Long Term Observation of CML Patients after Imatinib Resistance Associated with BCR-ABL Mutations.

The major cause of imatinib resistance in CML patients (pts) is clonal selection of cells harboring mutations of the BCR-ABL tyrosine kinase associated with various degrees of insensitivity to the drug. Management and outcome of these pts have been a matter of debate. In particular, retrospective studies suggested a poor prognosis of pts with P-loop mutations, whilst others did not confirm this observation. We sought to determine the outcome of 121 CML pts (73 m, 48 f; median age 61 yrs, range 28–80) with BCR-ABL mutations detected by direct sequencing and associated with hematologic (n=105), cytogenetic (n=14), or molecular (n=2) resistance. Imatinib was commenced between 8/99 and 3/05 after a median of 33.1 mo (range 0–312.3) after diagnosis, the median exposure to the drug was 17.0 mo (0.9–55.4). Resistance occurred after 18.1 mo (0.9–55.4) in chronic phase (CP, n=63), 18.3 mo (5.1–52.8) in accelerated phase (AP, n=33), 11.4 mo (0.9–47.7) in myeloid blast crisis (BC, n=20), and 9.1 mo (1.8–45.6) in lymphoid BC (n=5). Resistance was associated with 41 different mutations leading to a change of 32 amino acids (aa) of the P-loop area (n=38), T315I (n=15), the activation loop (n=11), and other sites (n=74). 15 pts showed 2, 2 pts 3, and one patient 4 mutations in parallel. In addition, resistance was associated with a loss of 27 aa including the P-loop in 2 pts. ABL polymorphisms were detected in 3 pts (T315T, n=1; K247R, n=2). Median time until resistance was 12.1 mo (0.9–52.8) for pts with P-loop mutations, 11.5 mo (2.8–47.7) for T315I, 17.2 mo (5.0–35.5) for activation loop mutations, and 20.4 (2.0–55.4) for others. After hematologic resistance, imatinib therapy was continued for 3.5 mo (0–54.1). Dose was increased in 42 pts up to 800mg/d, which was associated with disease stabilization in 7, hematologic improvement in 4, and cytogenetic or molecular improvement in 2 cases. In 46 pts, imatinib was combined with other drugs, 33 received low dose cytarabine. After imatinib withdrawal, treatment was continued with hydroxyurea, n=53, cytarabine, n=32, and/or other cytostatic drugs. 10 pts underwent allogeneic stem cell transplantation (SCT), 5 of them died. Recently, 2nd generation tyrosine kinase inhibitors were administered to 35 pts (dasatinib, n=19, AMN107, n=15, both drugs, n=1). Censoring pts at the time of SCT or start of alternative tyrosine kinase inhibitors, overall survival after resistance was 29.7 mo (0–52.3). 38 pts died; of these, mutations were located in the P-loop, n=16, at T315, n=4, or at other sites, n=18. Median survival after hematologic resistance was 29.7 mo for pts with P-loop mutations, 13.0 mo for pts with T315I, and 33.5 mo for others (n.s.). Considering CP pts only, median survival in case of P-loop mutations was 35.8 mo and has not been reached for pts with T315I and others (n.s.). In advanced disease, median survival with P-loop mutations was 17.0 mo, for T315I 13.0 mo, and for others 29.5 mo (n.s.). In conclusion, this prospective analysis failed to reveal a significantly impaired survival of pts with P-loop mutations compared to others. There is, however, a trend for a survival disadvantage in pts harboring T315I. Outcome of CML pts with mutations associated with insensitivity to imatinib, in particular P-loop and T315I mutations can be improved by early detection of the mutation and rapid withdrawal of the drug.