Acute myeloid leukemia (AML) primarily affects older adults. Treating these patients requires balancing risk of toxicity with risk of ineffective therapy. Here we examine the ability of decitabine combined with cytarabine (DecitAC) to strike such a balance. Our objectives were to 1) examine the likelihood DecitAC would improve 6-month survival probability from historical 60% (poor risk)-65% (good risk) in older patients treated with azacitidine + gemtuzumab on SWOG study - S0703 (Nand S et al Blood. 2013;122(20);3432-3439) to 80% in patients age ≥60 with newly diagnosed AML or high risk MDS (10-19% blasts); 2) determine response rate in this population; and 3) evaluate the effect of cytarabine dose escalation in patients who had a response short of CR after the first 2 cycles of the combination.

Patients and Methods:

In addition to age ≥ 60, patients needed a treatment related mortality (TRM) score <22.9 corresponding to <25% chance of death within 28 days of beginning 7+3 or high dose cytarabine-containing induction (Walter et al. JCO 2011). Prior use of azacitidine, lenalidomide, or decitabine, but not more intense therapy, was permitted. Doses were: decitabine 20mg/m2 once daily days 1 through 10, and cytarabine 100mg/m2once daily days 1 through 7. Two courses were planned to be given 28-35 days apart, regardless of blood counts. Bone marrow evaluation was planned after DecitAC course 2. The protocol allowed for cytarabine dose escalation after course to 1gm/m2 daily x 5 days for patients with ≥ 20% blasts by flow cytometry and if non-hematologic toxicities were < grade 3. If blasts were 5-19% by flow after two cycles, the doctor would choose between the original dose of cytarabine or the dose escalated dose. For patients that had ≥ grade 3 non-hematologic toxicities, the number of DecitAC dosed days were shortened if <5% blasts by flow after 2 cycles, or the patient was taken off study if marrow had ≥5% blasts by flow. Treatment was given outpatient or inpatient and patients were often managed by their referring physicians.


12 newly diagnosed patients (10 AML, 2 MDS) were treated on protocol at our academic center. 18 others (16 AML, 2 MDS) were treated off study because of higher TRM scores or by community oncologists, with chart review on an IRB approved protocol. TRM scores ranged from 0.5 - 61 with a median of 10. Ages ranged from 46-92 with a median of 73. Fourteen of the 26 AML patients had an antecedent hematological disorder. Two AML patients had ELN favorable risk disease, 12 had intermediate, and 12 adverse.DecitAC produced CR in 14/30 patients (47%; 95% CI 28-65%), CRi was seen in 7 other patients. No patients died within the first 28 days (95% CI 0-11%, i.e. up to 3 patients) contrasted with 5 expected deaths based on TRM score. Dose escalation of cytarabine to 1 gm/m2 on course 3 was conducted in 2 patients (TRM scores 14.2 and 9.8), each of whom died on this couse.Median survival in the 30 patients was 8 months and is 19 months in the 9 patients still alive as of August 2016. At 6 months survival probability was 57% in this group considered poor risk by definition in the comparison group vs. 60% in the controls. Results were not affected by whether patients were managed here or by their local physicians. Grade 3 adverse events included infections, diarrhea, hepatic dysfunction, and others, as expected with other regimens treating newly diagnosed AML and MDS patients.


DecitAC resulted in a CR+Cri rate of 21/30 (70%) caused less TRM than expected. Whether this reflects the general decline in TRM or a specific effect of DecitAC is unknown. DecitAC can be given equally well in the community as at an academic center. However pending multivariate analysis it does not appear to improve survival relative to that seen previously.


Becker:GlycoMimetics: Research Funding.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution