The value of “high dose” cytarabine (HDAC) during initial induction therapy for patients with newly diagnosed AML remains controversial with some prior reports suggesting improved relapse-free survival (RFS) but not complete remission (CR) or overall survival (OS) (Kern and Estey, Cancer 2006; 107: 116-24), and others finding no benefit (Lowenberg, Blood 2013; 121:26-28). The benefit of HDAC in the management of more favorable subsets of AML (such as core binding factor AML) is well-established, demonstrating higher sensitivity of this group to cytarabine. It is commonly acknowledged that AML is several diseases that may respond differently to a given therapy. With the increasing use of NPM1 and FLT3-ITD to stratify patients, it seemed plausible that the value of HDAC during induction might vary according to patient’s genotype as defined by NPM1 and FLT3-ITD mutation status. In particular, patients with normal cytogenetics/ NPM1 mutated/FLT3-ITD negative AML clearly have more favorable outcomes and, hence, are potentially a group that might benefit from HDAC therapy. Herein we examine this possibility.


We analyzed 1080 patients (median age 53, range 17-88) with newly diagnosed AML. The standard dose cytarabine (SDAC) group comprised 531 patients who received 7 days of cytarabine at 100 mg/m2 daily + 3 days of daunorubicin (3+7) on SWOG protocols (S0106 N=368, S9031 N=73, S9333 N=46, S9500 N=44). The HDAC group consisted of 549 patents treated on various MD Anderson (MDA) induction protocols, each of which administered cytarabine at either 1.5g/m2 daily X 4 days by continuous infusion or at 1-2 g/m2 daily X 5 days, most commonly combined with idarubicin, fludarabine, or clofarabine +/- other agents. Once in CR, both the SDAC and HDAC groups received HDAC consolidation. Only 16% of patients at MDA received allogeneic transplant in first CR (off-protocol transplant data not available for SWOG). We used log-rank tests to compare RFS and OS in the SDAC and HDAC groups and multivariate Cox regression to adjust for other potential prognostic covariates: numerical age, sex, performance status (0-1 vs. >2-4), numerical WBC count, platelets, % marrow blasts, and cytogenetics (SWOG criteria).


15% of the 1080 patients were positive for NPM1 mutation (N+)/negative for FLT3-ITD (F-) (13% MDA, 17% SWOG), 10% N-/F+ (9% MDA, 11% SWOG), 10% were N+/F+ (9% MDA, 11% SWOG), and 65% were N-/F-(69% MDA, 61% SWOG). CR rates were similar in each of the 4 groups regardless of the cytarabine dose (interaction p=0.42). On multivariate analysis, among N+/F- patients, OS was superior in the SDAC group (HR=0.50, p=0.02)). Multivariate analysis indicated that OS was similar between SDAC and HDAC in each of the other cohorts and that RFS was similar in all four cohorts. , When we restricted the analysis to the 545 patients age 60 with de novo AML treated since 2000 (N+/F- 15%, N-/F+ 9%, N+/F+ 10%, N-/F- 65%) there was no evidence that SDAC or HDAC was associated with CR, RFS or OS. When we restricted the analysis to the 418 patients with normal cytogenetics (N+/F- 27%, N-/F+ 13%, N+/F+ 20%, N-/F- 40%), on multivariate analysis, among N-/F- patients, OS and RFS were superior in the SDAC group (OS: HR=0.4, p=0.001; RFS: HR=0.48, p=0.02). SDAC and HDAC were similar with respect to OS and RFS in the other three cohorts and had similar CR rates in all four cohorts.


Our data do not suggest that higher doses of cytarabine during induction improve response, RFS or OS in any of the 4 subsets defined by NPM1 and FLT3 mutations in newly diagnosed AML, overall. This may be at least partly due to the use of HDAC in consolidation in both groups.


Faderl:Sanofi-Aventis: Membership on an entity’s Board of Directors or advisory committees, Research Funding. Kantarjian:Sanofi-Aventis: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution