Benefit of TKI-Treatment in CML after Failing Molecular or Cytogenetic Milestones

No information is available on long-term survival of tyrosine kinase inhibitor (TKI)- treated patients failing the response milestones as defined by the European LeukemiaNet (ELN). Such information would help with decision making in failing patients when risks associated with comorbidities, drug toxicities or transplantation preclude switching to another TKI or other treatments.

To systematically analyze survival after reaching, or not reaching, ELN-response milestones, 1342 imatinib-treated patients from CML-study IV with newly diagnosed CML in chronic phase and regular molecular and cytogenetic tests were studied. Survival was analyzed by landmark analyses at 3, 6, 12 and 24 months and followed up to 14 years. In patients who failed the failure milestones (>10% BCR::ABL1^IS at 6 months, >1% BCR::ABL1^IS at 12 months), 10- to 12-year survival ranged around 70%, 10-20% less than in responding patients. An example is shown in Fig. 1a which depicts survival of patients reaching, or not reaching 1% BCR::ABL1^IS at 12 months. Similar survival curves were observed for the other landmarks that indicate failure. Inbetween response rates of >1%-10% BCR::ABL1^IS at 12 months indicated survival probabilities similar to those of patients failing 1% BCR::ABL1^IS, whereas patients reaching >1%-10% BCR::ABL1^IS at 24 months showed a survival similar to that of patients failing 10% BCR::ABL1^IS. Switching to alternative therapies, mostly to dasatinib and nilotinib, was observed in 372 patients (26.9%) and did not change the main results.

Although cytogenetic tests were increasingly replaced by molecular tests in the course of the study, sufficient cytogenetic analyses were available for comparison with molecular analyses. Complete and partial cytogenetic remissions have been shown to be roughly equivalent to 1% and 10% BCR::ABL1^IS, respectively. Wheras reaching, or not reaching cytogenetic and molecular reponse markers at 6 and 12 months was equivalent, not reaching a complete or partial cytogenetic remission at 24 months identified a small subgroup of patients (5.9%) with inferior survival (Fig. 1b).

Age (more or less than 60 years) had no major impact on survival differences, but on hazard ratios for early death. Hazard ratios at 10% BCR::ABL1^IS by 6 months were 4.0 (95%CI:2.0-8.1) for the younger patients, but only 0.9 (95% CI: 0.4-1.8) for the elderly. The reasons for early death in the younger patients were mostly CML-related, whereas the elderly primarily died of other reasons.

The data show that TKI-treated patients not reaching failure milestones still may derive benefit from continuing TKI-treatment and provide a basis for individualized decisions, if failing patients are confronted with risks of alternative treatments.