Quantitative Analyses of BCR-ABL Mutations Associated with Imatinib Treatment in CML.

Acquired mutations in the BCR-ABL gene encoding the kinase domain constitute an important explanation for clinical resistance to imatinib in CML. Therapeutic options upon detection of resistance and mutation may include increased imatinib dosage, substitution of imatinib to a novel tyrosine kinase inhibitor or hematopoietic stem cell transplantation. Recently, imatinib resistant clones without evident association to clinical resistance have been detected using assays with high sensitivity. This observation emphasizes the importance of clarifying whether detection of single mutations motivates reassessment of clinical treatment options. To elucidate this question, blood samples from 38 chronic phase CML patients demonstrating suboptimal (<3 log reduction) or failing (increasing BCR-ABL transcripts or Ph-pos cytogenetically) clinical response to imatinib were assessed for mutations in the ABL gene fragment encoding the kinase domain. We used sequencing of BCR-ABL specific PCR products and a novel quantitative, single step PCR approach (qPCRMUT) for selective detection of mutated clones targeting 12 important mutations (all P-loop mutations, T315I, M351T, sensitivity 0.1 to 1%). Whenever mutations were detected, we aimed at a quantitative, longitudinal analysis of all available samples by qPCRMUT. We detected 28 BCR-ABL clones from 26 patients containing mutations in the ATP binding region. Three of these mutations appeared to be novel variants: two point mutations targeting amino acids located near the P-loop (positions G236E and D241G) and a deletion variant (deletion of 81 bases including the P-loop) induced by a 3′-splice site introduced by the L248V. All other detected mutations have been described before in association with imatinib resistance. Seventeen of the 28 mutated clones were followed quantitatively by qPCRMUT. We found two patterns of kinetic development: In 7 subjects the clones were detected only in small amounts in a single sample and quantitative follow up did not show any evidence of further selection. In 8 subjects increasing amounts of mutated BCR-ABL clones could be followed from their first detection until they finally replaced wild type BCR-ABL, a course accompanied by clinical progression. In samples of 10 other subjects large amounts of mutated clones were dominating BCR-ABL at the time of clinical resistance, indicating that these clones may indeed have triggered the resistance. No mutations were found in 10 clinically imatinib resistant patients. Four subjects were monitored quantitatively after withdrawal of imatinib. We noticed selection of P-loop mutated clones upon treatment with busulfan and deselection during treatment free periods. In conclusion, among 38 CML patients with suboptimal or failing response to imatinib we detected mutations in 26 (68%). Not all detected clones were selected upon imatinib treatment. Repeated quantitative assessment of mutated clones provided early indications of clinically significant disease progression. Monitoring of mutated BCR-ABL clones may provide a valuable supplement in CML surveillance, particularly in the setting of imatinib resistance.