Abstract 3811


Azacitidine is an effective therapy for high risk myelodysplastic syndrome (MDS). Neutropenic fever is a common life threatening complication during azacitidine therapy, however predicting it, is challenging. Despite a number of large scale prospective studies, there are no established indications for primary or secondary prophylactic antibiotics or for the use of granulocyte colony-stimulating factor (G-CSF) (Pierre Fenauxa et al. Leukemia Research 2010). We used a retrospective survey of 98 high risk MDS and AML patients treated with Azacitidine, to develop a predicting model for infection during each cycle of Azacitidine therapy.


We retrospectively studied 82 high risk MDS and 16 AML patients treated with 456 azacitidine cycles between 9.2008 and 7.2011at 11 institutions from Israel. Information, of complete blood count, creatinine and liver enzymes was documented prior to initiation of each cycle.


Patients' median age was 71 (range 27–92) and 57 (58%) of them males. Poor cytogenetic abnormalities were detected in 30.8% (25 of 82 patients with available cytogenetic) and 65 (67%) were transfusions dependent. The median interval between the initial diagnosis and the initiation of azacitidine therapy was 187 days (range 4 days – 18 years). Azacitidine was administrated as first line therapy in 24 (24%) of patients, 37 (38%) had failed growth factors, 5 (5%) were relapsing after allogeneic transplantation and 32 (33%) were given different chemotherapies prior to azacitidine therapy. Doses and schedule of azacitidine data were available for 98% (446/456) of cycles. The prevalence of 7 days cycles of 75mg/m2, 5 days cycles of 75mg/m2 or attenuated doses were 50.4%, 30%, 16.9% respectively.

Adverse events were obtained from patient's charts. 13 major bleeding and 78 infections episodes (2.85% and 16.9% of all cycles) were recorded. Due to the low number of bleeding events we focused on factors predicting infection episodes. Infection rates of 22.7%, 14.2% and 6.9% correlated with azacitidine dose (75mg/m2x7d Vs 5d) and lower respectively). Excluding 87 cycles of doses lower than 75mg/m2 for 5 days, predictors of infections were evaluated in 369 cycles.

Nine parameters were included in final analysis: age, sex, cytogenetics, being transfusion dependent prior to first cycle, time from diagnosis to the first cycle, azacitidine dose and neutrophil, thrombocyte and creatinine values prior to each cycle. The odd ratio off infections related to neutrophils count was higher than ANC, so we used neutrophils counts as a predictor. For each cycle we considered full 7 days Vs 5 days schedule, neutrophil above or below 500 cells/mcl, platelet above or below 20,000 cells/mcl and creatinine level prior to the first day of cycle. In univariate analysis neutrophil below 500, platelet below 20,000, creatinine level, azacitidine dose and being transfusion dependent were correlated with infection. In a multivariate analysis (table 1) transfusion dependency and platelets lower than 20,000 were the only significant parameters. Risk of infection was higher when a full seven days cycle was administrated but haven't reach statistical significance (p=0.07).

Table 1
Sig.Odd ratio95% C.I
Transfusion dependent .007 3.783 1.43–10.02 
platelet <20000 .029 2.262 1.1–4.7 
Sig.Odd ratio95% C.I
Transfusion dependent .007 3.783 1.43–10.02 
platelet <20000 .029 2.262 1.1–4.7 

Transfusion dependency prior to first cycle and platelets lower than 20,000 prior to each cycle, are the main significant risk factors for infections during azacitidine therapy. Neutropenia and age are known risk factors for infections in general, but were not significant in our study. We assume that in high risk MDS patients when most off the patients are old and neutropenic, thrombocytopenia is a surrogate marker of disease status which makes the patient more prone to infections. Therefore physicians should considerer these two parameters prior to every azacitidine cycle as guidance in the debate of concurrent prophylactic antibiotics, G-CSF or a tolerable dose of azacitidine. Our findings should be confirmed in a larger sample set but may pave the road for prospective studies of infection prophylaxis during azacitidine therapy.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution