PRIMARY OBJECTIVE:
I. To validate significant associations between SNPs and treatment outcome and toxicity on Children's Cancer Group (CCG)-1891 on an independent sample set from a successor CCG study for standard risk acute lymphoblastic leukemia (ALL), CCG-1952.
II. To evaluate the role of SNPs in drug metabolizing enzymes and the development of veno-occlusive disease in patients on CCG-1952.
III. To evaluate interactions among genotypes and other risk factors for treatment response in a combined data set of CCG-1891 and CCG-1952 with recently developed analytic tools for high dimensional data.
IV. To develop predictive models utilizing genetic information obtained in Aim 1.1 and clinical data to predict treatment response and toxicity.
OUTLINE:
Tumor tissue samples undergo genotype assessment on the Pyrosequencing platform. Contingency tables and X\^2 test performs a univariate analysis of the risk of relapse and genotype, and multivariable analyses using logistic regression. Cox proportional hazards evaluate the risk of relapse given genotype and other confounders. Genotype patterning, classification and regression trees, and multifactor dimensionality reduction evaluates for patterns of single nucleotide polymorphisms associated with toxicity and relapse risk.
