Acute lymphoblastic leukemia (ALL) is the most common cancer in children. Patients vary widely in their response to drug therapy. This difference can result in too little treatment, leading to treatment failure and disease relapse, or to too much treatment, leading to toxic side effects. Variations in genes that code for enzymes that are important for drugs effects and thus named candidate genes, may underlie these differences, offering a powerful tool in predicting treatment outcomes and thus for the adjustment of treatment based on patient genetic profile. Within our study of ALL pharmacogenetics, we identified variations in particular candidate genes that confer higher risk of relapse in ALL patients. In this study, we will validate these findings by analyzing the effect of these polymorphisms by means of in vitro assays and in vivo, using a humanized mouse model of ALL.