Acute myeloid leukemia (AML) in older patients is associated with a poor prognosis, with lower complete remission (CR) rates and worse overall survival compared to younger patients. Moreover, most patients over the age of 70 years do not tolerate standard induction chemotherapy. Alternative therapy with hypomethylating agents can improve CR rates and survival compared to best supportive care, but overall outcomes remain poor with current therapeutic options. Preclinical studies suggest that "epigenetic priming" using decitabine followed by cytarabine increases the cytotoxicity of cytarabine. It is hypothesized that this is due to the reactivation of genes that have been silenced by the malignancy. The aim of this study is to evaluate the efficacy and safety of a novel induction regimen using decitabine followed by cytarabine in older patients with newly diagnosed AML who are not candidates for intensive chemotherapy. Interim response rates were reported at the ASH Annual Meeting in 2014. Here we present updated response rates, treatment-related mortality, and overall survival.


This is a phase 2, single arm study at the University of Pittsburgh Cancer Institute (NCT01829503) for patients over the age of 70 years with newly diagnosed AML, or patients over the age of 60 years who are considered not to be candidates for intensive chemotherapy. The induction regimen consisted of decitabine 20mg/m² intravenously (IV) x 5 days followed by cytarabine 100mg/m² continuous IV infusion x 5 days. Patients with no evidence of disease on day 15 bone marrow biopsy proceeded with maintenance decitabine after count recovery; patients otherwise proceeded with a second cycle of induction using the same regimen. Patients with progressive disease after 1 cycle were taken off study. After a second induction cycle, patients who achieved a complete or partial remission proceeded with maintenance decitabine. Maintenance cycles consisted of decitabine 20mg/m² IV x 5 days every 4-8 weeks until disease progression or toxicity. Response rates were determined by the International Working Group Response Criteria in AML. Four-week and 8-week mortality rates were assessed.


Forty-six subjects were enrolled as of August 2015, 36 of whom were evaluable for response at the time of analysis. Median age was 76 years (range 67-88 years). There were 21 females (45%) and 26 males (55%). The median ECOG performance status was 1. There were 21 patients with poor risk cytogenetics at diagnosis. Of 36 patients who were evaluable for response, 20 patients had a CR and 5 patients had a CRi (CR/CRi rate 69%). Six patients had a partial remission, and 5 patients had resistant disease. All evaluable patients except for 6 received 2 cycles of induction. There were no 4-week mortalities and 4 (8.6%) 8-week mortalities. Deaths were attributed to subdural hemorrhage, multifactorial respiratory failure, progressive AML, and neutropenic sepsis. At a median follow up of 13.5 months, the overall survival is 12.4 months (95% CI:9.7-12.5).


We have shown that an induction regimen using decitabine as an epigenetic primer followed by cytarabine induces high CR/CRi rates with low treatment-related mortality in older adults with newly diagnosed AML who are not candidates for intensive chemotherapy, a patient population in whom there exists a dire need for novel treatment strategies. In the updated report of this phase 2 study, 69% of patients achieved a CR/CRi, and 4- and 8-week mortality were 0% and 8.6%, respectively. This compares favorably with historical outcomes of both intensive chemotherapy and decitabine monotherapy in older adults in terms of safety and efficacy, respectively. Overall survival was 12.4 months, which also compares favorably to previous reports of survival in this patient population. Methylation analyses at baseline and after decitabine in patients who achieved CR compared to patients who did not respond are ongoing and will be reported. We have demonstrated that decitabine followed by cytarabine is safe and effective in older adults with newly diagnosed AML.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.