High-Resolution Molecular Allelokaryotyping of Chronic Myeloid Leukemia Patients in Blast Crisis by 6.0 SNP-Arrays Shows a High-Frequency of Uniparental Disomy and Focal Copy Number Alterations Affecting the Whole Sequence or Specific Exons of Oncogenes and Tumor Suppressor Genes.

Abstract

Abstract 2176Poster Board II-153Progression from chronic phase to blast crisis (BC) remains a major hurdle on the road to effective treatment of chronic myeloid leukemia (CML). BC is known to be associated with accumulation of additional genetic alterations, but these alterations have so far been only partially characterized. The development of SNP-arrays as a tool for high-resolution karyotyping now allows to perform high-throughput genome-wide screens for submicroscopic genomic alterations with unprecedented informativity and resolution and to precisely map all the genes involved in these alterations. We have used Human 6.0 SNP Arrays (Affymetrix) to perform high-resolution molecular allelokaryotyping of 25 DNA samples from BC (myeloid, n=16; lymphoid, n=9) CML patients (pts). The 6.0 SNP Array technology relies on 1.8 million markers evenly spaced across the genome, with a median inter-marker distance <700 bp. Loss of Heterozygosity (LOH) analysis identified several recurrent regions of uniparental disomy (UPD) ranging from 970Kb to 2.4Mb: 3p21.31-3p21.2 (19 pts); 4p15.1 (n=18 pts); 14q23.3 (n=18 pts); 8q22.2 (n=15 pts); 7q31.31 (n=14 pts); 3q11.2 (13 pts); 17q23.2 (n=13 pts); 12q24.11-12q24.13 (n=12 pts); 15q15.2-15q15.3 (n=12 pts); 16q22.1 (n=12 pts); 10q22.1-10q22.2 (n=11 pts); 1p34.3 (n=10pts); 7q11.22 (n=10 pts); 8p11.12 (n=10 pts); 15q23-15q24.1 (n=9 pts); 20q11.22-20q11.23 (n=7 pts); 16q11.2 (n=6 pts); 17q11.2 (n=5 pts). Three pts had evidence of UPD involving the whole long arm of chromosomes 5, 14 and 19, respectively. Macroscopic copy number alterations (CNAs) (+8, +19, +14q; +21q; -7; -18, -16q; -17p; -6p; -6q; -9q) were frequent and easily detected. A variety of submicroscopic CNAs were also detected. However, we decided to exploit the unprecedented resolution power of Human 6.0 SNP Arrays and the ability of Genotyping Console 3.0.2 (Affymetrix) software to precisely pinpoint the borders of these CNAs. We thus aimed our analysis to the identification of very small CNAs that may have been missed by previous studies - all using less sensitive assays. This approach revealed a high number of focal gains or losses ranging from 4 to 47Kb, affecting single genes or even some exons only. Genes involved in >2 pts are listed in the Table below. Gains/losses mapping to known regions of copy number variation (CNV) were excluded.Gene SymbolType of CNAExons involvedChromosomal locationNo of pts harboring CNAsMDM4aAll7q22.115 ptsHGFaAll1q32.14 ptsNRAS*aAll1p13.23 ptsHRAS*aAll11p15.53 ptsKRAS*aAll12p12.15 ptsAKT3*mdOne or more1q4313 ptsCDC73mdOne or more1q31.28 ptsRB1md+aOne or more13q14.28 ptsERGmd/bdOne or more21q22.38 ptsETS1aOne or more11q24.38 ptsETV1aOne or more7p21.27 ptsSMAD2ma+dOne or more18q21.17 ptsPIK3CA*ma+dOne or more3p26.326 ptsSKAP2aOne or more7p15.25 ptsEPHA3aOne or more3p11.25 ptsKALRN*mdOne or more3q21.25 ptsFHITmdOne or more3p14.25 ptsJAK1dOne or more1p32.35 ptsJAK2ma+dOne or more9p24.14 ptsAKT2*mdOne or more19p13.24 ptsSOS1*mdOne or more2p22.14 ptsPTEN*ma+dOne or more10q23.314 ptsFGFR2aOne or more10q26.34 ptsETV5aOne or more3q27.24 ptsMETaOne or more7q31.24 ptsPBX4mdOne or more19p13.114 ptsNRG1mdOne or more8p124 ptsFAF1mdOne or more1p334 ptsGAS2mdOne or more11p14.33 ptsTECaOne or more4p123 ptsPTCH1ma+dOne or more9q22.323 ptsPIK3R1*aOne or more5q13.13 ptsGRB2*aOne or more17q243 ptsPAK2*mdOne or more3q293 ptsRUNX1T1aOne or more8q21.33 ptsAbbreviations: md, monoallelic deletion; bd, biallelic deletion; a, amplification; ma+d, complex pattern of monoallelic deletion of some exons and amplification of other exonsAll the genes found to harbor CNAs were transcription factors, adaptor proteins, receptor and non-receptor kinases involved in cell proliferation and apoptosis - with a known role as oncogenes or tumor suppressors or oncogene/tumor suppressor interactors. Although these results confirm a high degrees of heterogeneity in the alterations detectable in BC CML pts, members of the RAS pathway (indicated with an asterisk) were the most frequently altered genes. Further characterization by polymerase chain reaction and sequencing is ongoing. In conclusion, the power of 6.0 SNP Array technology allowed us to detect previously unidentified alterations targeting whole or part of key oncogenes or tumor suppressors whose deregulation may play a role in determining the aggressive phenotype of BC CML, and which may represent potential therapeutic targets. Supported by European LeukemiaNet, AIL, AIRC, PRIN, Fondazione del Monte di Bologna e Ravenna. Disclosures:No relevant conflicts of interest to declare.