Please note that we have stopped the regular imports of Gene Expression Omnibus (GEO) data into ArrayExpress. This may not be the latest version of this experiment.
E-GEOD-8879 - Transcription profiling of human atypical T-ALL
Submitted on 27 August 2007, released on 25 October 2008, last updated on 10 June 2011
Despite improved therapy, approximately one-fifth of children with acute T-lymphoblastic leukemia (T-ALL) succumb to the disease, suggesting unrecognized biologic heterogeneity that may contribute to drug resistance. We studied leukemic cells, collected at diagnosis, to identify features that could define this high-risk subgroup. A total of 139 patients with T-ALL were treated consecutively from 1992 to 2006 at this institution. Their leukemic cells were examined with multiparameter flow cytometry, single nucleotide polymorphism arrays and other methods of genomic analysis. Survival rates and probabilities of treatment failure were calculated for subgroups considered to have biologically distinct forms of T-ALL. The lymphoblasts of 17 patients (12.2%) had immunophenotypes that clearly differed from those of the remaining patients. Most striking features were the low or absent expression of common T-lineage markers (CD5, CD1a and CD8), and expression of myeloid and stem cell markers, suggesting stem-cell like properties. These atypical blasts also showed a distinct gene expression profile and increased genomic instability overall. Patients with this form of T-ALL had a very low probability of surviving for 10 years after diagnosis: 19% ± 12% (mean ± SE) versus 84% ± 6% for all other patients ( P<0.001). The cumulative incidence of remission failure / hematologic relapse was 72% ± 16% versus 10% ± 3% (P < 0.001). A subset of children with T-ALL have atypical blast cell immunophenotypes that predict a dire outcome with use of standard intensive chemotherapy. Alternative treatment strategies are need for patients with this sybtype of leukemia. Experiment Overall Design: 55 Diagnostic T-ALL samples were analyzed
transcription profiling by array, unknown experiment type