Expression data from ALL patients included in the set used to construct a classification signature (COALL cohort)

Childhood acute lymphoblastic leukemia (ALL) comprises a large group of genetic subtypes with a favorable prognosis characterized by a TEL-AML1-fusion, hyperdiploidy (>50 chromosomes) or E2A-PBX1 fusion and a smaller group with unfavorable outcome characterized by either a BCR-ABL-fusion, MLL-rearrangement or T-ALL. About 25% of precursor B-ALL are currently genetically unclassified and have an intermediate prognosis. The present study used genome-wide strategies to reveal new biological insights and advance the prognostic classification of childhood ALL. A double-loop cross validation was used to construct a classifier based on gene expression in ALL cells from 190 newly diagnosed cases (COALL cohort, GEO GSE13425) with a prediction accuracy of 90%. T-ALL, TEL-AML1-positive, hyperdiploid and E2A-rearranged cases were identified with 100% sensitivity and ≥94% specificity. The classifier accuracy was confirmed in an independent cohort of 107 cases (87.9%, DCOG cohort, GEO GSE13351). Keywords: gene expression study for classification of ALL subtypes 190 bone marrow and peripheral blood samples were collected at diagnosis and frozen. They were later thawed and hybridized to Affymetrix U133A arrays.