ALBERT

Description: Despite the risk stratification based on the prognostic relevant translocations, approximately 20–25% of childhood ALL patients cannot be classified according to genetic hallmarks. Recent studies based on single nucleotide polymorphism (SNP) analyses showed that Loss of Heterozygosity (LOH), with or without loss of genetic material, is frequently related to childhood ALL and AML. Moreover, there is an increasing interest in integrating the gene expression profiles with chromosomal localizations to identify new genetic subgroups. We applied an integrated approach composed of DNA index estimation, PCR (and/or RT-PCR) and cytogenetics, to exclude patients with known molecular and cytogenetics aberrations. Aim of the study was to identify criptic abnormalities in childhood ALL patients by performing an integrative analysis of gene expression and copy number changes (CNC) data. The 30 patients included in our study met the following inclusion criteria: a) B-cell precursor childhood ALL b) DNA index 1; c) negativity at t(4;11), t(12;21), t(9;22), t(1;19) RT-PCR screening; d) cytogenetics revealing normal karyotype or not technically feasible. Patients genomic DNA has been analyzed by the Affymetrix GeneChip Mapping 100K SNP arrays to identify genomic regions of LOH and CNC. In parallel, biotin-labeled cRNA has been synthesized and hybridized on HG-U133Plus2 Probe Arrays in accordance with the MILE (mircroarray innovation in leukemia) protocols; all samples were part of the B-ALL class without known genetic aberrations. The presence of del(9)(p21) was found in 9/30 (30%) patients, with an homozygous commonly deleted region involving CDKN2A and CDKN2B genes in 7/30. Hemizygous losses on 9p13 were found in two cases. Four patients (13,3%) showed the hemizygous deletion of chromosome 12p13.2, with a commonly deleted region involving the ETV6 gene. In one ETV6 deleted patient FISH analysis identified a translocation involving the first ETV6 exon and an unknown partner gene. Four patients, including three ETV6 deleted, showed the presence of the intrachromosomal amplification of chromosome 21 (iAMP21) with a common region of amplification between 31.3–43.5 Mb and a common region of deletion in 2/4 patients between 43.5–47 Mb. Four patients showed large regions of LOH not associated with CNC. Three patients did not show CNC and random microdeletions were found in single cases. Remission samples were analyzed for 5 patients carrying different aberrations, showing small regions of LOH in both the diagnosis and remission samples in 4 cases. No CNC was found in the remission samples. We found a statistically significant reduction for the expression of the CDKN2A transcript located within the (9)(p21) region. We are applying a locally adaptive statistical procedure to identify genes whose expression signals are specifically modulated by CNC. The integration of genomic variation and gene expression profiling might may be useful to identify new hidden genetic lesions, and to learn how critical regulators of tumor are linked to genomic alterations in cancer cells.

Description: Abstract 2943 Poster Board II-919 Introduction: T-cell acute lymphoblastic leukemia (T-ALL) and lymphoma (T-LBL) share common morphologic and immunophenotypic features and are treated with similar therapeutic approaches. Nonetheless, they show distinct clinical presentations suggesting that they may represent two different biological entities. In order to gain insights into the biological characteristics of these T-cell malignancies we applied genomic and transcriptomic approaches on residual diagnostic specimens from pediatric patients affected by T-ALL or T-LBL. Patients and methods: Genome-wide gene expression profiling (HG-U133Plus2.0, Affymetrix) was performed for 20 patients affected by T-LBL and 10 patients affected by T-ALL. In order to control for normal cells contamination present in the nodal biopsies (T-LBL) and bone marrow aspirates (T-ALL), we used gene expression profiles of B-cell lymphoblastic lymphoma and common (CD10+) lymphoblastic leukemia, respectively. Results: Genes differentially expressed in nodal versus bone marrow-derived samples were subtracted from the T-LBL versus T-ALL signature allowing the identification of a subset of genes able to discriminate the two T-cell malignancies regardless of their localization. This gene signature includes genes involved in chemotactic response, cell adhesion and angiogenesis which may play a role in the different tumor cell localization. Indeed, genes involved in promoting angiogenesis, invasion and nodal localization were up-regulated in T-LBL. Copy number analysis was performed using single nucleotide polymorphism (SNP) arrays (Human Mapping 100K arrays, Affymetrix) on a subset of the samples (9 T-ALL and 9 T-LBL) analyzed by gene expression profiling. This analysis detected approximately 200 genetic loci recurrently affected by copy number alterations in T-ALL and/or T-LBL. The most common aberration was the 9p21.3 deletion which includes CDKN2A/B. Consistent with previous reports amplifications involving MYB were identified in two cases of T-ALL. Although most aberrations were commonly found in both entities several were recurrently detected in T-LBL but not in T-ALL and vice versa. Conclusion: Taken together these results suggest that T-LBL and T-ALL share a large fraction of their biological features; nevertheless each malignancy displays also a unique pattern of genetic lesions and specific gene expression signatures, which may contribute to the understanding of the distinct evolution of these T-cell malignancies. Disclosures: No relevant conflicts of interest to declare.