Relapsed childhood ALL: Gaining insights and identifying targets by gene expression profiling.

8511 Background: In spite of the significant progress made in childhood ALL, the outcome for the 25% of patients who relapse is dismal, necessitating a search for new targets.,Bone marrow samples at relapse from 57 patients enrolled on Children's Oncology Group (COG) protocol, AALL01P2, for treatment of first marrow relapse were obtained. Paired samples from initial diagnosis for 10 of these patients were retrieved from COG cell banks. Total RNA was extracted, amplified and hybridized to Affymetrix U133-A microarrays. Data was normalized by Robust Multichip Analysis and Genetraffic software was used for unsupervised analysis. Differentially expressed genes were identified by SAM (Significance Analysis of Microarrays).,As timing of relapse is the most important predictor of outcome, we identified 156 genes with significant differences in expression between early (< 36 months from diagnosis) and late (>= 36 months from diagnosis) relapse. The median false discovery rate of these genes was less than 2%. Genes from several functional groups were differentially expressed. For example, regulators of apoptosis (survivin, AVEN), cell cycle regulators (cyclins B1, B2, A2, CDC2) and DNA repair genes (TOP2A, APEX1, RAD51) were among those upregulated in the early relapse group. Many of these genes have been shown to be poor prognostic markers in a variety of tumors. In a separate unsupervised analysis of diagnostic and first relapse paired samples, diagnostic samples from patients with early relapse tended to cluster with their respective relapse pair (5/7 pairs), while late relapse pairs did not cluster together (3/3 pairs). Differences in the gene expression profiles between initial diagnostic vs. relapse samples were also observed with upregulation of several emerging and established therapeutic targets (e.g.survivin, cyclin B1, PTTG1, TOP2A) in the relapse cohort.,Distinct gene expression profiles characterize newly diagnosed vs. recurrent ALL, and further distinguish relapse that recurs early from that which occurs late. These analyses may provide insight into mechanisms of ALL progression and clonal origins of recurrent disease. No significant financial relationships to disclose.