a Karyogram of a cell from a patient with hyperdiploid plasma cell myeloma. Chromosomal gains, and loss of one X chromosome, are indicated by arrows. b Chromosome pairs showing the primary IGH translocations in plasma cell neoplasms (the t(4;14) is cytogenetically cryptic and cannot be seen by G-banding). Translocations from top to bottom (chromosome 14 is on right in each pair): t(6;14); t(11;14); t(14;16); t(14;20)
Molecular Cytogenetic Analysis (FISH)
Conventional G-banding analysis is a well-established way to screen the genome for disease-associated abnormalities . However, conventional cytogenetic analyses in plasma cell neoplasms are hampered (a) by low resolution (each G-band might have 5–10 Mb of DNA), (b) by the fact that terminally differentiated plasma cells are often not mitotically active and thus normal bone marrow elements may be analyzed, and (c) because some abnormalities are cryptic (i.e., undetectable by G-banding due to their small size and/or banding characteristics). As a result, over 50 % of plasma cell neoplasms at diagnosis yield a normal result by G-banding. However, fluorescence in situ hybridization (FISH), which can be performed on interphase cells, obviates the need for dividing cells. Although FISH can be performed on metaphase cells, in neoplastic disorders, it is most frequently performed on interphase (nondividing) cells. A large number of interphase cells (typically at least 200) can be rapidly evaluated, thus making it much faster and more sensitive than G-banding analysis. Sensitivity can be increased further by enriching for specific cell populations (described below).
Excellent references are available describing protocols for evaluating specimens by FISH [2, 4, 6, 7], a method adaptable to a variety of substrates including cell suspensions, touch imprints, disaggregated tissues such as tumors or trephine core biopsies, and paraffin-embedded tissues (of note, however, FISH cannot be readily performed on specimens, such as bone marrow trephine cores, that have undergone decalcification). Critical steps in this process include applying fluorescently labeled probes (DNA sequences typically several hundred kb long and complementary to known genomic sequences) to cells that have been dropped onto a glass slide . The DNA of the probe and specimen are concurrently heat denatured for several minutes, after which the slide is kept at 37 °C for 6–14 h to allow the probe to bind to its target sequence. After washing and the addition of a nuclear counterstain such as 4′6′-diamidino-2-phenylindole (DAPI), the cells are examined under a fluorescence microscope. Because a number of different probe types are commercially available (or can be developed within the laboratory) (Table 1), it is critical that the technologist know the specific probe type used, the locus being evaluated, and the expected signal pattern. He or she must also be vigilant for unexpected signal patterns (e.g., gain rather than rearrangement of a locus) .
Enumeration: Directed against pericentromeric region of chromosome (Fig. 2a)
Gain or loss of specific chromosomes
Large, bright signals; easy to read
Repetitive sequences may result in cross-hybridization to other chromosomes
Locus specific: Directed against specific genes or loci (Fig. 2b)
Gain or loss of specific loci
Flexible; can also be used as markers for gain/loss of chromosomal regions, not just loci targeted
Partial gene deletions may be difficult to detect
Dual-fusion: Two genes are labeled in different colors (Fig. 2c)
Gene fusion in translocations
High specificity (low false positivity)
Variant translocations (including deletions at breakpoints) may be difficult to detect
Breakapart: 5′ and 3′ ends of gene labeled in different colors (Fig. 2d)
Rearrangement of genes with multiple partners
Easier to interpret than dual-fusion probes in some settings (e.g., paraffin-embedded tissue)
Partner gene cannot be identified/confirmed
Examples of different FISH probe configurations. a Probes to the centromeric regions of chromosomes 9 (aqua) and 15 (red), and to the D5S23/D5S721 loci (5p15.2) (green) (Abbott Molecular, Abbott Park, IL, USA). The interphase cells show 2 or 3 aqua, 3 red, and 3 green signals, indicating trisomies for chromosomes 5, 9, and 15 b Locus-specific probes to the ATM (green) and TP53 (red) loci (Cytocell, Cambridge, UK). The interphase cell shows two green and only one red signal, indicating loss of the TP53 locus. c Dual-fusion probes to the IGH (green) and CCND1 (red) loci (Abbott Molecular). The interphase cells show one normal green signal, one normal red signal, and two small red/green fusion signals (arrows), which appear yellow by fluorescence microscopy. d Breakapart probe to the IGH locus (Abbott Molecular). The interphase cells show one red/green (yellow) fusion signal, representing an intact (nonrearranged) IGH locus. One smaller red and green signal are also present, representing splitting of the IGH locus into its 3′ (centromeric, red) and 5′ (telomeric, green) components
The greater sensitivity of FISH compared with G-banding can be increased even further by enriching for specific populations of cells. This is particularly important in plasma cell neoplasms, in which it is well established that the rate of detection of abnormalities and the number of abnormalities detected per case increases dramatically when a plasma cell enrichment procedure is used [8−14]. Plasma cell enrichment can be accomplished by a number of different methods, including anti-CD138 magnetic bead sorting, concurrent staining of cells for cytoplasmic immunoglobulin, cell sorting by flow cytometric immunophenotyping, and morphologic selection of plasma cells [8, 9, 11−13, 15−18]. Because of the high yield of abnormalities found using these methods, the International Myeloma Working Group now recommends that FISH in plasma cell neoplasms be performed only on enriched specimens . As these methods may require the processing of uncultured bone marrow, it is good practice to send sufficient bone marrow (ideally at least 2 ml) to the cytogenetics laboratory at the time the specimen is obtained .
Because conventional and molecular cytogenetic analyses provide complementary information about disease status, both are typically performed at diagnosis. Although FISH is often performed in hematologic disorders to measure response to therapy, in myeloma the plasma cell enrichment process precludes its use as a quantitative assessment of therapeutic response. However, the markedly increased sensitivity obtained by the enrichment process enables the detection of very low levels of residual disease that may be undetectable by morphologic or flow cytometric analysis (depending on the flow cytometry laboratory’s methodology, number of events analyzed and limit of detection) .
Oligonucleotide Comparative Genomic Hybridization (CGH) and SNP (Single Nucleotide Polymorphism) Arrays
G-banding and FISH have been routinely used in the evaluation of plasma cell neoplasms for decades, and they remain a critical part of the evaluation and monitoring of these diseases . Array-based genomic technologies are now also becoming more widely employed. These assays, which use extracted DNA rather than metaphase or interphase cells as the substrate, can detect copy number gains or losses with a markedly increased resolution (< 100 kb compared with 5–10 Mb for G-banding) . Like FISH , these arrays evaluate specimen DNA via the use of fluorescently labeled probes. Unlike FISH, however, in which a single probe set is applied to patient DNA affixed to a slide, in array-based testing there are typically > 100,000 probes affixed to a slide, to which specimen DNA is applied. Oligonucleotide CGH identifies regions of copy number gain or loss; SNP arrays, which evaluate the specimen at highly polymorphic regions throughout the genome, can also identify copy number neutral loss of heterozygosity, in which two copies of a chromosomal region are present, but derived from one parent [20−22]. These array-based techniques more precisely determine the size of these aberrant regions as well as their gene content . As with FISH, enrichment for plasma cells improves detection of abnormalities [9, 14, 18, 20] .
Gene Expression Arrays
Gene expression studies have also been used in the risk assessment  and identification of subgroups of patients [25−27], as well as the elucidation of abnormalities acquired during disease progression from MGUS to myeloma . Panels of genes with prognostic significance, ranging from 15 to 70 genes, have been published [29, 30].
Next-generation sequencing (NGS) has also been used to elucidate the underlying genomic architecture of plasma cell neoplasms . NGS permits the detection of single-nucleotide polymorphisms (SNPs) or mutations in selected genes or, in the case of whole-exome sequencing, in the complete coding sequence of a genome. Array and sequencing techniques have been particularly important in identifying the complex intraclonal heterogeneity found in myeloma and its precursor lesions (MGUS and SMM) [32−34].
Thus, a number of different methods can be employed to evaluate the genomic abnormalities underlying plasma cell neoplasms. Some techniques, such as G-banding and FISH, have been routinely used by cytogenetics laboratories for decades. Others, such as next-generation sequencing, are highly complex assays not yet in routine clinical use, although the number of laboratories offering this testing is growing rapidly.
As early as the 1960s, before banding techniques had even been developed to permit identification of individual chromosomes, a number of reports of cytogenetic abnormalities in plasma cell disorders had been published. However, it was not until 1985 that the clinical significance of cytogenetic abnormalities in myeloma was established in a large study. Dewald et al.  reported that patients with cytogenetic abnormalities had a shorter survival than patients with normal karyotypes. In the decades since this study was published, great progress has been made in identifying the major cytogenetic pathways by which normal plasma cells become transformed and progress through MGUS and SMM to myeloma and even to plasma cell leukemia or extramedullary disease. These changes are summarized in Tables 2 and 3.
Primary IGH translocations
Standard risk in newer risk stratification systems 
Neutral prognostic factor; associated with lower immunoglobulin level and lymphoplasmacytic morphology; rare in hyperdiploidy 
Poor prognosis in myeloma but not in MGUS 
Genetic abnormalities in the development and progression of plasma cell neoplasms
0–9 % (pPCL)
~17 % (sPCL)
IGH translocation (partner gene)
t(14;20) (MAFB), t(6;14)
Other IGH translocations
20–75 %/ 30 %
Ploidy is one of the most basic means of classifying plasma cell neoplasms, which fall into two main categories: hyperdiploid (48–75 chromosomes) and nonhyperdiploid. The latter group includes those cases with translocations involving the immunoglobulin heavy chain (IGH) gene and one of five partner genes, as well as hypodiploid (< 48 chromosomes), near-diploid, pseudodiploid (46 chromosomes but with structural abnormalities), and near-tetraploid (> 75 chromosomes, which may represent doubling of a hypodiploid or near-haploid complement) cases. Because the hyperdiploid and nonhyperdiploid pathways are the two primary initiating genetic pathways in the development of plasma cell neoplasms, their associated abnormalities can be found at all stages of disease development, from MGUS to myeloma, as well as other manifestations of plasma cell neoplasms such as AL amyloidosis (Table 3).
Hyperdiploid plasma cell neoplasms are typically characterized by trisomies of two or more odd-numbered chromosomes (particularly, in order of decreasing frequency, chromosomes 15, 9, 5, 19, 3, 11, 7, 21) . Hyperdiploidy has been associated with better overall survival than nonhyperdiploid cases , which may be due to dosage effect of genes involved in tumor suppression or drug sensitivity . Patients with hyperdiploid myeloma who also have structural abnormalities typically associated with more aggressive disease such as IGH translocations and 1q gain have a worse prognosis [39, 40] .
Five IGH translocations in particular are considered primary abnormalities in the development of myeloma (see Tables 2 and 3) and, with the exception of the t(11;14) and t(6;14), are associated with an adverse prognosis . Because these abnormalities are seen in MGUS and SMM as well as myeloma, these abnormalities are considered necessary but not sufficient for the development of myeloma [44−46]. Although IGH translocations in the hyperdiploid category are uncommon, these two pathways are not mutually exclusive: in 10–15 % of cases, these translocations are found in the setting of a hyperdiploid karyotype, which may mitigate the translocation’s negative prognosis [38, 47, 48]. These recurrent primary translocations, which result from aberrant class-switch recombination, result in fusion of IGH with five partner genes: CCND1 (11q13), FGFR3 and MMSET (4p16), CCND3 (6p21), MAF (16q23), and MAFB (20q12) [44, 49]. As a result of these translocations, the partner genes become dysregulated by coming under the control of IGH enhancers; in the case of the t(4;14), both MMSET (on the derivative chromosome 4) and FGFR3 (on the derivative chromosome 14) may become upregulated . Common to all of these rearrangements is the direct or indirect dysregulation of cyclins D1, D2, and D3, whose normal function is to help regulate the transition between the G1 and S phases of the cell cycle [49, 51]. These translocations target cyclins either directly (e.g., CCND1 and CCND3) or indirectly (e.g., CCND2 via the 14;16 and 14;20 translocations) and lead to deregulation of the G1/S transition. Cyclin D dysregulation is not limited to these IGH translocations ; it is also found in the setting of hyperdiploidy . Additionally, biallelic dysregulation of CCND1 is reported in approximately 40 % of myeloma cases , the mechanism of which is unclear. Bergsagel et al.  have used these translocations to devise a classification of myeloma into eight different translocation/cyclin D groups.
Some studies show that there may be differences in the presenting clinical findings depending on the underlying cytogenetic abnormality .
A number of secondary genetic events have been reported (see Table 3); these are typically acquired by the malignant cells during the course of the disease, after the initiating genetic event has occurred. Many of these were initially detected by G-banding and/or FISH and have been well studied; newer genomic techniques are characterizing other clinically significant aberrations. The major secondary abnormalities of prognostic significance include monosomy 13 or loss of 13q, loss of 17p (including the TP53 locus), loss of 1p, and gain or amplification of 1q (particularly 1q21). NGS has also played an important role in the elucidation of intraclonal heterogeneity, a critical genomic event that occurs early in the course of disease and impacts the progression from MGUS to myeloma [32−34, 57]. Intraclonal heterogeneity is recognized as having implications not only for prognosis but also for choosing proper therapy.
Monosomy 13 (~ 85 %)/Deletion(13q) (~ 15 %)
Although losses or all or part of chromosome 13 used to be associated with an adverse prognosis, this is now thought to be due to its close association with IGH rearrangements and other high-risk genetic lesions [52, 58]. Studies have suggested that the timing of the acquisition of chromosome 13 abnormalities depends on other concomitant genetic abnormalities . It has also been postulated that losses involving chromosome 13 play a role in the progression from MGUS to MM, particularly in the setting of t(11;14) and t(6;14) .
Deletions involving the short arm of chromosome 17 are strongly associated with adverse prognosis [60−64]. When detected by FISH, its negative effect is most pronounced when the percentage of abnormal plasma cells is at least 60 % [60, 61]. Deletions of 17p may also be accompanied by mutation of the other TP53 allele: in one study, 37 % of MM patients with del(17p) had a TP53 mutation, whereas no mutation was found in patients lacking a deletion .
A number of studies have shown an association between deletions of 1p and adverse prognosis, including shorter progression free survival and overall survival in patients receiving high-dose chemotherapy and autologous hematopoietic stem cell transplant (HSCT) [22, 66−69]. Candidate genes implicated in this negative prognostic effect are FAM46C(1p12) and CDKN2C(1p32.3) ; deletion of CDKN2C is associated with increased proliferation (especially if a homozygous deletion) and worse overall survival . Deletion of 1p often occurs in conjunction with gain of 1q and loss of chromosome 13 [66, 69, 72]. Gene expression studies have shown that decreased expression of genes on 1p and elevated expression of genes on 1q are associated with high risk and short survival .