Clinical impact of dose intensity of initial tyrosine kinase inhibitor (TKI) therapy in accelerated-phase CML (CML-AP).

Abstract

6587 Background: For patients (pts) presenting with CML-AP, TKIs are recommended. Although generally well tolerated, dose reductions and/or interruptions are often required. The impact of TKI dose intensity (DI) on CML-AP outcome has not been described. Aim: To determine impact of TKI dose interruptions and/or reductions on outcome in de novo CML-AP. Methods: Overall, 58 CML-AP pts (median age 46 yrs) were treated on consecutive or parallel clinical trials with TKIs as initial therapy: imatinib (n=36), dasatinib (n=5), or nilotinib (n=17). CML-AP features included: blasts ≥15% (n=8), basophils ≥20%, (n=22), platelets <100x109/L (n=3), cytogenetic clonal evolution (n=22), or >1 feature (n=3). We examined the impact of dose reductions, treatment interruptions, and DI (ratio of the dose received vs. the intended dose for the entire treatment duration) on event-free (EFS), transformation-free (TFS), and overall survival (OS). Results: Among 58 pts, 37 required treatment interruptions and/or dose reductions (29 required dose reductions). 12 pts required dose reductions and/or interruptions due to hematologic toxicity and 23 for non-hematologic toxicities. Pts were divided into 3 DI tiers: 100% (n= 21), 90-99%, (n=10), and <90% (n=27). Outcomes were analyzed based on DI tiers and presence or absence of dose reductions as shown in the table. Pts with dose reductions and/or interruptions for hematologic toxicities vs. non-hematologic toxicities had a 24-mo EFS rate of 87% vs. 100% (p=0.05). 24-mo EFS by dose reductions and/or interruptions within 6 mo or later were 100% and 91%, respectively (p =0.7). Conclusions: DI has no significant impact on de novo CML-AP outcome. However, dose interruptions and/or reductions for hematologic toxicities are associated with worse 24-mo EFS, suggesting myelosuppression may in some patients be more a feature of the disease than toxicity from therapy. [Table: see text]