Leukemia cells, from children with acute myelogenous leukemia (AML) at diagnosis, harbor gene expression signatures that may be useful for risk stratification and therapy assignment. Our hypothesis is that prognostic gene expression profiles among patients with de novo childhood AML treated on different clinical protocols differ because gene expression signatures are retrospectively identified, and usually defined by therapy and outcome. In contrast, gene expression signatures useful for risk stratification and therapy assignment should be robust, and independent of treatment. We propose to complete gene expression profiling in samples from the POG #9421 study, and perform array comparative genomic hybridization, i.e. surveying gene amplifications and deletions across the genome, for correlation with gene expression profiles. Furthermore, the gene expression signatures identified in the POG #9421 training set will be cross-validated with patients enrolled in the AML 2002 clinical protocol. The overall objective is to identify gene expression signatures that can be used in clinical decision making.
Specific aims of this proposal are:
Aim 1. To Profile Gene Expression of Childhood de novo AML Specimens from patients enrolled in POG #9421 and AML 2002 Studies. Gene expression signatures that determine outcome in patients with FLT3 mutations, and inv(16) cytogenetics have been identified in our analysis of 130 samples from the POG #9421 study. These signatures require cross-validation to assess their suitability for prospective validation in frontline pediatric AML studies. Analysis of 145 additional samples is in progress while AML 2002 samples are accruing from a consortium of Pediatric Oncology Institutions.
Aim 2. To Cross-validate Gene Expression Signatures from POG #9421 and AML 2002 Studies. Cross-validation of key prognostic signatures between specimens from each study will be done. Each cohort will be used as a test set for expression signatures derived from the other cohort. Using cohorts of patients treated differently may facilitate the identification of robust signatures that are prognostic across treatment schemas.
Aim 3. To Profile Array CGH in de novo AML specimens from patients enrolled in POG #9421. Genome-wide gene copy numbers will be analyzed to identify prognostic copy number changes, correlation with cytogenetic groups, and gene expression profiles.
| O'Brien, Maureen M; Taub, Jeffrey W; Chang, Myron N et al. (2008) Cardiomyopathy in children with Down syndrome treated for acute myeloid leukemia: a report from the Children's Oncology Group Study POG 9421. J Clin Oncol 26:414-20 |
| Becton, David; Dahl, Gary V; Ravindranath, Yaddanapudi et al. (2006) Randomized use of cyclosporin A (CsA) to modulate P-glycoprotein in children with AML in remission: Pediatric Oncology Group Study 9421. Blood 107:1315-24 |
| DiMartino, J F; Lacayo, N J; Varadi, M et al. (2006) Low or absent SPARC expression in acute myeloid leukemia with MLL rearrangements is associated with sensitivity to growth inhibition by exogenous SPARC protein. Leukemia 20:426-32 |
| Lacayo, Norman J; Meshinchi, Soheil; Kinnunen, Paivi et al. (2004) Gene expression profiles at diagnosis in de novo childhood AML patients identify FLT3 mutations with good clinical outcomes. Blood 104:2646-54 |
