83

p53, which regulates the cell cycle and acts as a tumor suppressor represents the most commonly inactivated tumor-suppressor gene in cancer.^84,85 In newly diagnosed PCNs, p53 mutations occur in 5 percent of patients. The frequency of p53 mutations increases with disease progression; while infrequent in newly diagnosed myelomas, 30 percent of plasma cell leukemia patients present with p53 mutations.^86,87 Furthermore, p53 mutations are negatively correlated with survival.⁸⁸ Similar in function to p53, the retinoblastoma (RB) gene regulates the cell cycle. RB functions by inhibiting the effects of the cyclin D proteins with the help of the p18^INK4 proteins. However, overexpression of cyclin D or decreased expression of RB can lead to cell-cycle progression and neoplastic growth. Decreased expression of the two RB pathway regulators p18^INK4a and p18^INK4c can inhibit the regulatory effects of RB and result in cell-cycle release. Decreased expression of the two proteins is considered to be a late disease progression event.⁷⁶

Genetic Expression Changes in Myeloma

Myeloma is genetically heterogeneous and more closely resembles solid tumors than other hematologic malignancies. Because of its high degree of tumor heterogeneity, myeloma gene-expression microarrays have proven invaluable to our understanding and treatment of myeloma. Four stratification models exist designed to classify myeloma based on the risk profile as determined by gene-expression profiles.⁸⁹ Of the four, one model has proven most reliable, as it retains its prognostic relevance with newer therapeutic regimens. This model compares the expression of 70 different genes to devise a scoring system that ranks a sample as high risk (13 percent of patients) or low risk.⁹⁰ Abnormalities of expression that map to chromosome 1 are disproportionately represented within the model, with an impressive number of upregulated genes mapping to 1q and a high percentage of downregulated genes mapping to 1p. The 70-gene model offers a reliable method for the detection of high-risk myeloma. However, it does not take into account all important genes related to myeloma. One important gene not found in this model is MYC, which is a transcription factor that influences the expression of many genes through the binding of consensus sequences within the noncoding region of genes. It is thought to regulate the expression of 15 percent of all genes.⁹¹ MYC expression within myeloma and monoclonal gammopathy varies depending on the state of the disease. In high-risk monoclonal gammopathy and early stages of myeloma, the increase in MYC expression is primarily a result of a decrease in transcriptional regulation, whereas in late disease states, the 8q region encoding MYC translocates to the immunoglobulin loci, resulting in dysregulation of expression because of the influence of neighboring regulatory elements.³²

Drug Resistance

Refractory myeloma has several causes, with dysregulation of gene expression contributing significantly to the overall development of drug resistance. The serine/threonine kinase NEK2 induces drug resistance in myeloma through the activation of efflux drug pumps.⁹² Overexpression of NEK2 results in the activation of the AKT pathway and subsequently of NF-κB, which results in the upregulation of ABC drug efflux transporters. Overexpression of the antiapoptotic molecule MCL-1 induces drug resistance through the inhibition of apoptosis.⁹³ More specifically, MCL-1 overexpression results in the inhibition of the proapoptotic BCL-2 family members, which results in blocking apoptosis.

Next-Generation Sequencing

Next-generation sequencing or deep sequencing relies on the ability to sequence large amounts of small DNA fragments quickly and assemble the output into a complete coherent data set. This technique is providing new insights into our understanding of existing neoplastic mechanisms and identifying novel mechanisms and genes that had evaded detection thus far. One of the first reports to use next-generation sequencing was from a study that compared 38 whole-tumor genomes from patients with myeloma to matched normal DNA.⁹⁴ Novel genes involved in histone methylation, protein translation and blood coagulation were identified. Mutations to the toll-like receptor (TLR) 4, the adaptor molecule TRAF3, and to the kinases CYLD, RIPK4, and BTRC also were discovered. A greater-than-anticipated change in NF-κB was observed, including identifying 11 mutated genes involved in the NF-κB pathway. Furthermore, the prognostic value of next-generation sequencing was tested by using the sequencing technology to detect minimal residual disease (MRD) in myeloma patient samples.⁹⁵

Cytogenetic Microarrays

Fluorescence in situ hybridization (FISH) has become the gold standard for the detection and classification of myeloma. However, FISH cannot provide information about chromosomal abnormalities without the use of large scale panels of probes. Comparative genomic hybridization (CGH) arrays overcome some of the short falls of FISH technology by providing a genome-wide view of chromosomal changes, but do so at a reduced resolution of 10 to 20 megabases. To combat the low resolution of CGH arrays, small nucleotide polymorphism (SNP)-based technology has been employed. SNP arrays can detect copy number changes and have improved the resolution of CGH arrays to a submegabase level. An early report validated the use of SNP arrays by comparing its results to those obtained by FISH analysis using identical samples. Furthermore, uniparental disomy (UPD) was identified as prevalent in myeloma samples and may occur through several mechanisms, including mitotic nondisjunction and mitotic recombination.⁹⁶ A technical advance was the introduction of the Affymetrix Cyto Scan HD arrays. The Cyto Scan arrays incorporate the most up-to-date SNP library to generate a 2.6-million probe library across the human genome, allowing for resolution up to 50 kb. The technology has not been published in a myeloma study yet, but the resolution matches that of FISH and can provide a wealth of knowledge related to changes in copy number, mosaic chromosomes and loss of heterozygosity. Using these technologies numerous cytogenetic changes were identified that may be important for the development and progression of monoclonal gammopathy to myeloma.

FLUORESCENCE IN SITU HYBRIDIZATION

In PCNs, FISH has emerged as one of the most useful and reliable methods for the detection of chromosomal abnormalities and performing risk assessment on newly diagnosed patients. G-banding cytogenetic analysis requires actively dividing cells for identification of chromosomal changes, not commonly seen in plasma cells, whereas FISH analysis can overcome this limitation to an extent and detect structural abnormalities in non-dividing plasma cells.⁹⁷ A panel of FISH probes is used to query for commonly identified cytogenetic changes, including t(4;14), t(11;14), t(14;16), t(14;20), t(6;14), del(13q14), and del(17p13), and probes for hyperdiploidy (5, 7, 9, 11, 15, and 17). Cytogenetic abnormalities, using FISH, are observed in more than 90 percent of myeloma patients. Panels like the one described here provide essential information and paired with gene-expression profiles provides clinicians with essential data for diagnosis and the development of treatment regimens.

METAPHASE CYTOGENETICS

Approximately 30 percent of patients with newly diagnosed myeloma present with chromosomal abnormalities by metaphase cytogenetics, and this percentage increases to 50 percent in patients with relapsed myeloma. Cytogenetic changes broadly divide patients into one of two categories: patients with hyperdiploid chromosomal profile and nonhyperdiploid, encompassing hypodiploid and hypotetraploid cases. Hyperdiploid cytogenetic profiles are characterized by trisomies of many odd-numbered chromosomes, namely 3, 5, 7, 9, 15, 19, and 21 and are associated with a favorable outcome.⁹⁸ Nonhyperdiploid profiles are highly enriched with translocations at the IGH loci (14q32) with partner chromosomes, most importantly resulting in t(4;14), t(6;14), t(11;14), t(14;16), and t(14;20).⁹⁷ However, many of the translocations result in the activation of cell-cycle regulators and oncogenes like cyclin D, MMSET and c-MAF, MAFB. In addition, amplification of chromosome 1q and deletions to chromosomes 13q and 17p are commonly seen within myeloma patients.⁸⁶ With regards to deletion of chromosome 17p, the number of patients presenting with this abnormality increases as the disease progresses.^85,96 To accommodate the variability in cytogenetic markers, researchers have developed diagnostic criteria to aid in the management of patients with a variety of cytogenetic abnormalities with prognostic value and will be discussed below.

CYTOGENETIC ABNORMALITIES AS PROGNOSTIC MARKERS IN PLASMA CELL NEOPLASM

One model has become the standard for myeloma prognostic risk assessment: The Mayo stratification of myeloma and risk-adapted therapy (mSMART).^99,100 Historically, risk assessment was categorized into two groups, standard or high-risk. The mSMART guidelines add an intermediate-risk subgroup to better assess the appropriate treatment regimen. The standard-risk group consists of patients presenting with either the t(6;14) or t(11;14) translocation as well as the hyperdiploid group, whereas the intermediate-risk group comprises patients presenting with the t(4;14) translocation and deletions of chromosome 13 or hypodiploidy; the high-risk group is made up of patients presenting with the t(14;16), the t(14;20), or deletion 17p13. Patients with deletion of chromosome 13 by metaphase cytogenetics (not by FISH) are also considered high risk. The mSMART guidelines do not take into account all cytogenetic abnormalities seen in myeloma samples. There has been considerable research performed to identify prognostic markers within patients, including mRNAs and cell-surface receptors such as CD20+ samples as a prognostic marker.¹⁰¹ Chromosome 1 changes are important markers in myeloma; patients harboring the 1p deletion have a poor outcome.¹⁰² The amplification of chromosome 1q has also been identified as an important locus in myeloma. Amplification of 1q21 is a poor prognosis marker in PCN and the frequency of 1q21 amplifications increases as the disease progresses.³⁰ Furthermore, studies have identified the chromosome 1q genes, NEK2 and CKS1B, as markers of aggressive disease and poor prognosis.^92,103 Presently, alterations to chromosome 1 are the only chromosomal markers not incorporated into the commonly accepted prognostic models.

GENETIC ABNORMALITIES IN IMMUNOGLOBULIN LIGHT-CHAIN AMYLOIDOSIS

The clonal plasma cell burden in AL is usually small and similar to that seen in patients with monoclonal gammopathy. Because the proliferation rate of plasma cells is very low, chromosomal aberrations need to be assessed by FISH analysis and not by conventional metaphase cytogenetics. Approximately 70 percent of patients with AL have FISH abnormalities, the most common being IgH translocations (48 percent), including t(11;14) and t(14;16). Other chromosomal abnormalities seen in AL include deletion 13/13q− and hyperdiploidy.¹⁰⁴ The t(11;14) occurs more frequently in AL (39 percent) than in monoclonal gammopathy or myeloma.¹⁰⁵ Although cases with t(4;14) abnormality have been reported, those are exceptional and deletion 17p13 is not seen in this form of amyloidosis. FISH analysis is important in this disease, because t(11;14) is associated with an inferior prognosis in amyloidosis in contrast to myeloma, where it is associated with a good prognosis. In another large study assessing the prognostic significance of cytogenetic abnormalities, t(11;14) was not associated with an inferior outcome, but gain of chromosome 1q21 clearly was.¹⁰⁶

GENETIC ABNORMALITIES IN WALDENSTRÖM MACROGLOBULINEMIA

Alterations in one of myeloid differentiation primary response gene MYD88 have been identified in 90 percent of patients with macroglobulinemia.¹⁰⁷ MYD88 is an adaptor protein important in TLR and IL-1 receptor (IL-1R) signaling pathways. MYD88 is recruited to the activated receptor complex as a homodimer that complexes with the IL-1R–associated kinase (IRAK) 4 to activate IRAK 1. It leads to NF-κB activation via IκB α phosphorylation. Cell survival is enhanced by MYD88 overexpression, which leads to Bruton tyrosine kinase (BTK) phosphorylation. Combined use of IRAK and BTK inhibitors results in synergistic killing of MYD88-expressing macroglobulinemia cells. High rates of response have been observed in a clinical trial with a BTK inhibitor in relapsed and refractory patients.¹⁰⁸

The marrow microenvironment, discussed in detail in Chap. 5, provides a highly supportive tumor microenvironment (TME) for the development, growth, and survival of neoplastic cells in patients with myeloma. In the great majority of cases, the clonal expansion of these cells in the marrow is associated with increased blood vessel formation (neoangiogenesis) and, more importantly, myeloma bone disease (MBD).

Nonmalignant stromal cells in the marrow secrete cytokines and chemokines that promote myeloma cell growth and survival upon binding to specific receptors on the myeloma cell surface. Tumor promoters include IL-6, insulin-like growth factor 1 (IGF-1), VEGF, B-cell activating factor (BAFF), fibroblast growth factors (FGFs), stroma cell-derived factor 1α (SDF-1α, a.k.a. C-X-C motif chemokine 12 or CXCL12), and tumor necrosis factor α (TNF-α). Direct physical interaction of myeloma and marrow stromal cells by virtue of cell-to-cell adhesion may further enhance the cellular signaling pathways that are activated by cytokines and chemokines, thereby facilitating migration of myeloma cells to distant marrow and/or extramedullary sites (tumor dissemination). Myeloma-to-bone mesenchymal stem cell adhesion is also involved in the acquisition of drug resistance by tumor cells, underlying tumor relapse and/or refractory disease in patients with myeloma.¹⁰⁹

HOMING AND ADHESION OF MYELOMA CELLS

Homing and adhesion of myeloma cells to the marrow microenvironment involves the CXCL12/CXCR4 pathway and a number of homo- or heterotypic adhesion factors, including CD44 (an anionic, nonsulfated glycosaminoglycan called hyaluronan), very-late antigen 4 (VLA-4, composed of integrins α₄ [CD49d] and β₁ [CD29]) and its receptor, vascular cell adhesion molecule 1 (VCAM-1, CD106), leukocyte function–associated antigen 1 (LFA-1, CD11a), neuronal cell–adhesion molecule (NCAM, CD56), intercellular adhesion molecule 1 (ICAM-1, CD54), and, importantly, syndecan 1 (CD138). Syndecan 1 is a transmembrane heparan sulfate–containing proteoglycan that is usually expressed at high levels on the myeloma cell surface. Syndecan 1-mediated adhesion of myeloma cells promotes adhesion-dependent drug resistance and bone resorption via expression of matrix metalloproteinases, among other mechanisms.

Adhesion of myeloma cells to mesenchymal cells activates pleiotropic cellular signal transduction pathways that mediate the proliferative and survival-enhancing response of myeloma cells upon interaction with the marrow microenvironment. Resistance of myeloma cells to cytotoxic drugs is also promoted. These pathways include NF-κB (nuclear factor kappa light-chain enhancer of activated B cells); PI3K/AKT (phosphatidylinositol 3-kinase/protein kinase B/AKT oncogene), RAS/RAF/MEK/ERK (rat sarcoma protein subfamily of small GTPase/RAF protooncogene serine and threonine protein kinase/MAPK kinase/extracellular signal–regulated kinase) and JAK2/STAT3 (Janus kinase 2/signal transducers and activators of transcription 3). Engagement of these pathways results in upregulation of proteins that control cell cycle progression (e.g., D cyclins and Myc) and those that protect myeloma cells from programmed apoptotic cell death (e.g., BCL2, BCL-x_L, and MCL1).

The interplay of myeloma and mesenchymal cells is not the same in all subtypes of the disease. Instead, the genetic makeup of myeloma determines, in part, the interaction of tumor with the nonmalignant stromal cells in the marrow environment. For instance, myeloma cells harboring a MAF (v-MAF oncogene homologue)-activating chromosomal translocation, t(14;16), overexpress the adhesion factor, integrin β₇. This results in enhanced tumor cell adhesion to bone mesenchymal stem cells (BMSCs) relative to myeloma cells not containing a t(14;16).¹¹⁰ Signaling pathways that govern interactions of myeloma and normal marrow cells also promote neoangiogenesis in the TME. The formation of new blood vessels, which is driven to a large extent by CXCL12/CXCR4 signaling and production of VEGF, B-FGFs, matrix metallopeptidases (MMPs), IGF-1, interleukins (e.g., IL-8, IL-1), angiopoietin 1, transforming growth factor β (TGF-β), platelet-derived growth factor (PDGF), and hepatocyte growth factor (HGF), is critical for myeloma progression and thus an important target of myeloma drug development.

Increased bone resorption and suppressed bone formation leads to osteolytic lesions in patients with myeloma.^111,112 The main cytokines involved in that process are IL-6, receptor activator of NF-κB ligand (RANKL)/osteoprotegerin (OPG), BAFF, chemokine (C-C motif) ligand 3 (CCL3)/macrophage inflammatory protein (MIP)-1α, and VEGF. The main cell adhesion and integrin signaling pathways involved in that process are VLA-4/VCAM-1 and LFA-1/ICAM-1.^112,113 WNT/β-catenin signaling antagonists that inhibit osteoblast differentiation in myeloma also play an important role in the natural history of MBD. These include Dickkopf 1 (DKK1),¹¹⁴ soluble-frizzled receptor-like proteins (sFRPs), and sclerostin. Inhibition of osteoblast differentiation drives the bone disease and, presumably, facilitates the expansion of myeloma cells¹¹¹ because upregulation of β-catenin–dependent osteoblast activity in vivo results in significant inhibition of myeloma cell growth,¹¹⁶ relying on an ill-defined mechanism that appears to involve the small leucine-rich proteoglycan, decorin.¹¹⁷

RANK/RANKL/OPG PATHWAY

The receptor activator of NF-κB (RANK)/RANKL/OPG pathway is arguably one of the most important regulatory systems for homeostatic bone remodeling. RANK, a membrane-anchored signaling receptor of the TNF receptor family, is expressed on osteoclast (OC) precursors. Binding of its ligand, RANKL, expressed by osteoblasts (OBs) and BMSCs, induces osteoclastogenesis and activation of mature OCs. OPG, a soluble member of the TNF receptor family, is also produced by OBs and BMSCs. Because OPG acts as a soluble decoy receptor for RANKL, it blocks RANK-dependent OC maturation and activation. Thus, it is easy to understand that the ratio of RANKL and OPG significantly impacts the balance of bone resorption and bone deposition. Indeed, compared to normal individuals, marrow plasma of myeloma patients contains elevated levels of RANKL and reduced levels of OPG,¹¹⁷ consistent with a proosteoclastogenic TME. Also, low serum levels of OPG correlate with advanced MBD in patients with myeloma.¹¹⁸ In mouse models of MBD, treatment with OPG and OPG-like compounds prevented both bone destruction and myeloma growth in vivo.^119,120 Similarly, in patients with myeloma treated with thalidomide¹²¹ or autologous stem cell transplantation¹²² responding to therapy, the RANKL-to-OPG ratio normalized and bone resorption was inhibited. In aggregate, these findings indicate the RANK-RANKL-OPG axis is an important therapeutic target for MBD.

B-CELL ACTIVATION FACTOR

BAFF is a myeloma cell-promoting member of the TNF superfamily of proteins that is expressed by OCs and BMSCs. BAFF is significantly increased in the serum of patients with myeloma. In a preclinical study using a mouse model of myeloma, targeting BAFF by means of a neutralizing antibody led to reductions in tumor burden and osteolytic bone disease and increased survival.

DKK1 is an inhibitor of the WNT signaling pathway that plays a major role in MBD.¹¹⁴ DKK1 inhibits OB differentiation and promotes OC maturation.¹²³ Patients responsive to myeloma drugs exhibit reduced serum levels of DKK1.

Activin A, a member of the TGF-β superfamily of proteins, activates a signaling pathway that results in phosphorylation of SMADs (homologues of Drosophila’s mothers against decapentaplegic [MAD] and Caeno Rhabditis elegans’ small body size [SMA] proteins) followed by execution of a complex SMAD-dependent gene-expression program. Activin A is a stimulator of osteoclastogenesis and is overexpressed in myeloma.¹²⁴ IL-3, another stimulator of OCs, can induce the secretion of activin A by resident macrophages in the myeloma marrow microenvironment, providing a mechanism for the upregulation of activin A in patients with myeloma.¹¹² Levels of activin A in the marrow plasma of myeloma patients with osteolytic lesions (OLs) are increased.¹²⁵ RAP-011 is a soluble mouse activin A receptor that was shown to inhibit myeloma-like tumors and prevent OLs in laboratory mice.^124,126 Sotatercept (ACE-011), the human version of mouse RAP-011 is a soluble receptor of activin A and is undergoing clinical trials in patients with myeloma.

BRUTON TYROSINE KINASE

The metalloprotein, BTK, is a member of the Scr-related Tec family of protein kinases. BTK is a regulator of OC differentiation that is expressed in OCs (but not in OBs) and has been implicated in bone resorption. BTK is expressed in myeloma cells.^127,128 Genetic evidence indicates BTK may be crucial for neoplastic plasma cell development in laboratory mice.¹²⁹

Major progress has been made in the treatment of newly diagnosed myeloma patients with 80 percent achieving either a complete remission (CR) or near CR (nCR) and more than 50 percent surviving 10 years.¹³⁰ However, many patients ultimately relapse. Gene-expression profiles (GEPs) remain abnormal in myeloma patients with long-lasting remission (>10 years).¹³¹ This finding suggests that the persistence of a myeloma cell population with low proliferative capacity and limited sensitivity to our most intensive therapies, points to the concept of a myeloma stem cell problem. There is no agreement on the specific phenotype of the myeloma stem cell, but side-population (SP) cells, which constitute a minor fraction of the myeloma cells, have the features of stem cells. Such cells have a greater clonogenic potential and drug resistance, whether isolated from myeloma cell lines or primary myeloma samples.^132,133 Many different groups have tried to identity the phenotype of myeloma stem cells. A drug-resistant subpopulation of memory B-cell–like cells with the CD138−/CD19+/CD27+ phenotype was identified and termed myeloma stem cells.^134,135 The CD138− cell population derived from myeloma cell lines contains significantly higher levels of ALDH, a marker for stem cells. In contrast to the CD138+ cells, the CD138− cells were not affected by the drugs we typically use in myeloma, such as lenalidomide, bortezomib, dexamethasone, and cyclophosphamide.¹³⁴ In an earlier study, human CD34+/CD45_low clonotypic myeloma cells were shown to cause lytic bone lesions in a xenograft mouse model, and these CD34+ progenitors included approximately one-third of DNA aneuploid cells.^135,136 Mature CD138+ myeloma cell dedifferentiation to myeloma stem cells with a CD34+/CD138+/B7−/H1+ phenotype has been reported,¹³⁷ showing that myeloma cells have the plasticity to replenish the stem cell compartment if required, and that we should not think of stem cells as a static, but as a dynamic concept. CD38++/CD45− plasma cells proliferated successfully within an engrafted human fetal bone using the severe combined immune deficiency (SCID)-hu mouse model.^138,139 This could have been because of dedifferentiation of these more mature myeloma cells. It also has been demonstrated that the SP cells from different myeloma cell lines are able to generate more colonies than mature plasma cells. However, this SP cell population lacked correlation with CD138 expression.^132,133 To isolate multiple myeloma stem cells (MMSCs) from primary samples, we have relied on a functional characteristic of cancer stem cells, the ability to pump out the Hoechst dye, rather than on membrane markers, which are more likely to change with changing environmental conditions. To this functional marker, we have added a myeloma specific marker, either surface κ or λ light chain, based on the specific M-protein. Because drug-resistant myeloma cells are enriched in patients relapsing early after tandem autologous transplants, we compared GEPs at baseline and at relapse in 51 patients. The gene most differentially expressed at relapse was the retinoic receptor α (RARα). This receptor has two splice variants: α₁ and α₂. RARα₁ is present on all myeloma cells; RARα₂ expression is only present in one-third of newly diagnosed patients. The latter patients have a significantly inferior survival after tandem autotransplants.¹⁴⁰ In a subsequent study, we were able to show that increased RARα₂ expression conferred stem cell features to myeloma cells, such as increased drug resistance, increased clonogenic potential, activation of pathways typically found in a cancer stem cell (CSC) such as the Wnt and Hedgehog pathways, increased SP and ALDH levels, and increased expression of embryonic stem cell genes, such as Nanog, Oct4, and Sox2. These same characteristics were found in the SP+ and surface light-chain–restricted cells of primary myeloma samples. We also found that RARα₂ expressing CD138+ myeloma cells have a much higher expression of Oct4, Sox2, Nanog, TCF1, CCND1, and ABCC3. This indicates that RARα₂-positive myeloma cells show stem cell features and, therefore, have a worse outcome.¹⁴¹

TESTS OF SERUM

Because PCNs are B-cell malignancies derived from antibody-producing plasma cells in the marrow, which continue to secrete in the vast majority, an M-protein will be found in serum and/or urine in the majority of patients. The monoclonal protein is either an intact immunoglobulin or a component of the immunoglobulin. The most common types in myeloma are IgG and IgA; in Waldenström macroglobulinemia, it is IgM. The most common screening test to identify a serum M-protein is the serum protein electrophoresis (SPEP). While the quantity of M-protein is often considered a marker of tumor load, myeloma cells are ineffective producers of immunoglobulins and produce 10 to 100 times less immunoglobulin per cell per day than a normal plasma cell. In addition, the more immature and more proliferative myeloma cells are, the less immunoglobulin per cell per day they secrete. In myeloma, there is an imbalance of heavy chains and light chains in favor of the light chains. In general, the more aggressive the myeloma, the more pronounced the imbalance in favor of excess light-chain secretion. The excess light chains cannot bind to heavy chains and can pass freely though the renal glomeruli. Monomeric FLC, characteristic κ, are cleared from the serum in 2 to 4 hours at 40 percent of the glomerular filtration rate (GFR), while dimeric FLC, typically , are cleared in 3 to 6 hours at 20 percent of GFR. Removal may take 2 to 3 days in patients with complete renal failure. This is in contrast to IgG, which has a plasma half-life of 21 days.

Although it has long been possible to measure the total amount (free plus bound to heavy chains) of light chains in the urine, our ability to measure serum FLCs became a reality early in the 21st century. This test is not only of high affinity and allows measurement of low concentrations of FLC, but it is also of high specificity and only measures unbound light chains. Serum FLC are several orders of magnitude lower than serum light chains bound to intact immunoglobulin and the test is based on an antibody that exclusively binds to an epitope of the FLC that is hidden if the light-chain molecule is bound to a heavy chain. Because the half-life of serum FLCs is much shorter than that of the intact immunoglobulin, serial serum FLCs allow assessment of success of therapy much earlier than serial M-protein levels. Patients with highest amount of light chains have the worst outcome.¹⁴² The serum FLC ratio is an independent prognostic factor for progression of myeloma in monoclonal gammopathy and SMM.¹⁴³ Incorporation of serum FLCs into a staging system also improves risk stratification for patients with AL¹⁴⁴ and levels of serum FLCs correlate with tumor burden in macroglobulinemia.¹⁴⁵ The serum FLC assay is also an early indicator of relapse in myeloma. As explained above, both serum κ and concentrations increase with deteriorating renal function, but the ratio only changes slightly; with increasing renal failure the κ: ratio gradually increases. Diseases associated with generalized increased B-cell activation frequently have high concentrations of polyclonal immunoglobulins and high polyclonal FLC, but the κ: ratios will remain within normal range.

In addition to a SPEP and serum FLCs, it is important perform a serum immunofixation electrophoresis (sIFE) to identify the nature of the M-protein, and to measure the quantitative immunoglobulins. Many of the IgA and some of the IgG myelomas have their M-spike in the β-globulin region, where the M-protein level is less reliable because there is comigration of many other proteins. If an M-protein is identified on SPEP and sIFE does not identify the heavy chain, quantitative analysis of IgD and IgE should be performed. Most IgD myelomas are associated with lambda light chains.

TESTS OF URINE

A typical myeloma workup will have a 24-hour urine collection for total protein, urine protein electrophoresis (UPEP), urine M-protein, and urine immunofixation electrophoresis (uIFE). In contrast to the serum FLCs, the role of urine FLCs is much less clear and is generally not used. In the era of serum FLC assays, some experts believe that urine analysis can be eliminated. In a study by the Mayo Clinic of 428 patients, there were only two cases where a urinary monoclonal protein was missed by omitting urine analysis (0.5 percent). These two cases did not require any medical intervention.¹⁴⁶ When patients have a large amount of albuminuria in the presence of PCN with or without urine M-protein, the possibility of AL should be entertained, either primary AL or AL in the context of myeloma.

MARROW EXAMINATION

Marrow aspirate and biopsy are essential in the workup of PCN. It is best not to rely solely on a marrow aspirate. Myeloma cells have the tendency to form small or large clusters and myeloma is not evenly distributed throughout the marrow. The marrow may be falsely negative in the presence of overt myeloma. In a normal marrow with normal cellularity, there may be up to 2 percent plasma cells present. However, the percentage of plasma cells can be much higher in reactive marrows and hypoplastic or aplastic marrows in the absence of PCN. In a normal marrow, there can be some clustering of plasma cells around the blood vessels. However, the presence of larger clusters of plasma cells away from the blood vessels should alert one to the possibility of PCN, especially myeloma. In addition to the morphologic examination, the biopsy slides should be stained with a CD138 monoclonal antibody, which is a specific immune marker for plasma cells. If the diagnosis of PCN is not clear based on morphology and CD138 staining, in situ hybridization for cytoplasmic κ and λ should be used to evaluate whether the plasma cells are clonal (light-chain restricted).

It is customary to perform flow cytometry, metaphase cytogenetics, and FISH analysis on myeloma samples. Flow cytometry at diagnosis is mainly performed to identify aberrant surface markers on myeloma cells, which can be used subsequently to assess MRD. The identification of an accurate gating strategy is a critical component of a reproducible and sensitive immunophenotypic analysis. The best gating strategy uses a combination of CD38, CD138, CD45, and light scatter characteristics. The previously used gating method of CD38 and CD45 decreases the risk of contamination with other cells, but excludes the CD45+ cells, which can constitute the majority of the plasma cells. It is recommended that at least a four-color instrument be used, as at least two antigens (CD38 and CD138) are required after the initial analysis with CD138, CD38, CD45, and light scatter, to gate plasma cells accurately. When used for MRD, at least 100 neoplastic plasma cell events should be acquired. CD56 and CD19 should always be assessed. CD56 is negative on normal plasma cells but is positive on myeloma cells, while the opposite is true for CD19. It is also recommended to stain for CD117 and CD20, which are negative on normal plasma cells and can be aberrantly expressed on myeloma cells. CD28 and CD200 are negative or only weakly expressed on normal plasma cells, but can be strongly positive on myeloma cells, whereas CD27 and CD81 are strongly expressed on normal plasma cells, but weak or negative on myeloma cells. It is important to remember that there is no single marker that can systematically differentiate neoplastic cells from normal plasma cells. Also, the percentage of plasma cells detected by flow cytometry is typically lower than that found by morphology.¹⁴⁷ The importance of metaphase cytogenetics and FISH analysis was outlined above in the section Techniques to assess cytogenetic information. However, metaphase cytogenetic analysis remains important and is the best marker to differentiate between stroma-dependent (early) and stroma-independent (advanced) myeloma. The metaphases with normal cytogenetics obtained in most patients with myeloma are derived from the remaining normal hematopoietic elements and not from the myeloma cells. If myeloma cells are stroma-dependent, they will die soon after being removed from their supporting microenvironment, while stroma-independent myeloma cells are able to grow and divide in the absence of a supporting microenvironment.¹⁴⁸

IMAGING STUDIES

Bone disease is present in 70 percent of patients with myeloma at diagnosis. A skeletal X-ray survey has long been considered the gold standard for assessing bone disease in myeloma. However, the sensitivity and specificity of a skeletal survey is low. At least 50 to 75 percent of trabecular bone must be lost to see a lytic lesion. The false-negative rate varies between 30 and 70 percent, which leads to significant underestimation in diagnosing and staging of patients with myeloma. The skeletal survey does not provide us with a direct image of the tumor cells, but rather with the consequence of tumor cells being present or having been present. Bone lesions in myeloma seldom heal and in most cases, there is no healing at all or at the most there is a sclerotic rim around the bone lesion. Therefore, a skeletal survey is not a good technique to assess response to treatment or to diagnose early recurrence of myeloma. Much better imaging techniques have become available in the form of ¹⁸F-fluorodeoxyglucose positron emission tomography–computed tomography (PET/CT) and MRI. When used in combination, PET/CT scan and MRI were found to have specificity and a positive predictive value of virtually 100 percent, which is invaluable to clinicians assessing the efficacy of intensive and expensive treatment approaches with the goal to cure myeloma.¹⁴⁹

Magnetic Resonance Imaging

Between CT and MRI the latter is the more specific test. A complete MRI examination should include a series of sequences to permit identification of focal and diffuse marrow involvement, including spin echo (T1 and T2 weighted), gradient echo (T2), and short T1 inversion recovery (STIR) sequences. MRI should include the axial bones with active hematopoiesis in adults (skull, spine, sternum, shoulders and upper humeri, pelvis, and upper femora). In a study from the University of Arkansas for Medical Sciences, using these parameters and including 611 myeloma patients treated uniformly with tandem transplants, 74 percent had focal myeloma lesions compared with 56 percent on skeletal survey and 52 percent of patients with negative skeletal surveys had focal lesions on the MRI. Resolution of the MRI focal lesions after treatment occurred slowly, but ultimately was seen in 60 percent of patients and conferred a superior overall survival.¹⁵⁰ MRI should be used routinely for the staging, prognosis and response assessment of patients treated with curative intent.

¹⁸F-Fluorodeoxyglucose Positron Emission Tomography–Computed Tomography

One of the major advantages of the PET/CT scan is that it allows assessing the presence of extramedullary myeloma, which is present in 6 percent of the patients and associated with a significantly inferior overall survival based on two large studies. In a study of 239 patients receiving uniform therapy, the presence of more than three fluorodeoxyglucose (FDG)-avid focal lesions was associated with a significantly inferior overall and event-free survival. In contrast, complete FDG suppression prior to the first transplant conferred a significantly better outcome.¹⁵¹ These data were confirmed in an Italian study including 192 patients.¹⁵² A PET-CR occurs earlier than a clinical CR and a MRI-CR happens much later and less frequently.

OTHER IMPORTANT TESTS

β₂-Microglobulin

The serum β2-microglobulin (β2m) is one of the most important prognostic factors in myeloma; β2m is a small protein that associates with human leukocyte antigen class I and is almost exclusively catabolized in the kidneys. Its best-characterized function is to interact with and stabilize the tertiary structure of major histocompatibility complexn (MCH) class α chain. The main source of β2m in the serum is membrane turnover. It reflects tumor load and renal function; however, it predicts survival irrespective of renal function and Durie-Salmon stage.¹⁵³ The exact underlying biologic significance of β2m remains obscure, but interactions with drug resistance pathways may exist. β2m in high concentrations retards the generation of monocyte-derived dendritic cells.¹⁵⁴ Serum β2m and albumin levels form the basis for the International Staging System, which has major prognostic significance.⁹⁹

High-risk myeloma, defined by genetic abnormalities, is independent of the International Staging System and the latter had only significant prognostic value in patients with a low-risk gene expression profile and normal metaphase cytogenetics.¹⁵⁵

Serum Lactic Dehydrogenase Levels

The significance of high serum lactate dehydrogenase (LDH) levels in the absence of any other cause, such as liver disease or hemolytic anemia, predicting poor prognosis has long been recognized in myeloma.¹⁵⁶ Although rarely observed in the early phase of the disease, marked elevations of LDH were detected in up to 20 percent of patients with disease progressing after vincristine, Adriamycin, and dexamethasone chemotherapy. High LDH levels were associated with hypercalcemia, elevated serum β2m levels, extramedullary manifestations, plasmablastic morphology, and short overall survival duration, despite marked (usually very transient) antitumor responses to high-dose therapy. The same poor prognosis was noted in patients with initially normal LDH levels in who marked LDH increments were induced by treatment, presumably resulting from tumor lysis syndrome as an indicator of rapidly proliferative myeloma.¹⁵⁶ Close correlations with plasma cell labeling index and plasmablastic morphology has been demonstrated.¹⁵⁷ Elevated serum LDH levels at diagnosis also emerged as a significant prognostic factor in an analysis of 155 newly diagnosed patients who received at least one course of high-dose therapy with autologous stem cell transplantation (ASCT), irrespective of the deletion of chromosome 13.¹⁵⁸

MINIMAL RESIDUAL DISEASE

Methods to assess MRD include allele-specific oligonucleotide PCR (ASO-PCR), multiparameter flow cytometry (MPF), fluorescent-PCR (F-PCR) and high-throughput sequencing-based MRD assessment.

Allele-Specific Oligonucleotide Polymerase Chain Reaction

ASO-PCR can detect one clonal cell in 10⁵ normal cells. It remains a costly and labor-intensive assay to perform because a specific probe needs to be generated for each patient and it is unsuccessful in approximately 30 percent of patients.¹⁵⁹ In a relatively small study of 40 patients who achieved either a CR or very good partial response after autologous transplantation and who received consolidation with bortezomib, thalidomide, and dexamethasone (VTD), eight patients achieved a molecular remission: two only in one sample and six had at least two consecutive negative samples. With a median followup of 26 months from study entry, no clinical relapses were seen in patients achieving a molecular remission, although one molecular relapse was seen, while eight clinical relapses occurred in patients not a achieving a molecular remission. Importantly, VTD consolidation decreased the tumor load further after transplantation.¹⁶¹ In contrast to earlier held opinions that molecular remissions could only be obtained with allotransplantation, this study showed that molecular remission were also seen after autologous transplants followed by consolidation therapy with VTD.

Multiparameter Flow Cytometry

Multiparameter flow cytometry (MPC) has the advantage of being readily available and short turn-around time. It requires sophisticated analysis, but is automated. This technique can detect one clonal cell in 10⁴ normal cells. In the MRC Myeloma 9 study, MRD was assessed by MPF in 397 patients who received an autotransplant and 245 patients who were treated with a nontransplantation approach. A six-color panel was applied. For pretreatment samples 100,000 events were acquired, while for the posttreatment samples a minimum of 500,000 events were analyzed. In the transplantation group, absence of MRD at day 100 after transplantation, observed in 62 percent of patients, was highly predictive of a favorable outcome (progression-free survival [PFS]: p <0.001; overall survival [OS]: p = 0.018). This outcome advantage was seen irrespective of cytogenetic findings, but more clearly in patients with adverse cytogenetics (p <0.001 vs. p = 0.014). There was no complete agreement between immunofixation electrophoresis (IFE)-negative CR and absence of MRD by MPF. Approximately 15 percent of patients in CR still had measurable disease by MPF, while 25 percent of MRD-negative patients failed to achieve a CR by IFE. The effect of thalidomide maintenance therapy after transplantation was also assessed. The best outcomes were seen in patients who achieved MRD negativity and were maintained on thalidomide, while the outcome was worst in patients who failed to achieve MRD negativity and did not receive thalidomide maintenance. In the MRD-positive group at day 100, 28 percent of those receiving thalidomide became MRD-negative, while in those not receiving thalidomide maintenance only 3 percent became MRD-negative (p = 0.025). Finally, MRD assessment at the end of induction therapy in the nontransplantation group had no predictive value; only 15 percent of such patients became MRD-negative.¹⁶¹ It should be noted that there is a large heterogeneity in assessing MRD by MPF in the United States in terms of events acquired and number of antibodies used.¹⁶² If outcomes of therapies are based on MPF, it is critical that this technique becomes standardized and suitable quality controls are in place.

Fluorescent Polymerase Chain Reaction

F-PCR can detect one clonal cell in 10³ normal cells and is thus less sensitive than ASO-PCR and MPF. However, F-PCR is rapid, affordable, and easy to perform. High-molecular-weight DNA is isolated from 500 μL of marrow. Three different multiplex PCRs are used: IGH D-J, IGK V-J, and KDE rearrangements. The clonal population is identified at diagnosis. Patient with a visible lack of a clonal peak identified at diagnosis were considered F-PCR–negative. MRD was assessed in 130 newly diagnosed myeloma patients. The test was informative in 91.5 percent of patients. MPC was used in parallel with F-PCR. After induction, 64 patients achieved a molecular response and 66 did not. Median PFS was 61 versus 36 months (p = 0.001). The corresponding PFS with MPC was 67 versus 42 months for MPC− and MPC+ patients (p = 0.005).¹⁶³

High-Throughput Sequencing-Based Minimal Residual Disease Assessment

High-throughput sequencing-based MRD assessment relies on amplification and sequencing of immunoglobulin gene segments using consensus primers. It employs the IGH-VDJ_H, IGH-DJ_H, and IGK assays. The assay can detect one clonal cell in 10⁶ normal cells and is 1 log more sensitive than ASO-PCR. The prognostic value of this test was assessed in 133 myeloma patients enrolled in the GEM myeloma trials and was compared to MPC and ASO-PCR. The test was informative in 91 percent of patients. Concordance between deep sequencing and MPC or ASO-PCR was 83 percent and 85 percent, respectively. Patients who were MRD− by deep sequencing had a significantly longer time to progression (80 vs. 31 months; p <0.0001). In CR patients, the time to progression remained significantly longer in MRD− patients (131 vs. 35 months; p = 0.0009).⁹⁵

There is no doubt that assessment of MRD will receive more attention in the coming years to guide the clinicians in their treatment decisions and to individualize patient care. However, it should not be automatically concluded that MRD− obtained after continuation of aggressive therapy to ultimately reach MRD− status which was not achieved after a standard aggressive approach will have the same prognostic significance as MRD− obtained with standard intensive approaches. It will also be important to establish what degree of tumor reduction has the highest prognostic significance. Extremely sensitive techniques might not necessarily provide better prognostic information than somewhat less-sensitive techniques.