Limited data are available on azacitidine (AZA) treatment and its prognostic factors in acute myeloid leukemia (AML). One hundred and forty-nine previously untreated AML patients considered ineligible for intensive chemotherapy received AZA in a compassionate patient-named program. AML diagnosis was de novo, post-myelodysplastic syndromes (MDS), post-MPN, and therapy-related AML in 51, 55, 13, and 30 patients, respectively. Median age was 74 years, median white blood cell count (WBC) was 3.2 × 10⁹ /L and 58% of the patients had ≥ 30% marrow blasts. Cytogenetics was adverse in 60 patients. Patients received AZA for a median of five cycles (range 1-31). Response rate (including complete remission/CR with incomplete recovery/partial remission) was 27.5% after a median of three cycles (initial response), and 33% at any time (best response). Only adverse cytogenetics predicted poorer response. Median overall survival (OS) was 9.4 months. Two-year OS was 51% in responders and 10% in non-responders (P<0.0001). Adverse cytogenetics, WBC >15 × 10⁹ /L and ECOG-PS ≥ 2 predicted poorer OS, while age and marrow blast percentage had no impact. Using MDS IWG 2006 response criteria, among patients with stable disease, those with hematological improvement had no significant survival benefit in a 7 months landmark analysis. Outcomes observed in this high-risk AML population treated with AZA deserve comparison with those of patients treated intensively in prospective studies.
Copyright © 2013 Wiley Periodicals, Inc.