Validation Study on the Prediction of Response to Imatinib Mesylate in Chronic Myeloid Leukemia (CML) Patients by Genome-Wide cDNA Microarray Analysis.

Imatinib mesylate (IM) is highly effective in the first line therapy of CML. However, it remains unknown whether IM cures CML or not, and other therapeutic modalities like stem cell transplantation sometimes become necessary. Predicting the therapeutic effect of IM before the start of therapy would be useful, because other kinds of therapy could be planed during IM treatment. Previously, in order to predict the effect of IM therapy in CML, we established a prediction system by cDNA microarray containing 23,040 genes. According to the profiling data of RNA expression of peripheral blood mononuclear cells in 22 CML patients, who had previously failed in interferon-α therapy, 79 genes were identified that were expressed differentially between responders and non-responders to IM. By a prediction scoring system, 15 out of the 79 genes were selected which discriminated the two populations most clearly.(Kaneta Y et al. Jpn J Cancer Res 93, 849–856,2002) This time we applied this scoring system to another cohort of chronic or accerelated phase of CML patients who would receive IM as monotherapy. So far, 42 patients were analyzed. Of 40 patients who attained major cytogenetic response (MCR) by 6 months from the start of therapy, 38 (95%) had been predicted as the responder by the prediction system. On the other hand, two patients whose effect was less than MCR had been predicted as the non-responder. These data indicate that our prediction system for IM using cDNA microarray could effectively discriminate responders from non-responders before the start of therapy.