Abstract
High-dose cytarabine (Ara-C) has been reported with increased treatment-related mortality, whereas few data are available concerning intermediate-dose Ara-C for induction of acute myeloid leukemia (AML) with t(8;21) translocation. We retrospectively analyzed factors impacting complete remission (CR), event-free survival (EFS), cumulative incidence of relapse (CIR), and overall survival (OS) in 197 adults with t(8;21) AML, of whom 107 cases were induced with intermediate-dose and 90 with standard-dose Ara-C (as part of 3 + 7 protocol). After a single induction course, the overall CR rate was 87.6% (170/194), with a significant difference between the standard-dose (83/105, 79.0%) and intermediate-dose (87/89, 97.8%) groups (p < 0.001). Rather than general KITmut, the specific KIT-D816 independently led to a lower probability of achieving CR (HR = 3.29 [1.18-9.24], p = 0.023), worse EFS (HR = 3.53 [1.82-6.84], p < 0.001), and OS (HR = 5.45 [1.77-16.84], p = 0.003) in the standard-dose group, but not in the intermediate-dose group. CD19(+) represented the only independent factor predicting lower CIR both in the standard-dose group (HR = 0.32 [0.10-1.00], p = 0.050) and in the intermediate-dose group (HR = 0.11 [0.03-0.40], p = 0.001). When combined, KIT(+) plus CD19(-) conferred the most increased relapse risk (3-year CIR 60%; SE 0.12). Specific KIT-D816, instead of general KITmut, may be incorporated in prognostication model for t(8;21) AML. Combination of CD19 with KIT provides a more definite risk stratification profile for t(8;21) AML.