Gene expression-based imatinib response prediction in chronic phase CML

Tyrosine kinase inhibitor (TKI) treatment of chronic myeloid leukemia (CML) is guided by the pre-defined European Leukemia Net (ELN) or other response criteria. This allows patient stratification only during but not prior to treatment initiation. Gene expression profiling (GEP)-based response prediction might become a valuable tool for patient stratification in CML, but so far published data for response prediction are conflicting. We generated an imatinib response predicting gene signature by GEP from peripheral blood samples of pre-treated CML patients in late chronic phase. We generated an imatinib response predicting gene signature by GEP from peripheral blood samples of pre-treated CML patients in late chronic phase withdrawn prior to (n=135) and 6 weeks after starting imatinib therapy (n=63). According to the study protocol, patients received either standard (400 mg) or high dose (800 mg) imatinib for six months followed by standard dose maintenance therapy. GEP were generated by using the Affymetrix hgu133plus2 GeneChip. Prediction was based on linear discriminatory analysis applied in a nested cross-validation