Low chemosensitivity is the main reason for poor prognosis of relapsed acute lymphoblastic leukemias in children. In this study, the team tried to elucidate the genetic background of chemoresistance in relapses. By comparing 103 diagnosis-relapse samples pairs they identify genes with relapsed-specific mutation involved in drug metabolism. Next, they compared 208 samples from 16 patients taken during the time from diagnosis to relapse. By this approach, they could track the first appearance of resistant clone during therapy and determine if the relapse is due to pre-existing clone at diagnosis (observed in very early relapses) or due to acquired chemoresistance during treatment (observed in early and late relapses). These findings suggest that chemoresistance in relapses could be preventable by altered front-line treatment strategies.