Acute Leukemia Is The Commonest Malignancy In Children Biology Essay

Published: November 2, 2015 Words: 4849

Acute leukemia (AL) is the commonest malignancy in children less than 15-years of age {{664 Anonymous 2010}}. Approximately 3000 new cases of AL occur annually in the United States, of which 80% are acute lymphoblastic leukemia (ALL). The five-year survival rates for childhood AL, and especially ALL has dramatically improved from 61% in 1975-1978 to 89% in 1999-2002 {{663 Smith,M.A. 2010; 655 Pulte,D. 2008; 563 Johnston,W.T. 2010; }}. The remarkable success story of pediatric ALL is attributed to the exponential increase in knowledge of the molecular mechanisms of the disease and the impact of well-designed clinical trials adapted to risk -stratified subgroups based on prognostic indicators, including evaluation of early response to the treatment (minimal residual disease detection). This has been accomplished by genomic studies employing a host of modern techniques, for example, conventional cytogenetics, fluorescent in-situ hybridization (FISH), DNA and gene expression arrays, and proteomics. Many of these methodologies have moved from the research bench to clinical molecular diagnostics allowing their routine use in the diagnosis, classification, prognostication, and follow-up of acute leukemia. In this chapter we discuss the clinical features of acute leukemia and describe the evolution of the modern classification of acute leukemia as it reflects a transition from morphological to molecular genetic approaches. The subsequent chapter on AL in this volume describes molecular techniques routinely used in the diagnosis and prognostication of acute leukemia.

Clinical Features

AL is the commonest childhood malignancy with an age-adjusted incidence of 44.3 per 1,000,000 persons (ALL 36.8; AML 7.5){{484 Linabery,A.M. 2008; 664 Anonymous 2010}}. ALL is four times as common as AML, and occurs more often in male rather than female children {{484 Linabery,A.M. 2008; 664 Anonymous 2010}}. Since 80% of all childhood AL is ALL, the major discussion will be devoted to ALL, especially precursor B-ALL (80-85% of ALL). The applications of genomic studies to precursor B-ALL are similarly relevant to precursor T-ALL and AML; features specific to the latter will be highlighted separately. AL of infancy is biologically and clinically different from that in older children. Unlike older children, ALL and AML occur with nearly equal frequency in infancy, and there is a female rather than male preponderance {{1431 Schlis,K.D 2009; }}. The infant AL very often have high white blood cell counts and presence of the t(4;11)(q21;q23) putting them in the clinically high-risk category with poor outcome.

Clinical symptoms of AL are related to replacement of normal hematopoietic cells in the bone marrow by the leukemic cells, and to organ infiltration by blasts. Fever, pallor, weakness, bleeding manifestations, and bone pains are the most common presenting symptoms {{746 Chessells,J.M. 2002}}. Nearly 40% of affected children complain of joint aches, presumably from leukemic infiltration of the joint capsule {{1432 Sinigaglia, R. 2008;}}. Massive organomegaly and central nervous system (CNS) involvement at presentation are characteristic of AL in infancy {{1431 Schlis,K.D 2009; }}.. Extramedullary solid tumor masses or chloromas are seen at presentation in a small number of patients, mostly AML with monocytic differentiation; orbits, paranasal sinuses, and skin are the usual sites for chloromas. Isolated testicular masses are more a feature of relapse than primary presentation in acute leukemia. Nearly 10% children, almost always with precursor T-lymphoblastic leukemia/lymphoma, present with a life-threatening tracheobronchial or cardiovascular compression syndrome due the leukemic infiltration of the thymus and other mediastinal structures, requiring prompt intervention by systemic corticosteroids, or local radiation {{1430 Silverman,L.B. 2009}}.

Diagnosis and Classification of AL

In addition to establishing the diagnosis, the aim of physicians and laboratorians is to stratify patients according to the risk for optimal management. Risk-stratification is determined by 1) clinical features at presentation (age, white cell count, sex (in some protocols)), 2) cytogenetics (or molecular abnormalities), and 3) response to treatment (minimal residual disease). Laboratory methods are directed towards the latter two.

Morphology, immunophenotyping, and cytogenetic analysis form the cornerstone of diagnosis and risk-stratification in acute leukemia. The white cell count is usually high; counts exceeding 50,000/µL denote high-risk and are seen in 20% of the children at presentation, more often in infants. About 10% children present with hyperleukocytosis (leucocyte count >100,000/ µL ){{1433 Majhail,N.S. 20041434 Lowe,E.J. 2005}}. Circulating blasts are present in the peripheral blood in most patients, even those with normal white cell counts, although about 1%, may have an aleukemic or pancytopenic presentation. Anemia and thrombocytopenia may be mild or critically low.

Morphology

The diagnosis of AL based on cytomorphology and cytochemistry has undergone a drastic change in the last three decades. The initial French-American -British (FAB) classification of ALL {{68 Bennett,J.M. 1976 }} based on nuclear morphology (nuclear heterogeneity, contour, nucleoli), or the FAB classification for AML{{68 Bennett,J.M. 1976}} based on morphology, degree of maturation, and cytochemistry, are insufficient for prognostication and risk-stratification {{1317 Vardiman, J.W. 2009; 1318 Vardiman,J.W. 2002; 520 Brunning,R.D. 2003; }}. Still, morphological assessment is the first step towards diagnosis and guides subsequent investigation. In many instances, morphology provides a clue for the underlying genetic abnormality, for example blasts with cytoplasmic and nuclear vacuoles (Burkitt-like) are consistent with mature B-ALL and t(8;14), and abnormal heavily granulated promyelocytes are consistent with acute promyelocytic leukemia (APL) and t(15;17) (Figure 2-1A and 2-1B). Bone marrow evaluation is useful for diagnosis when the presentation is aleukemia or subleukemic. Bone marrow aspiration is however necessary in the initial estimation of cellularity and leukemic burden in order to obtain sample for karyotyping and molecular studies, to assess response to induction chemotherapy, and for predicting outcome based on early response to chemotherapy thereby predicting likelihood of relapse. Extensive necrosis in the bone marrow at presentation is not common {{1900 Invernizzi, R. 1995; 1901 Inoue, S. 2007; }}, but when necrosis is present it can make the morphological and immunophenotypic characterization very challenging.

Immunophenotype

The immunological classification of AL introduced in 1985 by the European Group on Immunological Classification of Leukemia (EGIL) was based on use of antigen panels for lineage determination and the hierarchical sequence of antigen expression by cells {{518 Bene,M.C. 1995 }}. The 2008 WHO classification system has proposed a simpler algorithm that relies on fewer markers for lineage determination and categorization as mixed phenotype acute leukemia (MPAL). Table 1 (adapted from the 2008 WHO classification), shows the critical antigens required for lineage delineation as myeloid, precursor B, precursor T-lymphoid, or MPAL.

INSERT TABLE 1

B-Acute Lymphoblastic Leukemia

Based on the presence of antigens that reflect degree of differentiation, precursor B-ALL has been subclassified as early or pro-B-ALL (TdT+, CD10-, CD19+, CD22+/-), intermediate stage or common ALL with expression of CD10 (CALLA), and pre-B-ALL (positive for cytoplasmic immunoglobulin (cIg)). Historically, the immunophenotype was used to predict outcome in ALL, for example the better outcome in ALL with expression of CALLA (CD10), and worse outcomes in CALLA (-) ALL, or ALL with a cIg + phenotype {{1324 Crist,W.M. 1990}}. Correlation of simple and complex phenotypes determined by multiparameter flow cytometry (MFC) with genotypes is well described {{1438 Paietta,E. 20041437 De,J. 20071435 Hrusak,O. 20021436 Kozlov,I. 2005}}. Later studies demonstrated that the phenotypes, and hence the outcomes, were highly associated with specific chromosomal anomalies {{829 Borowitz,M.J. 1990; 1321 Pui, C.H. 1990; }}. A higher representation of hyperdiploidy (>50 chromosomes) in the CD10 (CALLA)+ pre-B ALL, and of t(4;11) in the CD10 (-) ALL {{1332 Pui,C.H. 1991; 1329 Pui,C.H. 1998; }} were correlated with the observed favorable and unfavorable outcomes in these phenotypes, respectively. Likewise the presumed unfavorable outcome associated with co-expression of myeloid antigens in newly diagnosed ALL could be ascribed to presence of the associated MLL (11q23) and BCR-ABL1 [t(9;22)] translocations {{690 Seegmiller,A.C. 2009; 719 Khalidi,H.S. 1999; 1326 Vitale,A. 2007; 1332 Pui,C.H. 1991; }}. The early reports on the unfavorable outcome in ALL expressing cIg {{1324 Crist,W.M. 1990}} found a high correlation with t(1;19)(q23;p13) translocation, 90-95% of which have this phenotype {{829 Borowitz,M.J. 1990; 1321 Pui, C.H. 1990}}. In subsequent studies several authors showed that with intensive chemotherapy programs the negative impact of t(1;19) could be offset {{1322 Raimondi,S.C. 1990; 1323 Pui,C.H. 1994; }}. More recently, gene expression studies (GEP) have validated specific immunophenotypes predictive for different leukemia subtypes {{ 763 Haferlach,T. 2010; 767 Buldini,B. 2010; 1901 Basso, G. 2007; }}. MFC can suggest the underlying genetic aberration and guide the choice of appropriate panels for FISH studies if conventional cytogenetics is non-contributory (Tables 2 and 3).

INSERT TABLE 2

Genetic abnormalities

ALL is genetically heterogeneous. Multivariate analyses in several large clinical studies have clearly established that genetic abnormalities are the most important determinants of response to chemotherapy and outcome in ALL {{819 Sutcliffe,M.J. 2005;749 Schultz,K.R. 2007;732 Moorman,A.V. 2010;737 Moorman,A.V. 2007742 Moorman,A.V. 2003;553 Harrison,C.J. 2010;1285 Harrison,C.J. 2001;1302 Harrison,C.J. 2009;1302 Harrison,C.J. 2009;1291 Forestier,E. 2000;1291 Forestier,E. 2000;1188 Forestier,E. 2003;751 De Braekeleer,E. 2010}} . Their relevance is likely to increase as targeted therapies are introduced. The genetic abnormalities, which currently have the most significant impact on treatment and management are t(9;22)(q34;q11)/BCR-ABL1, t(4;11)(q21;q23)/MLL- AFF1, and near-haploid/low hypodiploidy that are poor prognostic markers, and to a lesser extent, t(12;21)(p13;q22)/ETV6- RUNX1 and high hyperdiploid that are favorable prognostic markers {{553 Harrison,C.J. 2010; 1303 Konn,Z.J. 2009; 737 Moorman,A.V. 2007; 732 Moorman,A.V. 2010; 1186 Harrison,C.J. 2010}} . The genetic abnormalities that are considered mandatory in the evaluation of ALL, and for which clinical testing is available, are shown in Table 2 along with their prognostic implication and the various genetic techniques for their determination. These and other novel genetic aberrations are discussed below.

Numerical chromosomal aberrations

The clinically significant numerical aberrations seen in ALL are high-hyperdiploidy (HeH) (51-65 chromosomes), near-haploidy (24-29 chromosomes), and near-diploidy (31-39 chromosomes) {{736 Moorman,A.V. 2007; 585 Kowalczyk,J.R. 2010; 585 Kowalczyk,J.R. 2010; 553 Harrison,C.J. 2010; 751 De Braekeleer,E. 2010 }}. HeH, which is present in 23-42% of newly diagnosed precursor B-ALL, results from gain of specific non-random chromosomes (4, 6, 10, 14, 17, 18, 20, 21, and X) and is grouped as those with 47to 50 or more than 50 chromosomes {{823 Ito,C. 1999; 819 Sutcliffe,M.J. 2005; 742 Moorman,A.V. 2003; }} . HeH with 51-65 chromosomes is considered a favorable risk factor when associated with other types of chromosomal gains, such as trisomy of chromosomes 4, 10, and 17 (Figure 2-2) {{742 Moorman,A.V. 2003; }} . HeH cells are more sensitive to apoptosis with certain drugs (especially methotrexate, mercaptopurine, and l-asparaginase) than non-hyperdiploid cells {{1284 Whitehead, V.M. 1998; }}. Low hypodiploidy, and near hypoploidy have a poorer outcome compared to near diploid chromosomes {{1067 Heerema,N.A. 1999; 965 Nachman,J.B. 2007; 1289 Harrison,C.J. 2005}}. Low hypodiplody often undergoes duplication, which can be misdiagnosed as hyperdiploidy; the distinction is crucial for prognostication. Analysis of metaphase chromosomes by G-banding is the gold standard and best method in the clinical laboratory to evaluate numerical abnormalities as it provides a global overview of the genome and a baseline to trace the evolution of the disease.

Common structural aberrations

t(9;22)(q34;q11)/BCR-ABL1 translocation/Ph+

The Philadelphia (Ph) chromosome results from the balanced translocation t(9;22)(q34;q11) and is the hallmark of chronic myelogenous leukemia (CML). Three to five percent of childhood and 25% of adult ALL have the Ph chromosome {{1895 Aric␠, M. 2010; }}. Until recently, the Ph+ ALL have been consistently considered as a very high risk group with dismal outcome {{734 Ribera,J.M. 2002; }}. Addition of specific tyrosine kinase inhibitors, such as imatinib to chemotherapy regimens have changed the traditional outlook of Ph+ ALL as a "hallmark high-risk" requiring stem cell transplantation {{749 Schultz,K.R. 2007}}, to one with a good outcome {{1440 Schultz, K.R. 2010; 1895 Aric␠, M. 2010; }}. There are two major forms of BCR-ABL1 translocations depending on the breakpoints on the BCR gene (Figure 2-3(A)). The translocation consistently involves exon 2 of the ABL gene, but occurs in different exons of the BCR gene. The fusion involving the major breakpoint region (MBR) between exons 12 and 13 or 13 and 14 leads to expression of an 8.5 kb transcript coding for a 210kD fusion protein (p210BCR-ABL1) that characterizes adult CML, and about 30% of ALL. In 70% of Ph+ ALL, the breakpoint in the BCR gene is in between alternate exon1 and exon 2 (minor breakpoint region/m-BR), resulting in a smaller 7.5 kb transcript coding for a 190 kD protein (p190 BCR-ABL1 ) {{1439 Verma, D. 2009; 1896 Westbrook,C.A. 1992; }}. Other less common fusions {{1441 Melo, J.V. 1997;}}, and co-expression of the p210BCR-ABL1 and p190 BCR-ABL1 encoding transcripts as a result of alternative splicing in the M-BR of BCR, have been reported {{1880 Lemes,A. 1999; 1887 Volpe,G. 2007; 1879 van Rhee,F. 1996; 1883 Solves,P. 1999; 1890 Saglio,G. 1996; 1888 Lichty,B.D. 1998; 1888 Lichty,B.D. 1998; 1875 Kunieda,Y. 1994; }} . rtPCR (reverse transcriptase polymerase chain reaction) and conventional cytogenetic analysis readily detect the BCR-ABL1 fusion transcript/t(9;22)(q34;q11); in those cases with a failed, cryptic, or normal karyotype result, FISH provides a reliable alternative method (Figures 2-3(B) and 2-3(C)){{1289 Harrison,C.J. 2005;}}. Dual-color, dual-fusion FISH probes have high specificity and can also detect associated deletions of derivative chromosome 9, and complex or variant translocations {{1308 Primo,D. 2003; 1367 Robinson,H.M. 2005; }}.

t(4;11)(p13;q23)/MLL-AF4

Chromosomal translocation between the C-terminal of MLL gene at chromosome band 11q23 and AF4 on chromosome 4p13 results in the MLL- AF4 fusion (Figure 2-4(A) and 2-4(B)). t(4;11) demonstrates an age-dependent distribution- it is observed in 50-70% of infant ALL and roughly 5% of pediatric and adult cases. Clinically it is associated with high-risk ALL. Most patients have a CD10-, CD19+ (pro-B) profile with co-expression of myeloid antigens. Breakpoints on MLL gene are dispersed over a wide region, and clustered differently in infant versus other patients. To date, 104 translocation partners are known of which 64 have been molecularly characterized {{1444 Meyer,C. 2009; }}. The biological behavior and distribution of the different MLL translocations depends on the translocation partner, for example t(4;11) AF4/MLL is almost exclusively seen in infants and has the worst outcome; t(11;19) ENL/MLL {{1025 Jansen,M.W. 2005}} occurs in both ALL and AML , and has a poor prognosis; whereas t(9;11) AF9/MLL present in AML has a favorable prognosis. At present, the combination of cytogenetic and molecular techniques, including FISH, specific rtPCR and genomic PCR methods, seems to be the best approach to identify MLL rearrangements, notably unusual, complex and cryptic chromosomal rearrangements IS THIS COPIED AND PASTED DIRECTLY FROM JOURNAL??? {{1442 De Braekeleer, E. 2010;1025 Jansen,M.W. 2005; }}. A systematic evaluation starting with a commercially available MLL split signal FISH probe is recommended for screening as a first step. Should an abnormality be detected, karyotype or genomic PCR-based molecular methods can be used to further delineate the recombination {{1442 De Braekeleer, E. 2010}}.

t(1;19)(q23;p13)/TCF3-PBX1 (E2/PBX1)

t(1;19) is detected in 5-6% of childhood ALL and can occur as a balanced or an unbalanced translocation. It occurs almost exclusively in pre-B ALL expressing cytoplasmic  (CD10+, CD19+, cIg+) {{829 Borowitz,M.J. 1990; 1321 Pui, C.H. 1990}}. Initially considered to have an unfavorable prognosis {{1324 Crist,W.M. 1990}}, an outcome similar to that for standard risk ALL was demonstrated in subsequent studies for TCF3-PBX1 ALL when patients received intensive chemotherapy regimes {{1322 Raimondi,S.C. 1990; 1323 Pui,C.H. 1994; }}.

t(12;21)(q21;q22)/ETV6-RUNX1 (TEL/AML1)

This is the most frequent structural abnormality in pediatric ALL. It is present in 25% of patients and is associated with a favorable outcome. ETV6/RUNX1 is a cryptic translocation and cannot be detected by conventional metaphase cytogenetics (Figures 2-5(A) and 2-5(B)) {{1365 Nordgren,A. 2002; 1364 Mathew,S. 2001; 1363 Douet-Guilbert,N. 2003}}.

t(8;14)(q24;q32)/MYC-IGH

The translocation, t(8;14)(q24;q32), and the variant forms t(2;8)(p13;q24) and t(8;22)(q24;q11), are found in Burkitt lymphoma, some diffuse large B cell lymphomas, and some ALLs. Mature B-cell ALL or Burkitt-like ALL are rare, comprising 1-2% of pediatric ALL {{717 Navid,F. 1999; }}. With a few exceptions, ALL blasts with translocations involving MYC have a FAB-L3 morphology and demonstrate a mature B cell phenotype, with clonal expression of immunogobulin light chain {{1449 Davey,F.R. 1992;1451 Hammami,A. 1991; 1450 Kaplinsky,C. 1998; 1447 Komrokji,R. 2003; 1446 Navid,F. 1999; 1448 Gupta,A.A. 2004; 1285 Harrison,C.J. 2001; }}. These translocations juxtapose the oncogene, MYC, located at 8q24, to the immunoglobulin heavy chain (IGH) at 14q32, kappa (IGK) or lambda (IGL) light chain genes at 2p12 and 22q11 respectively. Despite the low incidence of this type of leukemia and its poor response to conventional treatment for B-precursor ALL, it is recognized that patients with B cell ALL have an improved survival when treated with very intensive chemotherapy {{1446 Navid,F. 1999}}.

Precursor T-cell leukemia/lymphoma

About 10-15% of pediatric ALL have a T-cell immunophenotype. In comparison with precursor B-ALL, they are more frequent in the older children and adolescents, and present with a higher tumor burden {{1076 Pui,C.H. 2004; }}. Historically, T-ALL have been considered as high-risk with increased incidence of relapses, although the outcome has greatly improved with more intensive protocols {{1053 Goldberg,J.M. 2003; 1079 Pui,C.H. 2004; }} .

Genetic aberrations

In approximately 50% of T-ALL, structural chromosomal aberrations can be identified by conventional karyotyping {{1081 Rubnitz, J.E. 1999; 1080 Raimondi,S.C. 1999; 1117 Graux,C. 2006}}. Numerical changes are rare, except for tetraploidy seen in approximately 5% of cases, and are without prognostic significance {{1117 Graux,C. 2006; }}. Translocations involving the TRA and TRD (14q11) and TRB (7q34) are seen in 35% of patients {{715 Szczepanski,T. 2000; 1113 Ellison,D.A. 2005; 1117 Graux,C. 2006; 1116 Bellido,M. 2000; 1115 Cauwelier,B. 2006; }}, and result in upregulation of oncogenic transcription factors involved in T-cell differentiation. Other rearrangements include fusion translocations not involving the T-cell receptor loci, for example, the cryptic interstitial deletion of TAL1 at 1p32 resulting in the SIL/TAL1 chimeric gene (~30% T-ALL), t(10;11)(p13;q14)/ CALM-AF10 {{ 1048 Asnafi,V. 2003; }}, and t(5;14) (q34;q32) {{ 1051 Bernard,O.A. 2001; }} both of which are cryptic, and translocations involving the MLL gene {{1035 Raimondi,S.C. 1993; }}. Specific signatures obtained by GEP analyses can segregate the major oncogenic pathways, to which different genetic lesions converge. The major oncogene clusters are NOTCH1 {{1453 Aster,J.C. 2005; 1455 Rao,S.S. 2009; 1454 Real,P.J. 2009; }}; HOXA {{1035 Raimondi,S.C. 1993; 1456 Ferrando, A.A. 2003; 1457 Speleman,F. 2005; 1460 Dik,W.A. 2005; 1459 Soulier,J. 2005; 1458 Van Vlierberghe,P. 2008; }}; TLX1 (HOX11) AND TLX3 (HOX11L2) {{1098 van Grotel,M. 2006; 1099 van Grotel,M. 2008; 1100 Ballerini,P. 2008; 1101 Ballerini,P. 2002; 1088 Ferrando,A.A. 2004; 1042 Cave,H. 2004; 1121 Berger,R. 2003; }}; beta-helix-loop-helix (SCL/TAL1) {{1461 Kikuchi,A. 1993; 1462 Janssen,J.W. 1993; }}, LYL1{{ 1124 Ferrando,A.A. 2002; }}; ABL1 {{1057 Graux,C. 2004; 1043 De Keersmaecker,K. 2005}}; JAK1 {{1464 Flex,E. 2008; }}; and JAK2 {{1463 Peeters,P. 1997}}.

Similar to precursor B-ALL, the outcome in T-ALL is associated with molecular and cytogenetic abnormalities {{1038 van Grotel,M. 2006; 1045 van Grotel,M. 2008; }}. T-ALL with translocations involving the TAL1 gene [t(1;14) (p32;q11) or submicroscopic interstitial deletion resulting in the SIL-TAL1 fusion and HOX11 [ t(10;14)(q24;q11) or t(7;10)(q35;q24) ] have a better outcome, and those involving HOX11L2 [t(5;14)(q35;q32) and CALM-AF10 t(10;11)(p14;q14)], JAK1, and ABL1 do poorly independently, or in combination {{1461 Kikuchi,A. 1993; 1462 Janssen,J.W. 1993; 1037 Attarbaschi,A. 2010; 1042 Cave,H. 2004; 1100 Ballerini,P. 2008; 1038 van Grotel,M. 2006; 1045 van Grotel,M. 2008; 1104 Ferrando,A.A. 2004; 588 Szczepanski,T. 2010; 1464 Flex,E. 2008; }}.

Unlike precursor B-ALL, the clinical significance of immunophenotypic subtyping based on surface antigen expression in T-ALL is less clear {{1059 Pullen,J. 1999; }}, except for some recurrent chromosomal rearrangements that are associated with the expression of specific antigens suggesting a developmental stage of T-cell. For example, TAL1 translocations commonly have a surface CD3+, TCR/+, mature phenotype, and translocations involving HOX11 are frequently CD1a+, representing a cortical or intermediate stage of development. Early precursor T-ALL (ETP-ALL), a recently described distinct biological subtype by GEP, with high risk of remission induction failure or relapse, has a distinctive immunophenotype - (CD1a-, CD8-, CD5 (weak), and presence of stem-cell or myeloid markers, consistent with its origin from early T-cell precursors {{773 Coustan-Smith,E. 2009; }}. Table 3 shows the clinically important genetic abnormalities in precursor T-ALL and the recommended method for their evaluation.

INSERT TABLE 3

About 50% of the abnormalities in T-ALL can be detected by conventional cytogenetics, and the common translocations are detectable by FISH {{1117 Graux,C. 2006; 1044 Gorello,P. 2010; 588 Szczepanski,T. 2010; }}. A high percentage of cryptic abnormalities, such as cryptic deletions at 9p21 and 1p32, and translocations with breakpoints near terminal regions of chromosomes, for example 9q34 breakpoints, t(5;14)(q35;q32), and rearrangements of TRB at 7q34 are often disclosed only with appropriate FISH probes {{1117 Graux,C. 2006}}. A recent study established a novel Q-rtPCR assay for detection of NUP214-ABL1 {{1465 Burmeister, T. 2006}}. Although risk-stratified protocols have not been employed commonly in T-ALL, the recent discoveries of involvement of NOTCH1, ABL1, and JAK kinases in T-ALL have the potential for translation into novel targeted therapies {{1465 Burmeister, T. 2006; 1455 Rao,S.S. 2009; 1454 Real,P.J. 2009; 1057 Graux,C. 2004; 1464 Flex,E. 2008; }}.

Acute Myeloid Leukemia (AML)

Compared with ALL, AML is less common in the pediatric age group, accounting for 16% of AL in children less than 15 years of age, and 36% in adolescents and young adults in the 15-20 years age group. A bimodal distribution is observed, with increased frequency of AML in children less than 2 years compared to older children.

Genetic aberrations

Cytogenetics is the most important predictor of outcome in childhood AML {{1335 Kaspers,G.J. 2007; 1336 Rubnitz,J.E. 2008; 1337 Pui,C.H. 2004; 1152 Manola,K.N. 2009; 1186 Harrison,C.J. 2010; }}. Cytogenetic abnormalities, seen in 70-85% of pediatric AML, is higher than in adult AML {{1152 Manola,K.N. 2009; 1186 Harrison,C.J. 2010; }}, and form the basis of classification {{239 Swerdlow, S.H. 2008}} and risk stratification of childhood AML {{1186 Harrison,C.J. 2010; 1212 Lange, B.J. 2008; 1215 Meshinchi, S. 2007; }}. The cytogenetic and molecular subtypes of adult AML carry the same prognostic significance in pediatric AML but with a difference in age distribution. For example, the frequency of MLL (11q23) translocations decreases from nearly 40% of AML in infants to 10% in older children {{1152 Manola,K.N. 2009; }}. Certain cytogenetic and molecular abnormalities such as t(1;22)(p13;q13) translocation with aberrant expression of OTT-MAL fusion gene occurs exclusively in non-Down syndrome-associated AMkL in infants {{1283 Bernstein, J. 2000 ;}}.

INSERT FIGURES 3a, 3b, 3c???

Others such as NPM1 and FLT3 mutations are less commonly observed in childhood AML. NPM1 mutations in childhood AML occur in 8-10% of AML cases and in approximately 25% of those with a normal karyotype {{1170 Brown,P. 2007; 1158 Hollink,I.H. 2009; }}, as opposed to adult AML (overall frequency 35% ; normal karyotype AML 60%) {{1342 Thiede,C. 2006; 1338 Verhaak,R.G. 2005; }}. The frequency of FLT3 internal tandem duplication (FLT3 ITD) is 10-15% in children versus 20-30% in adults {{1350 Zwaan, C.M. 2003; 1351 Meshinchi, S. 2001; 1352 Schnittger, S. 2002;}}. NPM1 and FLT3 mutations have similar prognostic implications in childhood AML as in adult patients {{1344 Gale,R.E. 2008; 1350 Zwaan, C.M. 2003; 1351 Meshinchi, S. 2001; 1353 Meshinchi,S. 2006; }} - presence of NPM1 mutation in normal karyotype AML is associated with a favorable outcome when not associated with FLT3-ITD; coexistence of NPM1/FLT3-ITD is associated with an intermediate prognosis; whereas FLT3 ITD alone predicts the worst outcome for the patient {{1351 Meshinchi, S. 2001; 1170 Brown,P. 2007; 1346 Rau,R. 2009; }}. NPM1 mutation is stable during disease evolution, and represents a possible marker for minimal residual disease detection. Table 4 lists the cytogenetic and molecular genetic abnormalities common to and of prognostic importance in pediatric AML.

INSERT TABLE 4

Therapy-related AML/MDS (t-AML/MDS)

Use of intensive treatment protocols for childhood cancers has improved outcome for the primary tumors at the cost of increase in t-AML/MDS. It has been estimated that tMDS/tAML affects at least 1% of childhood cancer patients {{1501 Tucker, M.A. 1987; }}, with a 2% cumulative incidence at 15 years for secondary myeloid neoplasms. The frequency of t-AML/MDS in children depends on the nature of primary tumor, nature of therapy {{1606 Cohen, R.J. 2005; 1609 Bhatia,S. 2002; 1612 Pui,C.H. 1991}}, duration of therapy, and the underlying genetic predisposition {{1615 Bogni,A. 2006; }}. A higher incidence is seen in children treated for Hodgkin disease, sarcomas, especially Ewing's sarcoma, and ALL patients who received topoisomerase-II inhibitors, {1605 Kushner, B.H. 1988; 1606 Cohen, R.J. 2005; 1611 Neglia,J.P. 1991; 1612 Pui,C.H. 1991; 1608 Leung,W. 2001; 1610 Kreissman,S.G. 1992; 1609 Bhatia,S. 2002; }}. A higher incidence is also observed for children with neurofibromatosis 1 (NF1) and other germline DNA repair disorders {{1606 Cohen, R.J. 2005}} , or those with polymorphisms in drug metabolizing enzymes {{1613 Chen, H. 1996; 1614 Blanco, J.G. 2002; }}. An increased risk of t-AML was also observed in children who received granulocyte colony-stimulating factor (G-CSF) following ALL induction chemotherapy {{1616 Relling, M.V. 2003}}.

Most of the therapy related leukemias are t-AML. Therapy-related ALL comprises 5% of secondary leukemias. Molecular detection of IGH and TCR- gene rearrangements has facilitated the identification of t-ALL from the primary ALL. The outcome for t-myeloid neoplasms is very poor. Cytogenetic abnormalities are a major determinant of outcome. Abnormalities of chromosome 5, 7 or 21q22 translocation are commonly observed; the frequency of classical recurrent non-random translocation t(8;21), t(15;17), 11q23, and inv 16 are common in pediatric t-AML/MDS due to the frequent use of topoisomerase II inhibitors, especially in ALL. Developing individualized treatment protocols based on the genetic framework of the patient and targeted to a specific antigen or genetic pathway in the specific tumor is the goal for molecular medicine to check the increase in second neoplasms.

Acute leukemia associated with Down syndrome

Children with Down Syndrome (DS) have a 150-times greater risk of developing AML {{1227 Fonatsch,C. 2010;1259 Hasle,H. 2000; }}than those without DS. The most common AML subtype in DS is acute megakaryoblastic leukemia (AMkL), and the risk in DS is increased 500-fold {{1258 Zipursky,A. 2003}}. DS-AML has an excellent cure rate of nearly 80% compared to the non-DS AML . Virtually all children with DS- AMkL have somatic mutations in the GATA1 gene {{1265 Wechsler,J. 2002}}. The mutations result in a truncated GATA1 protein (also referred to as GATA1s) that contributes to uncontrolled megakaryoblastic proliferation. Somatic GATA1 mutations are specific to AMkL occuring in trisomy 21 patients {{1265 Wechsler,J. 2002}}; the underlying mechanism of this association is as yet unresolved.

DS children also have a 20-fold increased risk for developing ALL {{1259 Hasle,H. 2000; 1235 Xavier,A.C. 2009; }}. In contrast to DS-AML, DS-ALL is of very high-risk and has a worse outcome compared to non-DS ALL {{1275 Tigay,J.H. 2009; 1280 Whitlock,J.A. 2005; 1274 Maloney,K.W. 2010; }}. GATA1 mutations are not present in DS-ALL {{1273 Whitlock,J.A. 2006}}. Recent studies have shown presence of a R683 somatic mutation in JAK2 in 18-28% of DS-ALL patients {{1277 Kearney,L. 2009; 1281 Bercovich, D. 2008; }}. Aberrantly increased expression of cytokine receptor CRLF2 is present in the majority of DS-ALL with R683-mutated JAK2 {{1276 Hertzberg,L. 2010; 753 Russell,L.J. 2009; }}, suggesting a synergistic effect between the two mutations. The activation of the CRLF2-JAK-STAT signaling pathway in the majority of DS-ALL suggests a therapeutic potential for JAK inhibitors.

Conventional cytogenetics and FISH are important in detection of trisomy 21, the latter is especially useful in DS mosaics {{1226 Kudo,K. 2010; }}. GATA1 mutation testing may play a useful role in the initial stratification of AMkL into chemosensitive or chemoresistant groups (Stepensky P, Fuster). GATA1 mutation can be detected by sequencing and can be employed in follow-up for MRD. Down-syndrome related myeloid proliferations and their link to AL is discussed in greater detail in the chapter on pediatric myelodysplasia and myeloid proliferation.

Blastic Plasmacytoid dendritic cell tumor (BPDCN)

BPDCN, formerly known as blastic natural killer (NK)-cell lymphoma and CD4+/CD56+ hematodermic neoplasm {{1129 Jegalian,A.G. 2009; }}, is a rare subtype of AL, even less frequent in children{{1126 Jegalian,A.G. 2010; 1127 Hama,A. 2009}}. Based upon phenotypic (CD4+, CD56+, CD123+ BDCA-4 +, CD303/BDCA-2+, TCL1+,CD68-), functional, and genetic features, the cell of origin is believed to be the hematopoietic precursor committed to a plasmacytoid dendritic cell (PDC) lineage {{1129 Jegalian,A.G. 2009; 1128 Garnache-Ottou,F. 2009; }}. A leukemic presentation in the absence of skin lesions is described more often in the pediatric patients. Pediatric BPCDN are reported to have a fairly uniform and "lymphoblast"- like appearance and S-100 positivity in neoplastic cells compared to a variability of morphologic features and S-100 negativity seen in adult cases IS THIS COPIED {{571 Jegalian,A.G. 2010; 1485 Cota, C. 2010; }}. Two-thirds of BPDCN have complex, non-random cytogenetic abnormalities {{571 Jegalian,A.G. 2010; }}. GEP studies have identified a set of genes expressed in BPDCN not described in other hematopoietic cells, but highly expressed in neuronal cells and implicated in neurogenesis {{1486 Dijkman, R. 2007;}}.

Hereditary disorders with predisposition to AL

An increased predisposition for development of AML has been observed in children with inherited disorders that include disorders of DNA repair and increased chromosomal fragility, for example Fanconi's anemia and Bloom's syndrome; bone marrow failure syndromes, such as Kostmann's syndrome, Diamond-Blackfan anemia, familial platelet disorder and Shwachmann-Diamond syndrome; and in Neurofibromatosis type I {{1358 Shimamura,A. 2006; 1360 Rosenberg,P.S. 2008; 1362 Dror,Y. 2008; 1361 Alter,B.P. 2010; 1361 Alter,B.P. 2010; 1357 Alter,B.P. 2007; 1356 Jongmans, M.C. 2010}}. The development of leukemia in these disorders often occurs in adulthood. Highly sensitive and specific diagnostic tests based on the mutations identified for many of the inherited disorders are available {{1357 Alter,B.P. 2007}} for confirmation of the underlying germline genetic disorder and in investigation of family members of the affected individual. These tests and the different laboratories offering them are available at http://www.genetests.org