Chapter 13: Cytogenetics and Genetic Abnormalities

Over the past two decades, the genes that are located at the breakpoints of a number of the recurring chromosomal translocations have been identified. Alterations in the expression of the genes or in the properties of the encoded proteins resulting from the rearrangement play an integral role in the process of malignant transformation.^1,2 The altered genes fall into several functional classes, including tyrosine or serine protein kinases, cell surface receptors, growth factors and, the largest class, transcription factors. These latter proteins are involved in the induction or repression of gene transcription, often functioning in a tissue-specific fashion to regulate cell growth and differentiation.

There are two general mechanisms by which chromosomal translocations result in altered gene function. The first is deregulation of gene expression (Chap. 10). This mechanism is characteristic of the translocations in lymphoid neoplasms that involve the immunoglobulin genes in B-lineage tumors and the T-cell receptor genes in T-lineage tumors. These rearrangements result in the inappropriate or constitutive expression of an oncogene. The second mechanism is the encoding and expression of a novel fusion protein, resulting from the juxtaposition of coding sequences from two genes that are normally located on different chromosomes. Such chimeric proteins are “tumor specific” in that the fusion gene typically does not exist in nonmalignant cells. Thus, the detection of such a fusion gene or protein product can be important in diagnosis and in the detection of residual disease or early relapse. Moreover, they may also be appropriate targets for tumor-specific therapies. An example is the chimeric BCR-ABL1 protein resulting from the t(9;22) in chronic myeloid leukemia (CML) (see “Methods of Cell Preparation” below). All of the translocations cloned to date in the myeloid leukemias result in a fusion protein.

Chromosomal translocations result in the activation of genes in a dominant fashion. A number of human tumors result from homozygous, recessive mutations. These mutations lead to the absence of a functional protein product, suggesting that these genes function as “suppressor” genes, whose normal role(s) is to limit cellular proliferation. The hallmark of tumor suppressor genes is the loss of genetic material in malignant cells, resulting from chromosomal loss or deletion, as well as by other genetic mechanisms (Chap. 10).¹ A subset of tumor suppressor genes act by haploinsufficiency, whereby loss of one allele results in a reduction in the level of the protein product by half, thereby perturbing normal cellular processes. This mechanism is common in the recurring deletions in myeloid neoplasms (Chap. 83).

Extensive experimental evidence indicates that more than one mutation is required for the pathogenesis of hematologic malignancies. That is, expression of translocation-specific fusion genes or deregulated expression of oncogenes is required, but insufficient to induce leukemia. Thus, an important aspect of leukemia biology is the elucidation of the spectrum of chromosomal and molecular mutations that cooperate in the pathways leading to leukemogenesis. Where known, we describe the cooperating mutations associated with specific cytogenetic subsets of leukemia or lymphoma.

Cytogenetic analysis of malignant diseases should be based upon the study of the tumor cells themselves. In leukemia, the specimen is usually obtained by marrow aspiration and is typically cultured for 24 to 72 hours. When a marrow aspirate cannot be obtained, a marrow biopsy (bone core specimen) or a blood sample for patients who have circulating immature myeloid or lymphoid cells, can often be processed successfully. An involved lymph node or tumor mass specimen may be examined for the analysis of lymphoma cells.

For specimen collection, 1 to 5 mL of marrow are aspirated aseptically into a syringe coated with preservative-free sodium heparin and transferred to a sterile 15-mL centrifuge tube containing 5 mL of culture medium (RPMI 1640, 100 units sodium heparin). The use of Vacutainer tubes containing heparin as an anticoagulant should be avoided, as the heparin contains preservatives that suppress cell growth. If a marrow aspirate cannot be obtained, a marrow biopsy may be taken and placed into the collection tube. Approximately 75 percent of marrow biopsies can be minced to generate suspension of cells that will yield adequate numbers of metaphase cells for complete analysis. For blood specimens, 10 mL are drawn aseptically by venipuncture into a syringe coated with preservative-free heparin. To avoid loss of cell viability, it is critical that the specimen be transported at room temperature to the cytogenetics laboratory without delay. Overnight shipment of specimens frequently results in loss of cell viability, and most laboratories experience a high proportion (25–50 percent) of inadequate analyses using such specimens. For optimally handled specimens, approximately 95 percent of all cases should be adequate for cytogenetic analysis. Those cases that are inadequate generally represent samples from patients with hypocellular marrows.

Chromosomal abnormalities are described according to the International System for Human Cytogenetic Nomenclature (Table 13–1).³ To describe the chromosomal complement, the total chromosome number is listed first, followed by the sex chromosomes, and numerical and structural abnormalities in ascending order. The observation of at least two cells with the same structural rearrangement, for example, translocations, deletions or inversions, or gain of the same chromosome, or three cells each showing loss of the same chromosome, is considered evidence for the presence of an abnormal clone. However, one cell with a normal karyotype is considered evidence for the presence of a normal cell line. Patients whose cells show no alteration or nonclonal (single cell) abnormalities are considered to be normal. An exception to this is a single cell characterized by a recurring structural abnormality. In such instances, it is likely that this represents the karyotype of the mutated subclone in that particular patient.

FLUORESCENCE IN SITU HYBRIDIZATION

Cytogenetic analysis of human tumors is often technically difficult because of the presence of multiple abnormalities and requires highly skilled personnel. These factors have led investigators to seek alternative methods for identifying chromosomal abnormalities, such as fluorescence in situ hybridization (FISH).⁴ The FISH technique is based on the same principle as Southern blot analysis, namely, the ability of single-stranded DNA to anneal to complementary DNA.⁴ FISH can be performed on marrow or blood films, or fixed and sectioned tissue, as it does not require dividing cells. The target DNA is the nuclear DNA of interphase cells, or the DNA of metaphase chromosomes that are affixed to a glass microscope slide. Commercial probes are now available for the most common abnormalities, and are directly labeled with fluorochrome, which simplifies the technique by eliminating the probe preparation and detection steps. With the development of dual- and triple-pass filters, most laboratories now have the capacity to hybridize and detect two to three probes simultaneously. Table 13–1 summarizes the most frequently used commercially available FISH probes. Several types of probes can be used to detect chromosomal abnormalities by FISH. Hybridization of centromere-specific probes has been used to detect monosomy, trisomy, and other aneuploidies in both leukemias and solid tumors, as well as the sex chromosome complement in the transplant setting (Fig. 13–1).

Translocations and deletions can also be identified in interphase or metaphase cells by using genomic probes that are derived from the breakpoints of recurring translocations or within the deleted segment (see Fig. 13–1). In some cases, FISH analysis provides more sensitivity, in that cytogenetic abnormalities have been identified by FISH in samples that appeared to be normal by conventional cytogenetic analyses. Advantages of FISH include (1) the rapid nature of the method and the ability to analyze large numbers of cells; (2) its high sensitivity and specificity; and (3) the ability to obtain cytogenetic data from samples with a low mitotic index or from terminally differentiated cells. A further increase in sensitivity in cases with a low percentage of malignant cells can be achieved by performing FISH analysis on samples that have been enriched previously for specific subpopulations of cells. For example, the use of FISH in combination with plasma cell enrichment techniques is routinely applied in the clinical setting to maximize the detection rate of specific chromosome rearrangements in myeloma.^5,6 The major disadvantage of FISH testing is the inability to interrogate more than a few abnormalities. FISH is most powerful when the analysis is targeted toward those abnormalities that are known to be associated with a particular tumor or disease. In a clinical setting, cytogenetic analysis could be performed at the time of diagnosis to identify the chromosomal abnormalities in an individual patient’s malignant cells. Thereafter, FISH with the appropriate probes could be used to detect residual disease or early relapse, and to assess the efficacy of therapeutic regimens. For example, the use of FISH to detect the t(9;22) in CML patients following therapy with an oral tyrosine kinase inhibitor, or sex chromosome determination after a sex-mismatched transplant, is widespread. Material from patients newly presenting are often analyzed most efficiently by conventional cytogenetic analysis, combined with quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) analysis if a specific chromosome rearrangement is suspected, for example, a BCR-ABL1 fusion. Molecular qRT-PCR monitoring of the blood and marrow of CML patients is now part of the recommended testing for patient followup.⁷

MICROARRAY ANALYSIS

Several microarray-based technologies play an important role in the diagnosis and experimental analysis of hematologic malignancies, including high-density copy number/single nucleotide polymorphism (SNP) array testing (also known as chromosomal microarray analysis [CMA]), microarray-based gene expression profiling, and high-throughput SNP genotyping. CMA allows genome-wide detection of copy number abnormalities (deletions and duplications) at a much higher resolution than karyotyping; it also enables detection of loss of heterozygosity (LOH) that occurs without concurrent changes in the gene copy number, that is, uniparental disomy (UPD) (Chap. 10), and can be attributed to somatic mitotic recombination (also referred to as copy-neutral LOH) (Fig. 13–2). CMA is clinically available as an adjunct test to karyotyping and FISH, and it facilitates detection of genomic abnormalities in a substantial proportion of patients with myelodysplastic syndromes (MDSs) and leukemia with a normal karyotype; it can also be used as a cost-effective alternative to large panels of FISH probes, and as a very useful tool to characterize chromosomal abnormalities of uncertain significance. Microarray-based gene expression profiling has been applied to study a variety of hematopoietic malignancies, often revealing complex but unique expression signatures for each disease subtype. For example, gene expression profiling has shown that diffuse large B-cell lymphoma (DLBCL) comprises at least three different subtypes—germinal center B-cell (GCB)-like, activated B-cell (ABC)-like, and primary mediastinal B-cell lymphoma (PMBL)—each with a distinct oncogenic mechanism, prognosis, and response to therapies (Chap. 98).⁸ Expression profiling studies led to the recognition of a novel genetic subtype of high-risk B-cell acute lymphocytic or lymphoblastic leukemia (ALL), which has a similar expression signature as Ph-chromosome-positive ALL (Chap. 91).^9,10 High-throughput array-based SNP genotyping, enabling the analysis of large numbers of cases and controls, has facilitated genome-wide association studies for the identification of disease susceptibility loci.¹¹ Future diagnostic evaluation, identification of high-risk cases and management decisions will be increasingly based on genomic and even proteomic profiling of patients’ germline and tumor samples, which will likely utilize a combination of array-based and next-generation sequencing–based technologies.

CHRONIC MYELOID LEUKEMIA

The first consistent chromosomal abnormality in any malignant disease was identified in CML (Chap. 89). The Philadelphia (Ph) chromosome results from a translocation involving chromosomes 9 and 22, t(9;22)(q34.1;q11.2), (Fig. 13–3), and arises in a pluripotential stem cell that gives rise to both lymphoid and myeloid lineage cells. The standard t(9;22) is identified in approximately 92 percent of CML patients, whereas 6 to 8 percent have variant translocations that involve a third chromosome in addition to chromosomes 9 and 22 (see Chap. 89, Fig. 89–8). The genetic consequences of the t(9;22) or the complex translocations are to move a segment of the Abelson (ABL1) oncogene on chromosome 9 next to a segment of the BCR gene on 22. Analyses of leukemia cells from rare patients with typical CML lacking the t(9;22) has revealed a rearrangement involving ABL1 and BCR that is detectable only at the molecular level (1 to 2 percent of cases).¹²

The t(9;22) and resultant BCR-ABL1 fusion is the sine qua non of CML.¹² The BCR-ABL1 fusion protein is located on the cytoplasmic surface of the cell membrane and acquires a novel function in transmitting growth-regulatory signals to the nucleus via the RAS/MAPK, phosphatidylinositide 3′-kinase (PI3K)/AKT, and Janus kinase (JAK)/signal transducer and activator of transcription (STAT) signal transduction pathways. The tyrosine kinase activity of the BCR-ABL1 fusion protein can be specifically inhibited by several commercially available oral tyrosine kinase inhibitors (TKIs): imatinib mesylate (Gleevec/STI571, Novartis Pharmaceuticals, East Hanover, NJ), dasatinib (Sprycel, BMS-354825, Bristol-Myers Squibb, Princeton, NJ), and nilotinib (Tasigna, AMN107, Novartis Pharmaceuticals, East Hanover, NJ) (Chap. 89). Additional oral agents are also being tested in clinical trials.^13,14 The BCR-ABL1 translocation can be detected by cytogenetic and FISH analysis, qRT-PCR, and Southern blot analysis to diagnose the disease and detect residual disease. Studies of patients treated with TKIs show a strong correlation between BCR-ABL1 levels as measured in the blood by qRT-PCR and the percentage of Ph+ cells in the marrow.

Several types of genetic changes are associated with imatinib resistance, including point mutations leading to amino acid substitutions in the BCR-ABL1 kinase domain that interfere with imatinib binding, as well as the acquisition of additional copies of the Ph chromosome or BCR-ABL1 gene amplification, both of which can be detected by FISH.¹⁴ Besides duplications of the BCR-ABL1 fusion, additional TKI resistance-associated genomic lesions, including acquired regions of LOH on chromosomes 1, 8, 9, 17, 19, and 22, have been detected by SNP array analysis.¹⁵ Although some patients who achieve a complete cytogenetic response on imatinib develop clonal karyotypic abnormalities, most commonly +8, −7, or del(20q), the majority of them do not go on to develop the clinical features of MDS.¹⁶ The significance of these early findings will be elucidated by the analysis of a large number of patients who have had complete cytogenetic responses to TKIs and are being followed prospectively.

As they enter the more aggressive stages of accelerated and blast phase disease, historically 80 percent of CML patients showed karyotypic evolution with the appearance of new chromosomal abnormalities in very distinct patterns in addition to the Ph chromosome. A change in the karyotype was considered to be a grave prognostic sign.¹⁷ The most common changes, a gain of chromosomes 8 or 19, or a second Ph chromosome (by gain of the first), or an i(17q), frequently occurred in combination to produce modal chromosome numbers of 47 to 50. Other genetic changes identified in CML in blast crisis include mutations in the TP53, RB1, MYC, CDKN2A (p16), KRAS/NRAS, or RUNX1/AML1 genes. With the advent of TKI therapy, the natural history of CML has been altered, and the karyotype in blast phase appears to differ from that seen previously. However, the pattern of abnormalities is not yet well described.

Rarely, marrow biopsies from patients will appear similar to those patients with CML, but will lack a Ph chromosome or the BCR-ABL1 fusion. Most often these patients have a MDS or myeloproliferative neoplasm (MPN), most commonly chronic myelomonocytic leukemia, refractory anemia with excess blasts (RAEB), or the poorly understood disorder of “atypical CML.” Some of the latter have JAK2^V617F mutations and a phenotype consistent with chronic neutrophilic leukemia (Chaps. 84 and 89). Cytogenetic analysis of marrow biopsies from these patients commonly have a normal karyotype, +8, +13, del(20q), or i(17q). These patients have a substantially shorter survival than do those whose cells have the t(9;22). Because each of the oral TKIs blocks kinase activities in addition to BCR-ABL1, they have proven to be effective in other disorders, including chronic MPNs with platelet-derived growth factor receptor (PDGFR)-β rearrangements, a myeloproliferative variant of hypereosinophilic syndrome that expresses the FIP1L1-PDGFRA fusion protein, and in patients with mast cell malignancies that express an activating point mutation in KIT (Chap. 89).¹⁸

OTHER MYELOPROLIFERATIVE NEOPLASMS

A cytogenetically abnormal clone is present in 15 percent of untreated polycythemia vera patients compared with 40 percent of treated patients (Table 13–2).¹⁹ When the disease transforms to acute myeloid leukemia (AML), almost 100 percent have a cytogenetically abnormal clone. The presence of a chromosome abnormality at diagnosis does not necessarily predict a short survival or the development of leukemia. However, a change in the karyotype may be an ominous sign. Marrow cells frequently contain additional chromosomes (+8 or +9). Trisomy 8 and 9 may occur together which is otherwise rare.¹⁹ Structural rearrangements most often involve a del(13q) or del(20q), noted in 30 percent of patients. Loss of chromosome 7 (20 percent) and del(5q) (40 percent) are often observed in the leukemic phase, and may be related to the prior treatment received by these patients (Chap. 84).

Cytogenetic analysis of cells from patients with primary myelofibrosis has revealed clonal abnormalities in 60 percent of patients (Chap. 86).¹⁹ These abnormalities are similar to those noted in other myeloid disorders. The most common anomalies are +8, −7, or a del(7q), del(11q), del(13q), and del(20q).¹⁹ A change in the karyotype may signal evolution to AML. Fewer than 10 percent of patients with essential thrombocythemia have an abnormal clone (Chap. 85). Recurring abnormalities include +8 and del(13q). Although del(5q) and inv(3)/t(3;3) are associated with thrombocytosis, they are characteristic of MDS or AML, rather than essential thrombocythemia.

Mutant JAK2^V617F is a constitutively active tyrosine kinase that activates the STAT, PI3Ks, and mitogen-activated protein kinases (MAPKs) signalling pathways downstream of the erythropoietin receptor, thrombopoietin receptor, or the granulocyte colony-stimulating factor (G-CSF) receptor to promote proliferation and transformation of hematopoietic progenitor cells (Chap. 84). JAK2 mutations occur in polycythemia vera (95 percent), essential thrombocythemia (approximately 50 percent), and myelofibrosis (approximately 50 percent) (Chaps. 84 to 86).²⁰ In refractory anemia with ring sideroblasts (RARS) with thrombocytosis (RARS-t), a MDS/MPN, unclassified by the World Health Organization (WHO) classification, 60 percent of patients have the JAK2^V617F mutation, and present with higher white blood cell and platelet counts (Chap. 87).²¹ Less commonly, activation of the JAK-STAT pathway in MPNs may result from JAK2 exon 12 mutations (1 to 2 percent of polycythemia vera cases) or mutation of the thrombopoietin receptor, MPL (approximately 2 percent of essential thrombocythemia and approximately 5 percent of myelofibrosis). The majority of essential thrombocythemia and myelofibrosis patients with nonmutated JAK2 carry somatic mutations in the calreticulin gene (CALR) (Chaps. 85 and 86).^22,23

MYELODYSPLASTIC SYNDROMES

The MDSs are a heterogeneous group of neoplasms, including refractory cytopenia with unilineage dysplasia, RARS, refractory cytopenia with multilineage dysplasia (RCMD), RAEB–1,2, MDS with isolated del(5q), MDS unclassifiable, and childhood MDS, including refractory cytopenia of childhood (Chap. 87).²⁴ Clonal chromosome abnormalities can be detected in marrow cells of approximately 50 percent of patients with primary MDS at diagnosis (refractory anemia [RA], 25 percent; RARS, 10 percent; RCMD, 50 percent; RAEB –1,2, 50 to 70 percent; MDS with isolated del (5q), 100 percent) (see Table 13–2).^25,26 The proportion varies with the risk that a subtype will transform to AML, which is highest for RCMD and RAEB. The common chromosome changes, +8, del(5q), −7/del(7q), and del(20q), are similar to those seen in AML de novo. The recurring translocations that are closely associated with the distinct morphologic subsets of AML de novo are almost never seen in MDS. With the exception of MDS with isolated del(5q), the chromosome changes show no close association with the specific subtypes of MDS. MDS with isolated del(5q) occurs in a subset of older patients, frequently women, with RA, generally low blast counts, and normal or elevated platelet counts.²⁷ These patients have an interstitial deletion of 5q, typically as the sole abnormality, and can have a relatively benign course that extends over several years (Chap. 87).²⁷ Diagnostic and prognostic information for the patients with a normal karyotype can be provided by CMA, which can detect abnormalities in 10 to 15 percent of these cases. Some abnormalities detectable by CMA, including submicroscopic microdeletions in 4q24 affecting the TET2 gene, as well as LOH 7q, LOH 11q, and LOH 17p, were shown to be associated with a poor outcome in MDS.^28,29

Cytogenetic abnormalities in MDS are predictive of survival and progression to AML.²⁶ Patients with a “very good outcome” have −Y or del(11q) as the sole abnormality; those with a “good outcome” have normal karyotypes, del(5q) alone, or with one additional abnormality, del(12p) alone, or del(20q) alone; those with an “intermediate outcome” have del(7q), +8, +19, i(17q), or any other single or double abnormality; those with a “poor outcome” have −7, inv(3q)/t(3;3), double abnormalities, including −7/del(7q), and complex karyotypes with 3 abnormalities; and those with a “very poor outcome” have complex karyotypes with more than three abnormalities, typically with abnormalities of chromosome 5.²⁶ With larger data sets, the inclusion of additional rare recurring cytogenetic abnormalities has facilitated a refinement of the cytogenetic risk groups, and provided the clinician with more information to predict the expected outcome for their patient.^30,31

ACUTE MYELOID LEUKEMIA DE NOVO

Clonal chromosomal abnormalities are detected in 80 to 90 percent of patients with AML. The most frequent abnormalities are +8 and −7, which are seen in most subtypes of AML.¹ Specific rearrangements are closely associated with particular subtypes of AML as recognized by the WHO and French-American-British (FAB) classification schemes (see Table 13–2; Chap. 88).³²

Translocation 8;21

The 8;21 translocation [t(8;21)(q22;q22.3)], described in 1973, was the first translocation identified in AML (see Fig. 13–3). The t(8;21) is common and is observed in 5 to 10 percent of all patients with AML with an abnormal karyotype and in 10 percent of patients with AML with maturation. This translocation is the most frequent abnormality in children with AML and occurs in 15 to 20 percent of karyotypically abnormal cases. Loss of a sex chromosome (−Y in males, −X in females), or a del(9q) with loss of 9q22, accompanies the t(8;21) in 75 percent of cases. The presence of the t(8;21) identifies a morphologically and clinically distinct subset of AML, and most cases with the t(8;21) are classified as AML with maturation. AML with the t(8;21) has a favorable prognosis in adults (overall 5-year survival of 70 percent), but the outcome in children is poor.³³ At the molecular level, the t(8;21) involves the RUNX1/AML1 gene, which encodes a transcription factor, also known as core-binding factor, that is essential for hematopoiesis. The RUNX1 gene on chromosome 21 is fused to the RUNX1T1/ETO gene on chromosome 8 and results in a RUNX1-RUNX1T1 chimeric protein. Transformation by RUNX1-RUNX1T1 likely results from transcriptional repression of normal RUNX1 target genes via aberrant recruitment of nuclear transcriptional corepressor complexes.³³

Inversion 16 and Translocation 16;16

Another clinical–cytogenetic association involves acute myelomonocytic leukemia (AMML) with abnormal eosinophils, including large and irregular basophilic granules, and positive reactions with periodic acid–Schiff and chloroacetate esterase. Most patients have an inversion of chromosome 16, inv(16)(p13.1q22) (see Fig. 13–3), but some have a t(16;16)(p13.1;q22), and the WHO classification system now recognizes these as a distinct form of AML (Chap. 88). These aberrations are relatively common, occurring in 5 percent of AML and 25 percent of AMML patients.¹ These patients have a good response to intensive chemotherapy with a complete remission rate of approximately 90 percent and an overall 5-year survival of 60 percent.³³ The breakpoint at 16q22 occurs within the CBFB gene, which encodes one subunit of the RUNX1/CBFB transcription factor. Thus, like the t(8;21), the inv(16) disrupts the RUNX1/AML1 pathway regulating hematopoiesis. Secondary cooperating mutations of KIT, KRAS, and NRAS are common in core-binding factor-associated leukemias, although only KIT mutations confer a poor prognosis.³³

Translocation 15;17

The t(15;17)(q24.1;q21.1) (see Fig. 13–3) is specific for acute promyelocytic leukemia (APL) and has not been found in any other disease.³⁴ Rare variant translocations, which occur in less than 2 percent of cases, include the t(11;17)(q23.2;q21.1) and t(5;17)(q35.1;q21.1), which result in the ZBTB16 (PLZF)-RARA and NPM1-RARA fusion proteins, respectively. Establishing the diagnosis of APL with the typical t(15;17) is important, because this disease is sensitive to therapy with all-trans retinoic acid, whereas other cases of AML and some of the APL-like disorders associated with the variant translocations do not respond to this treatment (Chap. 88). The t(15;17) results in a fusion retinoic acid receptor-α protein (PML-RARA). The oncogenic potential of the APL fusion proteins appears to result from the aberrant repression of RARA-mediated gene transcription through histone deacetylase (HDAC)-dependent chromatin remodeling. Genetic mutations that cooperate with PML-RARA include FLT3 internal tandem duplications (ITDs), observed in 35 percent of patients.

Translocations Involving 11q

Recurring translocations involving 11q23.3 are seen in approximately 35 percent of acute monocytic leukemia patients and are of great interest in acute leukemia for at least three reasons.^1,35 First, there are more than 50 different recurring rearrangements that involve 11q23.3 and, thus, along with 14q32.3, 11q23.3 is one of the bands most frequently involved in rearrangements in human tumor cells.^35,36 The most common breakpoints in the translocation partners include 1p32, 4q21.3, and 19p13.3 in ALL (Chap. 91), and 1q21, 2q21, 6q27, 9p21.3, 10p12, 17q25, 19p13.3, and 19p13.1 in AML (Chap. 88). Second, these translocations occur in both lymphoid and myeloid leukemias. One common translocation in infants, t(4;11)(q21.3;q23.3), has a lymphoblastic phenotype, whereas other translocations, such as the t(9;11)(p21.3;q23.3) (see Fig. 13–3) and t(11;19)(q23.3;p13.1), are common in acute monocytic leukemias. Finally, translocations involving 11q23.3 have a very unusual age distribution, comprising about three-quarters of the chromosome abnormalities in leukemia cells of children younger than 1 year of age.³⁵ With the exception of the t(9;11) which may have an intermediate outcome, translocations of 11q23.3 are associated with a poor outcome.³² Translocations of 11q23.3 involve KMT2A/MLL, a very large gene (>100 kb) with multiple transcripts of 12 to 15 kb. KMT2A protein is a histone methyltransferase that assembles in protein complexes that regulate gene transcription via chromatin remodeling.³⁶ All of the KMT2A translocations identified to date result in fusion proteins.

Trisomy 11

Trisomy 11 is a rare abnormality, noted as a sole aberration in 1 to 2 percent of MDS or AML, and confers an unfavorable outcome.³⁷ It is notable that an ITD of the KMT2A gene is detected in 90 percent of AMLs with +11 as the sole abnormality and in 10 percent of AML cases with a normal karyotype. The rearrangement is the result of a duplication of KMT2A exons 2 to 6 or 2 to 8 mediated by recombination between Alu repetitive elements and may produce a partially duplicated protein.

Inversion 3 and t(3;3)

Each of the other recurring rearrangements in AML occurs in fewer than 3 percent of patients. A unique feature of abnormalities involving the long arm of chromosome 3 [inv(3)(q21.3q26.2) or t(3;3)(q21.3;q26.2)] is the presence of platelet counts greater than 100 × 10⁹/L, sometimes greater than 1000 × 10⁹/L, and an increase in marrow megakaryocytes, especially micromegakaryocytes.¹ Most of the recurring translocations described above occur in younger patients with a median age in the 30s, whereas other abnormalities, such as del(5q), or −7/del(7q), occur in patients with a median age greater than 50 years. Moreover, many of the latter patients have occupational exposure to mutagenic agents, such as solvents, petroleum, and pesticides.

Mutations

The prognosis of patients with AML is also determined by mutations, most commonly of the FLT3, NPM1, CEBPA, or KIT genes (Table 13–3).³⁸ Mutations of the FLT3 gene, including both point mutations within the tyrosine kinase domain and ITDs, are among the most common genetic changes seen in AML, occurring in 15 to 35 percent of cases. FLT3-ITD mutations may occur in any subtype of AML, but are most common in APL and AML with a normal karyotype, and are associated with a poor prognosis, particularly in those cases with loss of the remaining wild type FLT3 allele.³⁹ Mutations of the FLT3 tyrosine kinase domain (codons 835 or 836 of the second tyrosine kinase domain) are noted in 5 to 8 percent of AML.³⁹ Mutations of NPM1 also occur frequently in AML (35 percent of adult cases, and 80 to 90 percent of acute monocytic leukemia), but are less frequent in patients with recurring cytogenetic abnormalities. In the absence of FLT3 mutations, NPM1 mutations are associated with a favorable prognosis.⁴⁰ NPM1 mutations, most commonly involve exon 12, resulting in alterations at the C-terminus, that is, replacement of tryptophan(s) at position 288 and 290, and aberrant localization of the protein to the cytoplasm. CEBPA mutations (6 to 15 percent of all AMLs), are often biallelic, and are usually associated with intermediate risk cytogenetics, but are generally associated with a favorable prognosis.³⁸ Mutations of KIT are noted in 2 percent of AMLs, and have prognostic significance among AMLs with t(8;21) and inv(16)/t(16;16) (20 to 25 percent), in which they are associated with a poor prognosis.³³ With respect to epigenetic changes, transcriptional silencing via DNA methylation of the CDKN2B (p15^INK4B) gene is observed in a high percentage of patients with AML or therapy-related myeloid neoplasm, and is associated with −7/del(7q), and a poor prognosis.⁴¹

Whole-genome and exome sequencing studies implicated a number of additional genes in pathogenesis of AML, including DNMT3A, TET2, ASXL1, IDH1, IDH2, PHF6, WT1, TP53, RUNX1, and EZH2. Still, AML genomes have a relatively few mutations compared to other adult cancers (average of 13 mutations), and only approximately 20 genes appear to be mutated in AML with a significant frequency.⁴² Several of the newly identified genetic abnormalities have prognostic importance in AML, including DNMT3A, TET2, ASXL1, and PHF6 mutations, and KMT2A-PTD, which are all associated with a poor outcome.⁴³ As more data accumulate regarding the clinical relevance of the specific mutation in homogeneously treated clinical cohorts it will become important to develop and implement assays that allow for cost-effective, rapid molecular profiling in the clinical setting.

In addition, the emerging molecular analysis of tumor tissue presents an opportunity to identify individuals with germline predisposition to cancer.^44,45,46 Depending on the type of bioinformatic analysis that is performed on tumor tissue, it is possible to identify germline mutations when analyzing an individual’s primary tumor, as every cell in that person’s body will contain the germline mutation. For example, the current standard for molecular analysis in the case of a patient presenting with a new AML is to perform mutational analysis of CEBPA. CEBPA is mutated sporadically in AML, but the familial form is associated with biallelic CEBPA mutations, most commonly with the germline mutation found within the 5′ end of the gene, accompanied by the acquisition of a second 3′ mutation in the leukemia. Germline 3′ CEBPA mutations have also been identified.^47,48 In approximately 10 percent of AML patients found to have biallelic CEBPA mutations within their leukemia cells, one of these mutated alleles is actually a germline mutation and, therefore, any AML patient found to have biallelic CEBPA mutations should undergo genetic counseling and molecular testing of germline tissue.^44,49,50,51 This type of scenario may become more common as next-generation sequencing of tumor tissue is performed more frequently. The American College of Medical Genetics and Genomics (ACMG) has published a series of commentaries and recommendations regarding disclosure of genetic information when clinical genetic sequencing is performed.^52,53 The ACMG recommends disclosure of genetic information regarding 24 genes that confer germline cancer predisposition.⁵³ Figure 13–4 shows the frequency of common cytogenetic abnormalities in AML occurring de novo.

THERAPY-RELATED MYELOID NEOPLASMS

Therapy-related myeloid neoplasms (t-MNs), comprised of therapy-related MDS (t-MDS) and therapy-related AML (t-AML), are a late complication of cytotoxic therapy used in the treatment of both malignant and nonmalignant diseases.⁵⁴ In patients who received alkylating agents, the characteristic recurring chromosome abnormalities observed are loss of part or all of chromosomes 5 [del(5q)] and/or part or all of chromosome 7 [−7/del(7q)] (see Fig. 13–4). Clinically, these patients have a long latency period (5 years), present with MDS, which often progresses rapidly to AML with multilineage dysplasia and a poor prognosis. In our experience, 92 percent of t-MN patients had an abnormal karyotype and 70 percent had an abnormality of one or both chromosomes 5 and 7,⁵⁵ and these observations have been confirmed in other series.⁵⁶ In contrast, only approximately 20 percent of patients with AML de novo have a similar abnormality of chromosomes 5 or 7 or both.¹

By cytogenetic and molecular analysis, investigators have defined a 970-kb commonly deleted segment (CDS) containing 19 genes on the long arm of chromosome 5 (5q31.2) predicted to contain a myeloid tumor suppressor gene.⁵⁷ A second, nonoverlapping CDS in 5q32 is implicated in MDS with an isolated del(5q).⁵⁸ Parallel studies revealed a 2.5-Mb CDS within 7q22 containing 16 genes. Molecular analysis of the genes within these regions did not reveal inactivating mutations in the remaining alleles, nor was there evidence of transcriptional silencing.⁵⁷ These observations are compatible with a haploinsufficiency model (gene dosage effect resulting from the loss of one allele), and several candidate haploinsufficient genes (EGR1, APC, CSNK1A1, RPS14) have been identified on 5q. The EGR1 transcription factor is downstream of cytokine signaling pathways. In a mouse model, loss of a single allele of Egr1 cooperates with mutations induced by an alkylating agent in the development of myeloid diseases.⁵⁹ RPS14 encodes an essential component of the 40S subunit of ribosomes, and haploinsufficiency of this gene appears to be responsible for the defect in erythropoiesis in MDS with an isolated del(5q).⁶⁰ Other studies show that haploinsufficiency of two micro-RNAs, miR-145 and miR-146a, encoded by sequences near the RPS14 gene, cooperate with loss of RPS14 and mediate the megakaryocytic dysplasia seen in this disease.⁶¹ These studies raise the possibility that haploinsufficiency for one or more of these genes in hematopoietic stem cells (HSCs) may contribute to the pathogenesis of MDS or AML with a del(5q), and a study demonstrated that haploinsufficiency for two del(5q) genes, EGR1 and APC, together with loss of TP53 leads to AML in a mouse model.⁶²

A second subtype of t-AML has been identified that is distinctly different from the more common leukemia that follows alkylating agents or irradiation. This type of t-AML is seen in patients receiving drugs known to inhibit topoisomerase II, for example, etoposide, teniposide, and doxorubicin. Clinically, these patients have a shorter latency period (1 to 2 years), present with overt leukemia, often with monocytic features, without a preceding myelodysplastic phase, and have a more favorable response to intensive induction therapy. Balanced translocations involving the KMT2A gene at 11q23.3, or the RUNX1/AML1 gene at 21q22.3 are common in this subgroup.⁵⁴

ACUTE LYMPHOBLASTIC LEUKEMIA

ALL is the most frequent leukemia in children (Chap. 91). In both childhood and adult ALL, the identification of prognostic subgroups based on recurring cytogenetic abnormalities (Table 13–4) and molecular markers has resulted in the application of risk-adapted therapies.⁶³ The most useful prognostic indicators are karyotype (including ploidy), age, white blood cell count, and response to initial therapy (day 14 marrow response and end-induction minimal residual disease). Based on these parameters, the Children’s Oncology Group has defined four risk groups: lower risk (5-year event-free survival [EFS], at least 85 percent), with either the ETV6/RUNX1 fusion, or simultaneous trisomies of chromosomes 4, 10, and 17; standard and high risk (those remaining in the respective National Cancer Institute risk groups); and very high risk (5-year EFS, 45 percent or below), with extreme hypodiploidy (fewer than 44 chromosomes), or the BCR-ABL1 fusion, and induction failure.⁶⁴ Genome-wide profiling studies using CMA revealed a high frequency of submicroscopic copy-number abnormalities in pediatric ALL, including deletions of PAX5 (32 percent), IKZF1 (IKAROS, 29 percent), CDKN2A/B (50 percent), BTG1, and EBF1 (8 percent). Many of these abnormalities disrupt genes and pathways controlling B-cell development and differentiation, and the most clinically significant among them appears to be genetic alterations of IKZF1, which are invariably associated with a very poor outcome in B-cell progenitor ALL.⁹

Translocation 9;22

The incidence of the t(9;22) in ALL is 30 percent in adults (the incidence may approach 50 percent in adults older than 60 years of age) and 5 percent in children. Thus, the Ph chromosome is the most frequent rearrangement in adult ALL. Approximately 70 percent of the patients show additional abnormalities, a frequency that is substantially higher than that observed in CML with +der(22)t(9;22),+21, abnormalities of 9p, +8, −7, and +X (noted in descending frequency). Monosomy 7 is associated with a poorer outcome.⁶⁵ A chromosomally normal cell line is frequently noted in the marrow of Ph chromosome-positive ALL patients (70 percent), but is rare in untreated CML. Most cases have a B-lineage phenotype (CD10+, CD19+, and TdT+), but there is frequent expression of myeloid-associated antigens (CD13 and CD33). The disease in both adults and children is characterized by high white blood cell counts, a high percentage of circulating blasts, and a poor prognosis. As in CML, the t(9;22) in ALL results in a BCR-ABL1 fusion gene. However, in more than half of the patients, the break in BCR is more proximal, resulting in a smaller fusion protein with even greater tyrosine kinase activity (BCR-ABL1^p190). Genetic alterations of the IKZF1 gene are detectable in up to 80 percent of patients with Ph chromosome–positive ALL, and are associated with an unfavorable outcome even with the use of TKIs.⁶⁶

Translocations Involving 11q

Translocations involving the KMT2A gene at 11q23.3 are observed in 5 percent of ALL patients.⁶⁷ Of these, the most common is the t(4;11)(q21.3;q23.3) (Fig. 13–5). The t(11;19)(q23.3;p13.3) is second in frequency. However, this rearrangement is not limited to ALL in that approximately 50 percent of these cases have AML, usually monoblastic. Of note is the high frequency of translocations involving 11q23.3 in infant ALL (60 to 80 percent). Patients with the t(4;11) have a pro B-cell phenotype (CD10−, CD19+), with coexpression of monocytic (CD15+), or, less commonly, T-cell markers. Clinically, both children and adults have aggressive features with hyperleukocytosis, extramedullary disease, and a poor response to conventional chemotherapy.⁶⁷ Adults with the t(4;11) have a remission rate of 75 percent, but a median EFS of only 7 months. Rearrangements affecting KMT2A represent a major class of mutations in acute leukemia and identify patients with a poor outcome.

Translocation 12;21

The t(12;21)(p13.2;q22.3) has been identified in a high proportion (approximately 25 percent) of childhood precursor B-cell leukemia, but is uncommon in adults (approximately 5 percent of ALL cases) (Fig. 13–6).⁶⁸ The translocation is not easily detected by cytogenetic analysis because of the similarity in size and banding pattern of 12p and 21q. However, the rearrangement can be detected reliably using reverse transcriptase polymerase chain reaction (RT-PCR) or FISH analysis. The t(12;21) defines a distinct subgroup of patients characterized by an age between 1 and 10 years, B-cell lineage immunophenotype (CD10+, CD19+, HLA-DR+), and a favorable outcome, particularly when other favorable risk factors are present. In one study, patients with the t(12;21) had a 5-year EFS of 91 percent as compared to 65 percent for patients without this rearrangement. However, the t(12;21) may be associated with late disease recurrences. The t(12;21) results in a fusion protein containing the N-terminus of ETV6/TEL, a transcriptional repressor of the ETS family, and most of the RUNX1/AML1 transcription factor.

Hyperdiploidy

The leukemia cells of some patients with ALL are characterized by a gain of many chromosomes (see Fig. 13–6). Two distinct subgroups are recognized: a group with 1 to 4 extra chromosomes (47 to 50), and the more common group with more than 50 chromosomes. Chromosome numbers usually range from 51 to 60, and a few patients may have up to 65 chromosomes. Hyperdiploidy (>50 and usually <66 chromosomes) is common in children (approximately 30 percent), but is rarely observed in adults (<5 percent). Certain additional chromosomes are common (X chromosome, and chromosomes 4, 6, 10, 14, 17, 18, and 21). Chromosome 21 is gained most frequently (100 percent of cases). Patients who have hyperdiploidy with more than 50 chromosomes have all of the previously recognized clinical factors that indicate a good prognosis, including age between 1 and 9 years, low white blood cell count (median 6700/μL), and favorable immunophenotype (early pre-B cell or pre-B cell).⁶⁹ The favorable prognosis associated with high hyperdiploidy is associated with gains of chromosomes 4, 10, and 17, whereas a gain of chromosomes 5 and i(17q) is associated with a poor outcome.⁶⁹

Translocation 1;19 and translocation 8;14

The t(1;19)(q23;p13.3) has been identified in approximately 6 percent of children with a B-lineage leukemia (see Fig. 13–5). The leukemia cells have cytoplasmic immunoglobulin and are CD10+, CD19+, CD34−, and CD9+. A reciprocal translocation involving the long arms of chromosomes 8 and 14 [t(8;14)(q24.2;q32.3)] is observed in mature B-cell ALL (see Fig. 13–5).⁷⁰ These patients have a high incidence of central nervous system involvement and/or abdominal nodal involvement at diagnosis. Although the outcome for both children and adults with a t(8;14) has been poor, the use of high intensity chemotherapy has markedly improved the outcome (EFS of 80 percent in children).⁷⁰

Philadelphia Chromosome–like Acute Lymphocytic/Lymphoblastic Leukemia

Ph-like ALL is a novel subgroup of high-risk ALL, characterized by increased expression of HSC genes, and a similar gene expression profile to Ph-positive ALL. Like Ph-positive ALL, Ph-like cases are also characterized by a high frequency of IKZF1 deletions and mutations, which confer a poor prognosis.^9,10 Ph-like ALL comprises up to 15 percent of pediatric ALL and up to 30 percent of adult ALL and is associated with a higher risk of relapse compared to other Ph-negative cases. Genetic alterations responsible for the activated kinase and cytokine receptor signaling signature in Ph-like ALL are starting to be elucidated, and include point mutations and gene fusions affecting CRLF2, JAK2, ABL1, PDGFRB, EPOR, EBF1, FLT3, IL7R, SH2B3, and other genes.⁷¹

T-CELL ACUTE LYMPHOBLASTIC LEUKEMIA

T lymphoblastic leukemia/lymphoma has a distinct pattern of recurring karyotypic abnormalities.⁷² Rearrangements involving 14q11.2 (see Fig. 13–5) and two regions of chromosome 7, (7q34) and (7p14), are particularly frequent in T-cell malignancies (see Table 13–4). The most common are the t(10;11)(q24.3;q11.2) (7 percent of childhood and 30 percent of adult cases, TLX1 gene); the cryptic t(5;14)(q35.1;q32.1) (TLX3, 20 percent of childhood and 10 to 15 percent of adult cases), t(11;14) (p13;q11.2) (approximately 3 percent, LMO2 gene), and t(7;9)(q34;q34.3) (approximately 2 percent, NOTCH1 gene). Approximately 30 percent of patients have activating mutations of the NOTCH1 gene. Patients with T-cell ALL are most often young males and often have a mediastinal tumor mass, high white blood cell count, and leukemia cells in the cerebrospinal fluid. These same clinical characteristics are associated with lymphoblastic lymphoma, another T-cell malignancy.

CHRONIC LYMPHOCYTIC LEUKEMIA

The chromosomal abnormalities associated with chronic lymphocytic leukemia (CLL) have been delineated through the use of FISH (Chap. 92).⁷³ When conventional cytogenetic techniques are used, only 50 percent of CLL patients have detectable chromosomal abnormalities. The most common abnormality is trisomy 12 (20 to 60 percent), followed by structural abnormalities of 13q and 14q (see Table 13–4). However, when FISH analysis is used to study specific abnormalities, chromosomal abnormalities can be detected in greater than 80 percent of patients. The most frequent chromosomal changes seen by FISH are: loss or deletion of 13q (55 percent); deletion of 11q, the location of the ATM gene (18 percent); trisomy of 12q (16 percent); deletion of 17p, the location of the TP53 gene; and deletion of 6q (6 percent). The LOH affecting 17p frequently coincides with the TP53 mutations (7 percent) and can be detected by CMA.⁷⁴ Patient survival correlates with cytogenetic subtype, with a shorter median survival observed in patients with 17p (32 months) or 11q (79 months) deletions, than in those with no detectable abnormality (111 months), trisomy of 12q (114 months), or −13/del(13q) (133 months). Two micro-RNA genes (miR-16–1 and miR-15a) are possible target genes in the 13q14.3 region. FISH probes capable of detecting the deletions of 11q, 13q, and 17p, trisomy 12, and immunoglobulin heavy chain (IGH) translocations are commercially available, and facilitate the application of risk-adapted treatment strategies.

The prognosis of patients with CLL is also determined by two other molecular abnormalities: the status of the IGH variable region and the expression level of CD38 (Chap. 92). Patients whose CLL cells express IGH genes containing somatic mutations have a 24-year median survival compared to only 6 to 8 years in those patients who do not have somatic IGH gene mutations.⁷⁵ This simple grouping of patients based on the mutation status of the IGH gene may reflect the fact that CLL cells that have few or no IGH mutations also often contain chromosomal aberrations that confer a poor prognosis, for example deletions of 11q or 17p, or trisomy 12, whereas CLL cells with IGH mutations often contain deletions of 13q, which confer a more favorable clinical course. Unfortunately, testing for somatic mutations in the IGH gene is not currently commercially available. ZAP-70, an enzyme normally expressed in T lymphocytes and critical for T-cell activation, is upregulated in CLL cells that contain unmutated IGH genes, conferring a poor prognosis (Chap. 92).⁷⁶ Patients whose CLL cells have mutated IGH and lack expression of ZAP-70 and CD38, a membrane protein with signaling activity have the longest treatment-free period after initial diagnosis.⁷⁷

T-cell CLL and large granular lymphocytic leukemia are uncommon disorders in which the malignant lymphocytes have a T-cell immunophenotype. Rearrangements involving band 14q11.2 with or without an accompanying break in 14q32.1 have been reported in T-CLL as well as in T-cell lymphomas (see Table 13–4).⁷² The most common is inv(14)(q11.2q32.1).

LYMPHOMA

Cytogenetic analyses of patients with lymphoma have demonstrated that more than 90 percent of cases are characterized by clonal chromosomal abnormalities and, more importantly, many of the recurring abnormalities correlate with histology and immunophenotype (see Table 13–4; Chaps. 95 and 96).⁷⁸ For example, the t(14;18) is observed in a high proportion of follicular small cleaved cell lymphomas (70 to 90 percent), most patients with a t(3;22)(q27;q11.2) or t(3;14)(q27;q32.3) have DLBCL, and patients with a t(8;14)(q24.2;q32.3) have either small noncleaved cell or DLBCL (Chap. 98). Band 14q32.3, the location of IGH is frequently involved in translocations in B-cell neoplasms (approximately 70 percent). In contrast, a large proportion of T-cell neoplasms are characterized by rearrangements that involve 14q11.2, 7q34, or 7p14, the locations of the T-cell receptor genes (Chap. 104). Gene expression profiling has proven useful in distinguishing unique genetic subtypes of lymphoma.⁷⁹

The t(8;14) is characteristic of both endemic and nonendemic Burkitt tumors, as well as Epstein-Barr virus (EBV)–negative and EBV-positive tumors (see Fig. 13–5; Chap. 102). Moreover, the t(8;14) has also been observed in other lymphomas, particularly small noncleaved cell (non-Burkitt) and large cell immunoblastic lymphomas, HIV-associated Burkitt lymphoma (100 percent) and HIV-related DLBCL (30 percent).⁸⁰ Two other variant translocations also occur in Burkitt lymphoma, t(2;8)(p12;q24.2) and t(8;22)(q24.2;q11.2). All three translocations involve chromosome band 8q24.2. These same translocations have been seen in some patients with B-cell ALL. The t(8;14) involves a break within the IGH locus on chromosome 14, and a break either 5′ or within the MYC gene on chromosome 8, and relocates the MYC coding exons to chromosome 14. MYC is a transcription factor that plays a critical role in a number of cellular processes including DNA replication, proliferation, and apoptosis; its oncogenic properties are a result of its constitutive expression.

Between 70 and 90 percent of follicular lymphomas (Chap. 99) and 20 percent of DLBCL have the t(14;18) (see Fig. 13–6), in which the BCL2 gene at 18q21.3 is juxtaposed to the IGH J segment, leading to the deregulated expression of BCL2.⁸¹ Common secondary abnormalities include −7, +18, and del(6q). Other malignancies that overexpress BCL2, but do not harbor the t(14;18), include hairy cell leukemia and CLL. The BCL2 gene encodes a 26-kDa mitochondrial membrane protein that functions to increase cell survival through antiapoptosis and preventing programmed cell death.

The t(11;14) (q13.3;q32.3) is observed in virtually all cases of mantle cell lymphoma (Chap. 100), 3 percent of myeloma (Chap 107), and up to 20 percent of prolymphocytic leukemias (Chap. 92).^82,83 Many cases also have deletions or point mutations of the ATM gene (11q22.3). Mantle cell lymphomas are currently regarded as a poor prognostic group with a median survival from diagnosis of 3 years. This translocation results in the activation of the cyclin D1 (CCND1) gene by the IGH gene (J region).⁸² The CCND1 gene is located 100 to 130 kb away from the breakpoint on 11q13.3. The D-type cyclins act as growth factor sensors, causing cells to go through the restriction start point of the cell cycle at G₁ and committing them to divide via phosphorylation and inactivation of RB1.

The BCL6 gene was cloned from the recurring breakpoint at 3q27 in cells characterized by a t(3;22)(q27;q11.2), t(3;14)(q27;q32.3) or, rarely, t(2;3)(p12;q27).⁷⁸ BCL6 rearrangements occur in 40 percent of DLBCLs and, in some series, up to 10 percent of follicular lymphomas. The translocations lead to the truncation of the BCL6 gene within the first exon or the first intron, substitution of its promoter sequences with an IG promoter, and deregulated expression. The BCL6 gene product is a 96-kDa POZ/Zn finger, nuclear protein that acts as a potent transcriptional repressor. It is predominantly expressed in the B-cell lineage, particularly in mature B cells, and may suppress genes involved in lymphocyte activation, differentiation, cell cycle arrest, and apoptosis. Somatic mutations have been identified in the 5′ regulatory regions of BCL6 in approximately 20 percent of DLBCLs without translocations leading to deregulation of BCL6, suggesting that overexpression of BCL6 is more broadly involved than initially recognized.⁸⁴

Extranodal marginal zone B-cell lymphomas of mucosa-associated lymphoid tissue (MALT lymphoma) are comprised of several genetic subgroups, one characterized by trisomy 3 plus other abnormalities (60 percent), and another by the t(11;18)(q21.2;q21.3)(25 to 50 percent) and its variants (Chap. 101).⁸⁵ Of note is that the t(11;18) is not observed in primary large B-cell gastric lymphoma. The t(11;18) results in the fusion of the apoptosis-inhibitor gene BIRC3 (API2), to a novel gene at 18q21.3, MALT1, whose product activates the nuclear factor κB (NFκB) pathway.

A number of recurring chromosomal abnormalities have been recognized in T-cell leukemias and lymphomas (see Table 13–4; Chap. 104). Similar to B-cell neoplasms, in which rearrangements frequently involve the chromosomal bands containing the immunoglobulin gene loci, T-cell neoplasms often have rearrangements involving band 14q11.2, the site of the T-cell receptor α-chain (TRA) and δ-chain (TRD) genes or, less often, one of two regions of chromosome 7 (7q34 and 7p14) to which the T-cell receptor β-chain (TRB) and γ-chain (TRG) genes have been localized, respectively.⁷⁸ These translocations result from aberrant V-D-J recombination events. With few exceptions, the involved gene on the partner chromosome encodes a transcription factor, whose expression is deregulated or activated as a result of the rearrangement (see Table 13–4). As a consequence of a chromosomal rearrangement that brings an oncogene under the controlling influence of promoters and enhancers that are active in T-cell receptor synthesis, T-cells may gain a proliferative advantage, resulting in malignant clonal expansion.

A distinctive subtype of lymphoma, namely, anaplastic large cell lymphoma (ALCL) is characterized by a young age at presentation, and skin and/or lymph node infiltration by large, often bizarre lymphoma cells, which preferentially involve the paracortical areas and lymph node sinuses (Chap. 98). The majority of such tumors express one or more T-cell antigens, a minority express B-cell antigens, and some express both T- and B-cell antigens (the null phenotype). A reciprocal translocation, t(2;5)(p23.2;q35.1), t(1;2)(q25;p23), or variant rearrangement involving the ALK tyrosine kinase gene at 2p23.2 appears to be restricted to ALCL of either T-cell or null phenotype, and is present in a high percentage of these cases.⁸⁶ The tumor cells are positive for CD30 on the cell membrane and in the Golgi region, and ALK expression is detectable in 60 to 85 percent of cases, where it confers a more favorable outcome (5-year survival, 80 percent in ALK+ vs. 40 percent in ALK− tumors). The t(2;5) has also been found in CD30+ primary cutaneous lymphomas.

MYELOMA

As in CLL, the application of molecular cytogenetic tools, such as FISH, has led to the discovery of numerous chromosomal abnormalities in myeloma, its precursor essential monoclonal gammopathy (Chaps. 105 and 106), and plasma cell leukemia (Chap. 107).^83,87 Monoclonal gammopathy is characterized by chromosomal aneuploidy, IGH translocations (45 percent of patients), and deletions of 13q (15 to 50 percent). Plasma cell myeloma is a malignancy of postfollicular B cells and is characterized by the acquisition of complex chromosomal rearrangements. As in monoclonal gammopathy, the earliest changes involve deletions of 13q14, and translocations of the IGH gene, which deregulate the expression of oncogenes located near the translocation breakpoints. Loss of chromosome 13 or a del(13q) is the most frequently observed chromosomal loss in myeloma and confers a poor prognosis.⁸³ With the use of FISH, deletions of 13q are detected in 40 to 50 percent of patients with myeloma and may be associated with specific 14q translocations.

Among the most frequent chromosomal rearrangements noted in plasma cell malignancies are translocations involving the IGH locus on 14q32. IGH translocations are detectable by interphase FISH analysis in approximately 10 percent of patients with monoclonal gammopathy, 40 to 50 percent of patients with myeloma, and more than 60 percent of patients with plasma cell leukemia.⁸³ The t(11;14)(q13.3;q32.2) is found in 15 percent of cases, and results in cyclin D1 overexpression and may deregulate expression of MYEOV (myeloma overexpressed gene). The t(4;14)(p16.3;q32.3) is noted in approximately 15 percent of patients and deregulates the expression of the fibroblast growth factor receptor 3 gene (FGFR3) translocated to the der(14), and the WHSC1/MMSET domain remaining on the der(4) chromosomes. The t(14;16)(q32.3;q23), noted in 5 percent of cases, results in the overexpression of the MAF transcription factor gene. Cyclin D3 overexpression occurs in the context of the t(6;14)(p21.1;q32.3), observed in 4 percent of patients. The translocation partners for the remaining 10 to 15 percent of myeloma cases are currently unknown. The t(4;14) and t(14;16) are both associated with a poor clinical outcome, whereas the t(11;14) confers a favorable prognosis. Translocations involving unknown partners confer an intermediate prognosis.

Chromosome 1 abnormalities are prevalent in multiple myeloma, frequently resulting in both gain of 1q and loss of 1p, and are associated with a shorter survival.⁸⁸ Furthermore, gene expression profiling studies that identified a high-risk disease signature noted a significant enrichment of genes located on chromosome 1.⁸⁸ For this reason, it is now recommended that a comprehensive FISH testing panel for multiple myeloma include detection of chromosome 1 abnormalities, particularly using probes for 1q.

Additional events occur with disease progression, including mutations of NRAS and KRAS, MYC deregulation, and epigenetic alterations. Activating mutations of NRAS or KRAS have been identified in monoclonal gammopathy (approximately 5 percent), and at a higher frequency in myeloma (30 to 40 percent); but the frequency may be higher in patients who relapse (80 percent).⁸⁹ Several genes are silenced through aberrant promoter hypermethylation in both monoclonal gammopathy and myeloma, including DAPK1 (67 percent), SOCS1, CDKN2B (p15), and CDKN2A (p16).⁸³