Genomic Copy Number Alterations and Clustering Analysis Using Array CGH in Acute Myeloid Leukemia with Normal Karyotype.

Abstract

Abstract 1008Poster Board I-30 Introduction:Acute myeloid leukemia with normal karyotype (N-AML) is a clinically and molecularly heterogeneous group with gene expression changes and frequently mutated genes, such as NPM1. The copy number alterations identified by array comparative genomic hybridization (array CGH) technology, is also found in these N-AML patients. This study's aim is to establish the frequency of copy number alterations, explain some of the heterogeneity of N-AML with copy number alterations, and understand their potential significance for prognostic prediction and target-therapy in patients with N-AML. Patients and Methods:Eleven patients were selected from patients diagnosed as de novo AML in Asan Medical Center during 2003 ∼ 2005, which have the normal karyotype by G-banding without the recurrent genetic abnormalities by fluorescent in situ hybridization or reverse transcription polymerase chain reaction. Their bone marrow specimen at diagnosis included more than 30% of myeloblasts. High-resolution array CGH analysis were performed with Agilent's 244K whole-genome oligonucleotide array on the patient's DNA and Promega Human genomic male pooled DNAs as reference DNA. Four cases among selected patients were also performed array CGH analysis using the bone marrow specimens at complete remission, then compared with the result of specimen at diagnosis. The clustering analysis of the array CGH data was done and compared with their clinical features and prognosis. Results:N-AML patients had a lot of copy number alterations with variable size on variable locations. None showed the copy number alteration sized of ≥ 5 Mb. However, the cryptic 1.2 Mb deletion was found at 12p13.2, including ETV6, also known as TEL, and CDKN1B. The detection of this alteration was very difficult by G-banding and confirmed by TEL FISH. High copy number altered genes with frequency of ≥10% in specimens at diagnosis were TOP2B, SDHC, SATB1, DHX36, LARS, PDZD8, MED13L, ATP8B4, TRPM7, RNF125, DNMT1, THOP1, CHAF1A, EEF2, and TRIP10. The gained genes were associated with the transport or the metabolic process, and the lossed genes were the regulation of transcription, the chromatin modification, or the nuclear mRNA splicing. In the clustering analysis, two groups were found, i.e. cluster A and cluster B, which cluster A included the cases with poor response at chemotherapy, and cluster B included the cases with relapse. The overall survival duration was shorter in cluster B than cluster A, but was not significant statistically. The enrichment analysis of biological process GO terms showed that cluster A was associated with the DNA replication, the mitosis, or the regulation of cell proliferation, and cluster B was the regulation of cell cycle or the anti-apoptosis, significantly. Conclusion:We detected not only the cryptic genomic changes but also the gene-level copy number alterations in N-AML patients using array CGH. The copy number alterations on specimens at diagnosis were variable in size and location on the chromosomes, including fifteen genes with high copy number alteration and frequency of ≥10%. Among them one case had the cryptic partial deletion of 12p including ETV6 gene, known as proto-oncogene or tumor suppressor gene in hematologic malignancies. In the clustering analysis and enrichment analysis of biological process GO terms, these variable copy number alterations were clustered as two groups with different prognosis and showed some information about leukemogenesis or pathologic status in clusters of N-AML. Considering of several previously reported studies using gene expression chip, this result should support the hypothesis that the difference of gene-expression reflects the difference of gene dosage. Disclosures:No relevant conflicts of interest to declare.