Second-generation tyrosine kinase inhibitors (2G-TKI), nilotinib or dasatinib used after imatinib failure can induce complete cytogenetic response (CCR) in 50% of chronic myeloid leukemia (CML) patients. BCR-ABL transcript levels reduction in the initial months of treatment has been associated improved outcome.


to evaluate the predictive value of early molecular responses, at 3 and 6 months after treatment with 2G-TKI in CML patients with imatinib failure or intolerance; to correlate these responses with CCR, overall survival (OS), progression-free survival (PFS) and event free survival (EFS).


Between September 2007 and August 2012, 71 consecutive patients with CML resistant or intolerant to imatinib were treated with dasatinib (n= 50) or nilotinib (n=21). BCR-ABL transcripts were measured in peripheral blood using real-time quantitative PCR (RQ-PCR) at 3 months intervals. Results were expressed as BCR-ABL/ABL ratio, with conversion to the international scale (IS). Major molecular response (MMR) was defined as a transcript level ≤ 0.1% (IS). Cytogenetic analysis was performed at baseline, 3, 6, 12 and 18 months after starting therapy with 2G-TKI. BCR-ABL mutation analysis by direct sequencing was investigated in resistant patients. Probabilities of OS, PFS and EFS were calculated using Kaplan-Meier method. An event was defined as the loss of CHR, CCR, progression to advanced phases, death or 2G-TKI discontinuation. The CCR probabilities according to molecular responses were calculated by c2 method and cumulative incidence, considering as competitive event death or progression.


71 patients were analyzed, median age of 47 years (15-81); Disease status before 2G-TKI was: 50 (71%) CP; 13 (18%) AP and 8 (11%) BC. Responses: 59/71 (83%) obtained CHR; 38/55 (69%) CCR and 37/60 (62%) MMR. At 3 months of therapy, 81.5% (44/54) had a BCR-ABL ratio ≤10% and at 6 months 66% (33/50) had ≤ 1%. At 3 months, CCR was obtained 65% (19/29) pts with ≤10% RQ-PCR and 16% (1/6) with >10% RQ-PCR (p= 0.06). At 6 months, CCR was 100% (12/12) in pts with RQ-PCR ≤ 1% and 14% (1/7) in those with >1% (p< 0.0001). The probability to achieve RQ-PCR < 10% at 3 month was 43% (95% CI 32-54%). During treatment 3 (4%) progressed to AP and 5 (7%) to BC. The 5-year probability of OS was 78% (95% CI: 68-88%) albeit by disease status was 86% in CP, 92% in AP and 12% in BC (p< 0.0001). OS was inferior in pts with RQ-PCR > 10% at 3 months (60 vs 84%, p= 0.03) and >1% at 6 months (68 vs100%, p= 0.006). PFS was 68% in 5-year, and was lower in BC pts (p< 0.0001) and pts with RQ-PCR >1% at 6 months (p= 0.004). EFS was 53% and lower in BC pts (p< 0.0001), in pts with RQ-PCR > 10% at 3 months (p= 0.005) and > 1% at 6 months (p< 0.0001). RQ-PCR at 3 and 6 months were also predictive of a worse survival when patients in CP were analyzed separately. 2G-TKI was discontinued in 44% (31/71) due to: resistance (n=18), intolerance (n=5), death (n= 3), HSCT (n=3) and loss of follow-up (n=2). Eleven BCR-ABL mutations were detected in 36 pts; 3 previously 2G-TKI (L387M-1, E255K-1, M351T-1) and 9 after therapy (T315I-5, M244V-2, E255V-1, Y253H-1). OS by mutation was 45% with any mutation and 88% with no mutation (p= 0.05).


BCR-ABL transcript levels at 3 and 6 months can identify patients with worse prognosis and less chance to obtain CCR with 2G-TKI after imatinib treatment.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.