Hematopathology

B-lymphoblastic leukemia/lymphoma

B-lymphoblastic leukemia with t(v;11q23); KMT2A (previously known as mixed-lineage leukemia[ MLL ]) rearranged is characterized by a translocation between the lysine methyltransferase 2A ( KMT2A ) gene on 11q23 and any one of a large number of potential fusion partners, the most common of which is AFF1 on 4q21. These translocations confer high-risk disease due to aberrant regulation of the homeotic regulator KMT2A . Rearrangements of KMT2A represent the most likely recurrent genetic lesion in infantile B-lymphoblastic leukemia and may transpire in utero. Patients are typically younger than 2 years of age and characteristically present with marked leukocytosis and central nervous system involvement. KMT2A rearrangements in lymphoblastic leukemia are typically detected by FISH studies using a break apart probe spanning the 11q23 region.

B-lymphoblastic leukemia with t(9;22)(q34;q11.2) is a high-risk disease characterized by production of a fusion protein with constitutively active ABL1 tyrosine kinase activity; BCR-ABL1 (the Philadelphia chromosome). This is the most common recurrent genetic abnormality in adult B-lymphoblastic leukemia patients and confers poor prognosis among all age groups. Patients may benefit from tyrosine kinase inhibitor (TKI) therapy in addition to traditional high-dose chemotherapy. While t(9;22)(q34;q11.2) is reliably detected by FISH, the BCR-ABL1 breakpoint status should be confirmed at diagnosis by quantitative reverse transcription-PCR for purposes of ongoing monitoring during treatment and minimal residual disease (MRD) detection. The p190-kDa fusion protein (e1a2 transcript) predominates in pediatric disease while adults may demonstrate either the p190-kDa form or the larger p210-kDa form (e13a2 [b2a2] or e14a2 [b3a2] transcript). The latter is also seen in almost all cases of chronic myelogenous leukemia (CML). BCR-ABL1 fusion transcripts are described in more detail later in this section.

The balanced cryptic translocation t(12;21)(p13;q22); ETV6-RUNX1 can be detected by FISH and DNA or RNA fusion analysis techniques and is the most common recurrent abnormality in childhood B-ALL, accounting for 25% of cases. Its incidence decreases with age and ETV6-RUNX1 is only rarely seen in adult B-ALL. The event appears to occur early in leukemogenesis, and results in an abnormal fusion of the transcription factor ETV6 and the DNA-binding domain of the core binding factor subunit RUNX1 . The fusion protein interferes with the normal function of RUNX1, a factor that is critical for hematopoietic cell differentiation. Blasts in these cases often show aberrant expression of myeloid antigens such as CD13. Patients with B-ALL and ETV6-RUNX1 , especially if unfavorable risk factors are absent, have an excellent prognosis and are cured in greater than 90% of cases.

In contrast to mature B-cell malignancies, B-ALL rarely demonstrates translocations involving immunoglobulin loci. An exception is B-ALL with t(5;14)(q31;q32); IL3-IGH . The cytokine interleukin-3 (IL-3) is constitutively overexpressed as a result of being brought under the control of the IGH enhancer. One consequence is variable secondary eosinophilia, potentially leading to end-organ damage to sensitive tissues such as cardiac muscle. Otherwise, the clinical characteristics and prognosis associated with this rare disease are similar to B-ALL in general.

B-lymphoblastic leukemia with t(1;19)(q23;p13.3); TCF3-PBX1 accounts for approximately 6% of cases in children and less than 5% in adults. The genetic lesion is a translocation between the transcription factors TCF3 and PBX1 resulting in creation of a leukemogenic fusion gene in which the DNA binding domain of TCF3 is replaced with that of PBX1 . This results in constitutive transcriptional activity of genes regulated by the PBX protein family. The prognosis of TCF3-PBX1 positive B-ALL is similar to other subtypes with comparable risk factors.

BCR-ABL1– like B-lymphoblastic leukemia is defined by a gene expression profile similar to that seen in B-lymphoblastic leukemia with t(9;22) but lacking the Philadelphia chromosome. It accounts for 10 to 20% of pediatric and 20 to 30% of adult ALL, with increased incidence in Hispanic and Native American populations and is often referred to as Ph-like ALL. This disease is genetically heterogeneous and includes over 60 different rearrangements and gene mutations resulting in activation of kinase and cytokine receptor signaling. Common translocations involve CRLF2, EPOR, ABL2, PDGFRB, NTRK3, TYK2, CSF1R, JAK2, and ABL1 rearrangements with non- BCR partners. These alterations may be amenable to treatment with targeted therapies including TKIs and JAK2 inhibitors, ameliorating the typical poor prognosis associated with this disease. The majority of these abnormalities are not detected by conventional karyotyping and require additional testing such as FISH panels, an RT-PCR fusion panel, or RNA sequencing for diagnosis. Gene expression profiling or low-density arrays can identify these cases based on expression signatures.

Numerical chromosomal abnormalities are relatively common in B-lymphoblastic leukemia and define additional subtypes of the disease. Precise definitions vary, but generally cases with greater than 50 chromosomes are referred to as hyperdiploid, while those with less than 45 chromosomes are designated hypodiploid. Concurrent structural abnormalities may be encountered but are uncommon. B-ALL with hyperdiploidy is seen in 25% of childhood disease and decreases in incidence with age. The prognosis is very good and achievement of cure is highly likely with standard therapy. The number of chromosomes may be less important than the specific chromosomes involved. Patients with three copies of chromosomes 4, 10, and 17, so-called “triple-trisomy ALL,” have an excellent prognosis. Hypodiploid B-ALL is seen in less than 5% of patients overall. Prognosis for hypodiploid B-ALL is generally poor and appears to be correlated to the degree of chromosome loss, with near diploid (44 chromosomes) and high hypodiploid (40 to 43 chromosomes) patients faring relatively well. Low hypodiploid (32 to 39 chromosomes) cases show a high frequency of TP53 mutations (often germline) along with IKZF2 and RB1 abnormalities. Near haploid (24 to 31 chromosomes) B-ALL is associated with RAS mutations and other alterations targeting receptor tyrosine kinase signaling.

Intrachromosomal amplification of chromosome 21 (iAMP21) occurs as a primary genetic event in approximately 2% of childhood B-lymphoblastic leukemia and arises from the effect of “gene dosage” due to copy-number alterations of potentially hundreds of linked genes. iAMP21 in B-ALL is associated with older age at presentation (median age = 9 years) and with low white blood cell counts. FISH testing with probes to the RUNX1 locus demonstrates the presence of five or more signals on a single chromosome. Recent data show that B-ALL patients with iAMP21 fare poorly when treated according to standard risk protocols and benefit from high-risk ALL therapy.

T-lymphoblastic leukemia/lymphoma

The most frequent nonrandom cytogenetic abnormalities seen in T-lymphoblastic leukemia/lymphoma are translocations involving T-cell receptor (TCR) loci on chromosomes 7 ( TRA and TRD ) and 14 ( TRB and TRG ) in association with various partner genes. Many of these rearrangements involve dysregulation of T cell–specific cellular transcription factors, resulting in disruption of normal maturation or uncontrolled cellular proliferation. One frequent genetic target is TAL1 , which is rearranged in approximately 60% of T-ALL. Other genes implicated in T-ALL–related translocations include TLX1 , TLX3 , MYC , LMO1 , LMO2 , and LYL1 . The prognostic significance of these rearrangements is generally not well established, but TLX1 abnormalities appear to correlate with improved clinical outcome, while LYL1 and TAL1 rearrangements appear to be less favorable. The Philadelphia chromosome, t(9;22)(q34;q11.2), is rarely detected in T-ALL but confers a poor prognosis. Chromosomal deletions also occur in T-ALL, the most important of which involves the short arm of chromosome 9 and is seen in approximately 30% of cases. This results in loss of the important tumor suppressor gene CDKN2A and leads to a loss of normal cell cycle control. Notably, cryptic abnormalities (not identifiable by routine karyotype analysis) are frequently found in T-ALL. Activating mutations in NOTCH1 , which encodes a protein critical for early T-cell development, have been demonstrated in the majority of T-ALL patients and appear to be associated with a favorable outcome. Increased NOTCH1 signaling can be due to a point mutation, insertion, deletion, or rarely, translocation, and appears to play a central role in leukemogenesis. Additionally, mutations in the NOTCH1 regulator FBXW7 can lead to increased accumulation of the NOTCH1 protein.

Burkitt lymphoma

Burkitt lymphoma is an aggressive mature B-cell lymphoma. It is divided into three subtypes: endemic, associated with EBV infection and occurring primarily in young children in equatorial Africa; sporadic, associated with EBV in a third of cases and occurring in children and young adults; and immunodeficiency associated, often seen with HIV and EBV infection. The majority of cases present as an extranodal mass, often with bulky disease. Leukemic presentation is seen in rare cases, with only bone marrow and blood involvement. Burkitt lymphoma is characterized by monomorphic, intermediate sized B cells with basophilic, vacuolated cytoplasm. Tissue sections show a high mitotic rate and increased tingible body macrophages. The classic immunophenotype is positive for CD20, CD10, and BCL6, with expression of MYC and near 100% expression of Ki67. BCL2 is negative. There is evidence of EBV infection in the majority of cases, particularly in “endemic” Burkitt lymphoma and immunodeficiency associated cases. While the majority of cases show classic morphology, cases with mildly atypical appearance but the classic immunophenotype and MYC translocation may still be diagnosed as Burkitt lymphoma.

Burkitt lymphoma was the first lymphoma shown to harbor a recurrent genetic abnormality, and in many ways serves as an archetype for the study of mature B-cell non-Hodgkin lymphomas. Translocations involving the immunoglobulin locus and the MYC oncogene on 8q24 are primary genetic events in most cases of Burkitt lymphoma. The typical finding is t(8;14)(q24;q32), resulting in formation of a derivative chromosome 14 in which MYC is brought under control of the IGH locus, leading to constitutive MYC expression. Variant translocations involving MYC and the immunoglobulin light chain loci at 2p12 ( IGK ) and 22q11 ( IGL ) are seen in 5 to 10% of cases. MYC overexpression contributes to genomic instability through various mechanisms, including disruption of DNA double-strand break repair pathways and both activation and repression of transcriptional activity. It should also be noted that IG-MYC translocations are not specific for Burkitt lymphoma and occur as secondary events in a minor subset of other aggressive B-cell non-Hodgkin lymphomas and rarely in plasma cell myeloma and B lymphoblastic leukemia. Burkitt lymphoma is most reliably defined by gene expression profiling with a molecular signature distinct from diffuse large B-cell lymphoma (DLBCL), but this is unavailable in most clinical settings.

Rare cases with the morphologic and immunophenotypic features of Burkitt lymphoma that lack MYC rearrangement but show 11q aberrations should be diagnosed as Burkitt-like lymphoma with 11q aberration. Common abnormalities include interstitial gains in the 11q23.2–23.3 region and deletions telomeric to 11q24.1. These cases are more common in post-transplant patients.

Follicular lymphoma

Follicular lymphoma is a mature small B-cell lymphoma composed of germinal center B cells. The majority of cases are lymph node based and show a follicular growth pattern, with closely packed, uniform follicles that typically lack polarization and tingible body macrophages and have attenuated mantle zones. A smaller proportion of cases show a purely diffuse growth pattern or a mixed follicular and diffuse pattern. The cells typically express pan B-cell markers along with CD10. The neoplastic follicles show aberrant expression of BCL2.

A translocation involving IGH at 14q32 and BCL2 on 18q21 is present in up to 90% of cases of follicular lymphoma. BCL2 is thereby placed under the influence of the IGH promoter and the result is overexpression of the anti-apoptotic bcl-2. IGH-BCL2 translocations are also found in approximately 20 to 30% of cases of DLBCL. Surprisingly, t(14;18) can sometimes be detected in histologically benign lymph nodes or tonsillar tissue (usually at low levels) in otherwise normal individuals and is therefore not definitively diagnostic of lymphoma in isolation. Additional genetic aberrations are present in most cases of follicular lymphoma, including a variety of chromosomal gains and losses. Notably, 10 to 15% of follicular lymphoma, particularly high-grade cases, lack t(14;18). Approximately 80 to 90% of the breakpoints on 18q21 are located either in the major breakpoint region (MBR) found within the 3′ untranslated region of exon 3 of BCL2 or in the minor cluster region (MCR) found 3′ further downstream of exon 3 and are amenable to detection by multiplex PCR using consensus primer sets ( Fig. 70.1 ). Less common breakpoints (10%), many of which are found upstream of exon 1 of BCL2 , are not targeted by typical PCR assays. For this reason, FISH represents an attractive testing modality when evidence of t(14;18) is sought. BCL6 translocations are seen in 15% of follicular lymphomas and, along with MYC translocations, are associated with transformation to DLBCL.

Diffuse large B-cell lymphoma

DLBCL is a morphologically and genetically heterogeneous group of large B-cell lymphomas. DLBCL is more common in the elderly but can be seen in any age group. This disease classification includes both de novo lymphomas and transformation of low-grade lymphomas. DLBCL can be divided into two molecular subtypes: activated B-cell subtype and germinal center B-cell subtype, based on gene expression profiling. Up to 10% of cases do not fit either of these groups. Because gene expression profiling is not widely available for routine clinical use, immunohistochemical algorithms have been developed to assist in classification. Current treatment algorithms recommend more aggressive therapy in Activated B-Cell (ABC) subtype DLBCL.

DLBCL display a variety of underlying chromosomal abnormalities including t(14;18)(q32;q21). In addition to IGH-BCL2 rearrangement, which may signify evolution from preexisting follicular lymphoma, rearrangements of the transcriptional regulator BCL6 on 3q27 are common. The prevalence of BCL6 somatic mutations and translocations is explained by the fact that BCL6 is one of several other genes expressed in the germinal center that are known to normally undergo the process of somatic hypermutation, as does the variable region of IGH . MYC rearrangements are also present in approximately 10% of patients and are often seen in conjunction with a complex karyotype. Large cell lymphomas with non–Burkitt-like morphology and isolated MYC rearrangement should be diagnosed as DLBCL. These cases show inferior survival. Cases with translocations of MYC and BCL2 and/or BCL6 should be diagnosed as high-grade B-cell lymphoma (HGBL) with MYC and BCL2 and/or BCL6 rearrangements rather than DLBCL.

High-grade B-cell lymphoma

A subset of HGBL patients with variable histology harbor concurrent rearrangements of MYC in combination with BCL2 or BCL6 , a phenomenon commonly referred to as “double-hit” lymphoma. Rarely all three abnormalities may be observed, in which case the designation “triple-hit” B-cell lymphoma may be used. Cases of mature B-cell lymphomas with these genetic alterations should be diagnosed as HGBL with MYC and BCL2 and BCL6 rearrangements regardless of morphologic or immunophenotypic findings. The clinical phenotype is particularly aggressive. Due to the aggressive nature of the disease and a propensity for involvement of the central nervous system, prompt recognition of these patients is warranted for purposes of therapeutic decision making and prognostic stratification. This requires FISH testing at the time of initial diagnosis.

Mantle cell lymphoma

Mantle cell lymphoma is a mature small B-cell lymphoma that is typically associated with an aggressive behavior, although indolent variants do exist. There are a number of morphologic subtypes, including blastoid, pleomorphic, and small cell. The neoplastic B cells are usually positive for CD5, cyclin D1, and SOX11.

Mantle cell lymphoma is characterized by t(11;14)(q13;q32); IGH-CCND1 in nearly all cases. The juxtaposition of the CCND1 gene, encoding cyclin D1, with the IGH enhancer is a primary genetic event and results in increased progression through the cell cycle due to cyclin D1 protein overexpression. The latter is detectable by immunohistochemistry, which often renders direct molecular genetic demonstration of IGH-CCND1 unnecessary for diagnosis in most settings. However, clinical scenarios often arise in which mantle cell lymphoma is in the differential diagnosis but cyclin D1 immunohistochemical evaluation is not feasible, perhaps due to specimen limitations (e.g., peripheral blood). In these situations, alternative testing is required. FISH represents the most sensitive testing strategy and is capable of detecting nearly 100% of translocations, assuming that cells harboring the translocation are present above established sensitivity thresholds. Notably, t(11;14)(q13;q32) is also seen in 5 to 10% of plasma cell myeloma patients. Numerous breakpoints involving the CCND1 locus may be encountered in mantle cell lymphoma and span a 350 kilobase region on 11q13. Approximately 40% of these breakpoints are clustered in a 1 kilobase segment referred to as the major translocation cluster (MTC) found 110 kb downstream of the CCND1 locus ( Fig. 70.2 ). Most PCR-based assays do not interrogate CCND1 breakpoints outside the MTC. Translocations with breakpoints outside of this region are therefore not detected and the sensitivity of these assays is only 40 to 50%. Additional chromosomal abnormalities including loss of 1p, 13q, and 17p and gains of 3q are common. Rarely, t(8;14) is present and is associated with a worse prognosis. Cases without CCND1 translocation show translocations or mutations in other cell cycle regulators including CCND2 and CCND3.

Extranodal marginal zone lymphoma (MALT lymphoma)

Extranodal marginal zone lymphoma (MALT) is a small mature B-cell lymphoma, occasionally showing plasmacytic differentiation, occurring in sites including the stomach, eyes, skin, lungs, salivary glands, breasts, and thyroid. MALT lymphomas are often associated with chronic inflammation. For example, gastric MALT lymphomas are often due to chronic Helicobacter pylori infection. In early disease, eradication of the infection can induce remission of the lymphoma. Several recurrent translocations have been described in MALT lymphoma and their incidence varies with anatomic site of disease. The t(11;18)(q21;q21) occurs mostly in gastric and pulmonary MALT lymphoma and results in BIRC3-MALT1 fusion. Detection is important because patients with t(11;18) are less likely to respond to antibiotic therapy directed at H. pylori and only rarely progress to large cell lymphoma. Three additional translocations, t(14;18)(p14;q32), t(1;14)(p22;q32), t(3;14)(p22q32) are seen with relatively low incidence but reinforce the paradigm of proto-oncogenes ( MALT1 , BCL10 , and FOXP1 , respectively) under the control of the IGH enhancer complex in B-cell lymphoproliferative disorders.

Multiple myeloma

Multiple myeloma is a clinically and genetically heterogeneous clonal disorder of terminally differentiated plasma cells. Incorporation of clinicopathologic and radiographic findings, including bone marrow evaluation, is often necessary for definitive diagnosis. Karyotypic and/or FISH abnormalities are detectable in a large proportion of cases and include a variety of aberrations. Identification is important for prognostic stratification. The most common abnormalities appear to be early events and include either trisomies of various chromosomes resulting in hyperdiploidy or translocations involving the IGH locus at 14q32. IGH translocation partners include, in decreasing frequency, CCND1 (11q13), FGFR3/MMSET (4p16.3), C-MAF (16q23), CCND3 (6p21), and MAFB (20q11). Monosomy 13 or deletion of 13q14 is also observed with regularity in plasma cell myeloma (40%), particularly in its leukemic form (70%). Disease progression is associated with acquisition of various additional genetic abnormalities, perhaps the most important of which is deletion of the tumor suppressor TP53 at 17p13. TP53 deletion signals high-risk disease with significantly decreased overall survival and additionally may serve as a marker of extramedullary involvement. High-risk abnormalities include t(4;14), t(14;16), t(14;20), del 17p, TP53 mutation, and gain of 1q. All other genetic abnormalities are considered standard risk.

Anaplastic large cell lymphoma

Examples of recurrent chromosomal abnormalities are less common in mature T-cell lymphomas. Anaplastic lymphoma kinase (ALK)-positive anaplastic large cell lymphoma (ALCL), the hallmark of which is translocation of the tyrosine kinase receptor ALK , is one exception. ALCL is an aggressive T cell lymphoma derived from cytotoxic T-cells. Typically, these lymphomas feature characteristic “hallmark cells,” large cells with eccentric, wreath-like or horseshoe-shaped nuclei. ALCL stains with CD30 and ALK and shows variable expression of other T cell–associated antigens. CD3 is often negative but the majority of cases stain with CD2 and CD5, along with cytotoxic antigens. ALCL is most often CD4 positive, with rare cases positive for CD8.

The most common translocation, t(2;5)(p23;q35.1), accounts for 85% of cases and involves fusion of ALK to nucleophosmin ( NPM1 ), resulting in nuclear and cytoplasmic expression of ALK. Numerous less frequent variant translocation partners have also been described. ALK rearrangements are not disease-specific and are seen in a subset of non–small-cell lung cancers. NPM1-ALK and variant fusions can be detected by FISH and other means but this is usually unnecessary since normal postnatal human tissues, except for rare central nervous system constituents, lack ALK expression. Therefore the presence of an ALK translocation can be inferred quickly and cost effectively by the immunohistochemical detection of ALK protein expression in the neoplastic cells. Notably, ALK inhibitor therapies are in development, and treatment with these agents may lead to the acquisition of ALK kinase domain activating mutations conferring drug resistance.

ALK-negative ALCL is a T-cell lymphoma that shares essentially identical morphologic and immunohistochemical features with ALK-positive ALCL but occurs in older patients and lacks both ALK rearrangements and ALK protein expression. ALK-negative ALCL has a worse prognosis relative to ALK-positive ALCL, possibly justifying a more aggressive therapeutic approach. Recent studies have revealed genetic heterogeneity in these cases. Patients with ALK-negative ALCL and rearrangement of DUSP22 on 6p25.3 (demonstrated by FISH) fare relatively well, with overall survival resembling ALK positive disease. Rearrangements of TP63 on 3q27 appear to be mutually exclusive with regard to DUSP22 and imply a very poor prognosis.

Acute myeloid leukemia

Core binding factor-associated AMLs are unified by molecular pathophysiology and include t(8;21)(q22;q22); RUNX1-RUNX1T1 and inv(16)(q13.1;q22) or t(16;16)(p13.1;q22); CBFB-MYH11 . The core binding factor is a heterodimeric protein composed of alpha and beta subunits encoded by RUNX1 and CBFB , respectively, which is normally involved in regulation of hematopoiesis. The RUNX1-RUNX1T1 and CBFB-MYH11 abnormalities disrupt this function through impairment of differentiation and contribute to leukemogenesis. A finding of either of these two AML subtypes is diagnostic of acute leukemia, irrespective of blast count. Patients with either of these two AML subtypes respond well to cytarabine-based consolidation chemotherapy. Importantly, prognosis is adversely affected by activating KIT mutations in exon 8 or 17, which are present in up to 30% of cases. FISH is preferred for detection at diagnosis, while quantitative reverse transcription PCR is highly sensitive and ideally suited for disease monitoring. MRD testing has a clear prognostic significance in the setting of core binding factor AML. RUNX1- RUNX1T1 fusion transcripts are readily detectable by reverse transcription-PCR (RT-PCR) utilizing relatively simple primer sets due to the clustering of breakpoints at a limited number of intronic sites. Three dominant CBFB-MYH11 fusion transcripts (types A, D, and E) corresponding to different breakpoints account for over 95% of inv(16) positive AML cases and are also amenable to detection by RT-PCR.

The presence of the balanced reciprocal translocation t(15;17)(q24;q12); PML-RARA is diagnostic of acute promyelocytic leukemia. The chimeric fusion protein resulting from the translocation mediates an arrest in myeloid differentiation at the promyelocyte stage. Timely recognition is required due to the high risk of disseminated intravascular coagulation (DIC) seen in these patients and the associated clinical ramifications. Prompt DIC prophylaxis may prevent a bleeding diathesis which otherwise could be fatal. Morphologic and flow cytometric studies allow for a presumptive diagnosis in most cases, but genetic confirmation of PML-RARA is nevertheless required. FISH is the test of choice at diagnosis due to high sensitivity and typically rapid turn-around time. Rare cryptic PML-RARA fusions have been documented, which are not detectable by routine karyotyping. The disease is responsive to therapy with all- trans -retinoic acid, which drives the terminal differentiation of the neoplastic cells and is used in combination with anthracyclines or arsenic trioxide to induce durable remission in 80 to 90% of patients. Variant RARA fusion partners may be seen on occasion, the most important of which is ZBTB16 at 11q23. Recognition is important due to the lack of therapeutic response to all- trans -retinoic acid. Assessment of response to therapy by a quantitative RT-PCR-based monitoring test has emerged as standard practice and appears to improve clinical outcome. Three PML-RARA fusion transcripts (bcr1, bcr2, and bcr3) may be encountered corresponding to breakpoints at different regions of PML at 15q24 ( Fig. 70.3 ). RARA breakpoints are clustered within intron 2 in a 15-kb region on 17q12. The bcr3 fusion results in a relatively short transcript and may be associated with the presence of FLT3 mutations.

Translocations involving the KMT2A gene and various partner genes occur with some frequency in AML cases, most commonly in children. Over 80 KMT2A translocation partners have been described. In AML, the most common is t(9;11)(p22;q23 );MLLT3-KMT2A , which is routinely detectable by FISH. This disease is often associated with extramedullary manifestations, and patients classically present with gingival hyperplasia or cutaneous lesions due to tissue infiltration by the leukemia cells. Due to the heterogeneity of the translocation breakpoints, a widely applicable quantitative RT-PCR assay for disease monitoring is difficult to design, but testing can be performed on those patients with more common breakpoints.

A subset of acute myeloid leukemia cases may arise from a preexisting MDS and therefore some genetic features common to both diseases are observed. Two examples are t(6;9)(p23q34); DEK-NUP214 and inv3(q21;q26.2) or t(3;3)(q21;q26.2); RPN1-EVI1. ^, Both are associated with multilineage dysplasia and occur in 1-2% of AML patients overall. As expected, prognosis is generally poor. FLT3 mutations are frequently detected in AML with DEK-NUP214 , which has a tendency to affect younger adults and children, and it is unclear whether the negative prognostic implication is independent of FLT3 status. FLT3 mutations are relatively uncommon in AML with inv3 or t(3;3) in which the RPN1 acts as an enhancer of the oncogene EVI1 to drive cellular proliferation possibly in collaboration with RAS -pathway mutations.

A particularly rare form of AML occurs in infants and children with Down syndrome and involves fusion of RBM15 to MKL1 as a result of t(1;22)(p13;q13). The fusion gene’s precise role in leukemogenesis is unclear but transcriptional activation and modulation of chromatin organization are likely mechanisms. Somatic N-terminal truncating GATA1 mutations are common in patients with Down syndrome associated AML but do not affect prognosis. GATA1 mutations are also seen in individuals with Down syndrome who develop transient abnormal myelopoiesis—a disorder which may mimic AML but usually resolves in the first 6 months of life. These two conditions may not be readily distinguishable, especially in early infancy, and careful clinical correlation is necessary.

Chronic myelogeneous leukemia

The discovery in 1960 of a recurrent chromosomal abnormality in patients with a disease then known as chronic granulocytic leukemia by Peter Nowell and David Hungerford was followed 13 years later by Janet Rowley’s demonstration of a consistent reciprocal translocation between 9q34 and 22q11.2, resulting in a der(22q) (now known as the Philadelphia chromosome) found in virtually all cases of CML. ^, Another 10 years would pass before the precise breakpoints were cloned, and it was shown that the translocation results in juxtaposition of the ABL1 proto-oncogene on 9q34 to BCR on 22q11.2 with formation of a novel BCR-ABL1 fusion gene. This results in constitutive ABL1 protein tyrosine kinase activity and dysregulated cellular proliferation. Documentation of t(9;22)(q34;q11.2); BCR-ABL1 fusion is necessary for diagnosis and is readily demonstrated by metaphase cytogenetics, FISH and/or RT-PCR.

Distinct BCR-ABL1 fusion transcripts, which correspond to variably sized fusion proteins, are encountered based on the translocation breakpoints ( Fig. 70.4 ). The ABL1 breakpoint is largely conserved at a location upstream of exon a2. The bulk of ABL1 is therefore fused to BCR thereby preserving the ABL1 kinase domain. The major breakpoint cluster region (M-bcr) is the location of the majority of BCR breakpoints in CML, and up to half of the BCR breakpoints in Philadelphia chromosome positive adult B-ALL. MBR breakpoints occur between exons 13 or 15 (b2 or b4), resulting in a transcript consisting of e13a2 (b2a2) or e14a2 (b3a2), both of which encode the p210-kDa BCR-ABL1 isoform typical of CML.

Less frequently, BCR breakpoints occur in the minor breakpoint cluster region (m-bcr) between BCR exon 1 and exon 2, resulting in a shorter e1a2 fusion transcript encoding the p190-kDa fusion protein seen rarely in CML but commonly in pediatric Philadelphia chromosome positive B-ALL. At initial diagnosis, CML patients may harbor low levels of e1a2 transcript detectable by RT-PCR in addition to the M-bcr transcript. This phenomenon is likely due to alternative splicing and it is important to note that for purposes of disease monitoring, only the M-bcr transcript should be followed by RT-PCR. Occasionally, e1a2 is the exclusive transcript detected in CML and is associated with monocytic proliferation resembling chronic myelomonocytic leukemia. Rare cases of CML harbor BCR breakpoints that occur in the mu breakpoint cluster region (μ-BCR) at exon 19 (c3), resulting in an e19a2 (c3a2) fusion. The corresponding BCR-ABL1 protein is larger (p230 kDa) and is characteristically seen in association with marked neutrophilic maturation that may mimic chronic neutrophilic leukemia. Notably, most RT-PCR assays are not designed to detect the e19a2 fusion transcript, so FISH studies should be recommended if clinical suspicion for CML persists despite a negative RT-PCR result.

Despite dramatic improvement in long-term survival, treatment of CML with TKIs does not typically result in cure. This is evidenced by the fact that low levels of BCR-ABL1 persist even in patients who achieve a major molecular response. Over time, point mutations in the BCR-ABL 1 kinase domain may develop in a subset of patients and confer resistance to the TKI. This resistance may be overcome by changing therapy to a different TKI. Documentation of a BCR-ABL1 kinase domain mutation in a patient with suboptimal or failed response to first-line TKI therapy may therefore be indicated and can be accomplished by direct (Sanger) sequencing and, increasingly, by massively parallel sequencing. The latter strategy may offer additional testing benefits. This is discussed in more detail in Section II of this chapter.

Myelodysplastic syndromes

The most common chromosomal abnormalities observed in MDS take the form of unbalanced structural chromosomal deletions or gains (see Table 70.2 ). Deletion of the long arm (q arm) or outright loss of chromosome 5 or 7 is seen relatively frequently. Balanced abnormalities are rare in MDS, but do occur, and show some overlap with recurrent translocations found in AML. These findings provide evidence of clonality which is particularly useful when the differential diagnosis in a cytopenic patient includes reactive conditions which must be excluded. Indeed several chromosomal aberrations can provide presumptive evidence of MDS in the setting of persistent cytopenias, even in the absence of sufficient morphologic evidence of dysplasia. In addition, many of these cytogenetic abnormalities have established prognostic significance and are an integral component of the widely applied International Prognostic Scoring System for MDS. For example, a complex cytogenetic profile (greater than three unrelated defects) is a harbinger of evolution to AML and portends a very poor prognosis. Copy number alterations or copy number neutral loss of heterozygosity (e.g., TP53 ) also occur in MDS and such abnormalities are detectable by array methods. ^,

Myeloid and lymphoid neoplasms with eosinophilia

A rare but distinct group of myeloid and lymphoid neoplasms demonstrate variable clinical presentations but are unified by eosinophilia and gene fusions involving the receptor protein tyrosine kinases PDGFRA , PDGFRB , or FGFR1 (see Table 70.2 ). Patients with PDGFRA associated disease most often present with a MPN resembling chronic eosinophilic syndrome but a range of presentations including acute myeloid leukemia and T-lymphoblastic leukemia/lymphoma are possible. A FIP1L1-PDGFRA fusion is formed as a result of a cryptic deletion at chromosome 4q12. FISH is well suited for the detection of this abnormality and probes targeting FIP1L1 , PDGFRA, and CHIC2 (also located at chromosome 4q12) are often employed. Fusion of the FIP1L1 and PDGFRA loci results in loss of the CHIC2 locus and this is detectable with the aforementioned FISH probes. An ETV6-PDGRFB fusion resulting from t(5;12)(q33;p12) is the most commonly observed fusion involving the PDGFRB gene and typically occurs in the context of a chronic myelomonocytic leukemia-like disorder, often with eosinophilia. PDGRFB -associated translocations are detectable by routine karyotype. Prompt recognition of these diseases is crucial since end-organ damage due to eosinophilia may result in significant morbidity. In addition, most patients with PDGFRA and PDGFRB abnormalities are highly sensitive to treatment with the TKI imatinib. FGFR1 -related myeloid and lymphoid neoplasms are characteristically heterogeneous and may manifest as chronic eosinophilic leukemia, acute myeloid leukemia, or T- or B- lymphoblastic leukemia/lymphoma, among other possibilities. Various fusion partners have been reported, notably including BCR , but rearrangement of FGFR1 at 8p11 is constant and readily detectable by karyotyping. This is a very poor prognosis disease which is unresponsive to TKI therapy. More recently, rare myeloid neoplasms with similar clinical and pathologic features harboring PCM1-JAK2 fusion caused by t(8;9)(p22;p24) and resulting in JAK2 activation have been described and are included in this category. JAK inhibitors such as ruxolitinib may provide short-term therapeutic responses in these patients, which may allow bridging to stem cell transplantation.

Cytogenetics

Conventional cytogenetics refers to visual analysis of a karyotype composed of a complete set of metaphase-arrested chromosomes typically stained by Giemsa (G-banding). G-banding highlights light and dark zones corresponding to A-T-rich and G-C-rich regions of chromosomes. This decades-old technique remains very important for the comprehensive bone marrow evaluation of patients with acute leukemia or an MDS. G-banded karyotyping enabled the discovery of many of the recurring translocations characteristic of specific hematologic diseases. Karyotyping provides a genome-scale perspective but is only well suited for uncovering large-scale chromosomal aberrations (resolution > 5 Mb) including reciprocal translocations, large deletions, and aneuploidy. Cryptic rearrangements and small insertions or deletions are likely to be missed by conventional karyotyping and require more sensitive techniques for detection. In addition, karyotyping is a labor-intensive process that requires considerable expertise for interpretation. Because cell growth in culture is required, turnaround times often range from 3 to 7 days, making this technique inappropriate for urgent clinical scenarios such as detection of t(15;17)(q24;q12); PML-RARA.

Fluorescence in situ hybridization

FISH continues to be a powerful tool in the modern cytogenetics laboratory due to its ability to overcome several of the limitations of conventional G-banded karyotypes. This technique utilizes fluorescently labeled DNA probes designed to hybridize to specific sequences on metaphase or interphase chromosomal preparations. Resolution is improved to approximately 2 Mb. FISH allows detection of cryptic chromosomal abnormalities such as translocations or deletions that may not be identifiable by routine karyotype. The proliferation of FISH testing has been facilitated by the commercial availability of extensive libraries of probes targeting various recurrent abnormalities. A variety of specimen types can be used, importantly including paraffin embedded formalin fixed samples, since FISH assays do not require viable cells or growth in culture. So-called “break apart” probes allow for identification of gene rearrangements without a priori knowledge of the translocation partner, of which there may be numerous possibilities. For example, an MLL break apart probe ( KMT2A gene ) spanning the appropriate 11q23 region can effectively confirm or exclude the presence of an MLL rearrangement without having to query any of the dozens of potential MLL translocation partners. Global assessment of copy number changes is not possible by FISH since only a limited number of probes are utilized.