Retrospective Application of European LeukemiaNet Provisional Criteria for Second-Generation TKI Chronic Myeloid Leukemia

Abstract 2270

An update of the European LeukemiaNet criteria for monitoring response of chronic myeloid leukemia patients was recently published and provisional criteria to evaluate patients during second generation TKI therapy after resistance to imatinib were proposed. In our study we retrospectively tested these criteria in a large series of CML patients resistant to imatinib further treated with second generation TKIs with the aim to analyze the outcome of suboptimal response and failure patients compared to those with optimal response and to validate the provisional criteria for monitoring response. One hundred twenty-seven CML patients resistant to imatinib were collected from 6 different Italian hematologic centers. There were 66 males and 61 females, median age 54 years (range 25–80). Twenty-seven patients were in late chronic phase after IFN resistance. Ninety-seven patients received second-generation TKI after acquired resistance, whereas 30 patients had primary resistance. We found that at different time points (3, 6 and 12 months), patients classified as failure showed significantly worse 2-year overall survival (OS), progression-free survival (PFS) and event-free survival (EFS) than sub-optimal and optimal response patients. At 3 months, “failure” patients, had an OS of 83% compared to 86% of sub-optimal and 97% of optimal response patients (p=0.001); PFS was 77% for failure patients compared to 92% and 99% for sub-optimal and optimal response patients, respectively (p=0.001), whereas EFS was 41% for failure vs 59% for sub-optimal (p=0.001) and 85% and optimal response patients, respectively (sub-optimal vs optimal p<0.001). At 6 months, OS was 82%, 88% and 99% for failure, sub-optimal and optimal response patients (p=0.05), respectively; PFS was 82% for failure compared to 94% and 99% for sub-optimal and optimal response patients, respectively (p=0.001); EFS was 47% vs 69% for failure and sub-optimal response (p=0.001) and 86% for optimal response patients (sub-optimal vs optimal, p<0.001). At 12 months again OS was 84% for failure patients compared to 95% and 99% for sub-optimal and optimal response patients (p=0.04); PFS was 86%, 95% and 99% for failure, sub-optimal and optimal response patients, respectively (p=0.001) and EFS was 48% for failure, 67% for sub-optimal response patients (p=0.002) and 89% for optimal response patients (sub-optimal vs optimal, p<0.001). We found that ELN provisional criteria identified at any times worse EFS for sub-optimal response patients, similar to that of failure patients, and failure criteria at 3 months identified patients who had worse PFS and EFS. ELN provisional criteria for second-generation TKIs treated patients appear to clearly predict outcome and can be useful to identify patients at high risk of progression.