PMC:5003446 / 13792-19516 JSONTXT

{"project":"2_test","denotations":[{"id":"27600067-12036858-69477837","span":{"begin":592,"end":594},"obj":"12036858"},{"id":"27600067-14726504-69477838","span":{"begin":672,"end":674},"obj":"14726504"},{"id":"27600067-15695375-69477839","span":{"begin":703,"end":705},"obj":"15695375"},{"id":"27600067-16230371-69477840","span":{"begin":1025,"end":1027},"obj":"16230371"},{"id":"27600067-18506749-69477841","span":{"begin":1058,"end":1060},"obj":"18506749"},{"id":"27600067-18506749-69477842","span":{"begin":1293,"end":1295},"obj":"18506749"},{"id":"27600067-20016533-69477843","span":{"begin":1639,"end":1641},"obj":"20016533"},{"id":"27600067-19651600-69477844","span":{"begin":2547,"end":2549},"obj":"19651600"},{"id":"27600067-19651601-69477845","span":{"begin":3757,"end":3759},"obj":"19651601"},{"id":"27600067-20729466-69477846","span":{"begin":4324,"end":4326},"obj":"20729466"},{"id":"27600067-19770377-69477847","span":{"begin":4712,"end":4714},"obj":"19770377"},{"id":"27600067-22084373-69477848","span":{"begin":5278,"end":5280},"obj":"22084373"}],"text":"3. Acute Myeloid Leukemia\nAcute myeloid leukemia (AML) is genetically heterogeneous, ranging from cases with recurrent cytogenetic abnormalities to approximately 40% cases with normal karyotype, some of which have prognostically significant somatic mutations. AMLs with t(15;17), inv(16) and t(8;21) respond well to chemotherapy and have good prognosis, while AMLs with Philadelphia chromosome t(9;22), complex karyotype, −5/−5q, −7/−7q, and 11q23 translocations have poor prognosis. In AMLs with normal karyotype, FLT3 internal tandem duplications (ITDs) are associated with poor prognosis [39], while mutations of NPM1 and CEBPA are associated with favorable prognosis [40]. In 2005, Raghavan et al. [41] first applied a 10K SNP array in 64 cases of AML and identified partial uniparental disomy (pUPD) in approximately 20% cases. A similar study with 10K SNP array demonstrated homozygous mutations involving genes FLT3, CEBPA, and RUNX1 in approximately 50% patients with UPD involving corresponding chromosomal regions [42]. Subsequently, Gupta et al. [43] expanded the study with 10K SNP array on 454 cases of AML from young adults, and found nonrandom acquired UPD (aUPD) in 17% cases, preferentially affecting chromosomes 13q, 11p, and 11q, similar to the findings by Raghavan et al. [43], and additional recurrent aUPDs at 2p, 17p, 2q, 17q, 1p, and Xq. AMLs with FLT3-ITD had aUPD 13q involving the FLT3 gene, while AMLs with FLT3-ITD-/FLT3-TKD+ mutation did not have aUPD 13q.\nThese early studies suffered from the low resolution of the SNP arrays available at that time. With advances in SNP array technology, Bullinger et al. [44] were able to use 50 and 500K Affymetrix SNP arrays on 157 cases of cytogenetically normal AML. The cohort showed 12% aUPDs with chromosomal regions 6p, 11p, and 13q most commonly affected and all aUPDs were \u003e29 Mb. The aUPDs were associated with mutations in NPM1 or CEBPA, which suggested that aUPDs may affect genes critical for hematopoiesis. In terms of aCNAs, as expected, the 500K SNP array was much more sensitive than 50K SNP array, which missed approximately 60%–70% of CNAs detected by the 500K SNP array. aCNAs were identified in 49% cases, with genomic losses (1.21/case) more frequent than gains (0.25/case). The recurrent genomic deletions included bands 3p14.1-p13, 6q27, 8q23.3, 10q11.21, 11q25, 12p13.2, and 15q21.3, that harbor genes FOXP1 and RYBP (3p14.1-p13), RPS6KA2 (6q27), TRPS1 (8q23.3), HNRPF (10q11.21), ETV6 (12p13.2) and RFXDC2 (15q21.3). In a separate study, Walter et al. [45] applied a 1.85 million SNP array from Affymetrix in 86 adult patients with de novo AML, and identified a total of 201 aCNAs in 44% of the cases with a mean of 2.34 CNAs per genome. Acute erythroid leukemia and acute megakaryocytic leukemia had more CNAs (10–29 CNAs per genome) than other morphologic variants of AML. CNAs were detected in 24% AMLs with normal cytogenetics and in 40% AMLs with abnormal karyotype. 12 chromosomal regions (eight deletions and four amplifications) containing at least one gene implicated in AML or MDS (deletions of 3p14.1: FHIT, 5q31.1: CTNNA1, 12p12.3: ETV6, 16q22.1: CBFB, 17p13.1: TP53, 17q11.2: NF1, and amplifications of 8q23.2: MYC, 11q23.3: MLL, and 21q22.2: ETS2) were identified in multiple AML cases. CNAs in chromosomal regions 17q11.2 and 21q22.2 (CNAs spanning NF1 and ETS2) were found in at least five cases, most of which had complex karyotypes, and associated with worse overall survival. aUPD was infrequent occurring in eight of 86 genomes, with most in AML with normal cytogenetics. 50% of the AML studied had no CNA or UPD at this resolution. The identified CNAs did not predict overall or event-free survival independent of cytogenetics. Radtke et al [46] showed similarly very low burden of genomic alterations in pediatric de novo AML, with a mean of only 2.38 somatic copy-number alterations per case. These studies indicate that AMLs are not genomically unstable and the genes implicated in these CNAs likely play important role in the leukemogenesis of AML.\nWhile Walter’s study did not identify any prognostically significant CNAs or UPDs, three groups clearly showed prognostic significance of SNP array lesions in AML, likely due to utilization of more stringent SNP lesion detection algorithms. Parkin et al. [47] examined 114 previously untreated prospectively enrolled AML patients with Affymetrix SNP 6.0 arrays, and showed that ≥2 genomic lesions detected by SNP 6.0 array almost doubled the risk of death after controlling for age- and karyotype-based risk by multivariate analyses. P53 mutations, or P53 mutations coupled with 17p-LOH conferred an independent negative prognosis. Tiu et al. [48] performed 250 K and 6.0 SNP arrays on 140 cases of primary (p) and secondary (s) AML, and demonstrated that patients with genomic lesions including acquired somatic UPD identified by SNP array had worse overall survival (OS) and event-free survival (EFS) in pAML with normal cytogenetics and in pAML/sAML with abnormal cytogenetics. The SNP array lesions, AML type, and metaphase cytogenetics had independent predictive value for OS by multivariate analyses. In a study of 133 cases of AML with normal cytogenetics by Genome-Wide Human SNP 6.0 Array, Yi et al. [49] found at least one abnormal SNP lesion in 32.3% cases. Detection of abnormal SNP lesions by SNP-A karyotyping conferred an unfavorable prognosis for overall survival by multivariate analyses. All these studies confirmed the clinical relevance as well as prognostic significance of SNP lesions in AML, especially AML with normal cytogenetics, which would allow a better prognostic stratification of patients with AML for appropriate treatment."}