Hemoglobin Level at Azacitidine Onset Is a Prognostic Factor of Unachievement of Three Azacitidine Cycles in Myelodysplastic Syndromes and Acute Myeloid Leukemia

Background: Azacitidine (AZA) is the first line for myelodysplastic syndromes (MDS) and acute myeloid leukemia with less than 30% of blasts (Fenaux et al., Lancet Oncology 2009). This is also the treatment of AML with more than 30% of blasts uneligible for intensive chemotherapy (Dombret et al., Blood 2015). Expected response rate is around 50-60%. 50% and more than 80% of AZA responses were observed after 3 cycles and 6 cycles, respectively. Therefore, the FDA/EMEA approved schedule is 75mg/m²/day subcutaneously for seven days every 28 days for at least 6 cycles to evaluate the efficiency of the drug. Unfortunately, a proportion of patients can not achieved 3 or 6 cycles and is not evaluated. The aim of this study was to define prognostic factors predicting unachievement of 3 AZA cycles.

Methods: We analyzed retrospectively MDS and AML patients treated by AZA in 5 centers. All patients receiving at least one cycle were included. Patients were treated with conventional regimen (75mg/m²/day subcutaneously for seven days every 28 days). Responses were scored according to IWG 2006 criteria for MDS and Cheson et al. (JCO 2003) for AML. Non continuous variables were compared used khi square test. Mann Whithney test was used for continuous variables.

Results: The study population included 290 patients: F/M: 159/131; median age 71 years (range 27-94). Diagnosis at AZA onset was MDS in 208 patients (RAEB-1 = 22, RAEB-2=95, RAEB-t=87) and AML with more than 30% of blasts in 55 patients. In MDS/RAEB-t, cytogenetic according to IPSS was good in 100, intermediate in 35, poor in 56 and unknown in 17 patients; and cytogenetic according to IPSS-R was very good in 1, good in 103, intermediate in 38, poor in 18 and very poor in 28 patients. IPSS was low in 5, intermediate 1 in 32, intermediate 2 in 69 and high in 101. IPSS-R was very low in 1, low in 12, intermediate in 46, high in 86 and very high in 60 patients. Median number of cycles was 6 (range 1-69) in MDS/RAEB-t and 5 (range 1-34) in AML. Overall response rate was 42% and 44%; and median overall survival (OS) was 13.4 months (range 10.7-16.0) and 7.8 months (range 2.7-12.8) in MDS/RAEB-t and AML, respectively.

In MDS/RAEB-t and AML, we did not observe a significant difference for age, sex, WHO classification, neutrophil count, platelet count, blast count. In MDS/RAEB-t only, we did not observe a significant difference for IPSS cytogenetic, IPSS-R cytogenetic, IPSS scoring and IPSS-R scoring between patients receiving less and more than 3 AZA cycles. Nevertheless, only hemoglobin (hb) level inferior to 9g/dl is significantly associated to unachievement of 3 AZA cycles. 61% versus 24% of patients did not achieve 3 AZA cycles if hb is inferior or superior to 9g/dl respectively (p=0.001). In AML, we also did not observe a significant difference for age, sex, WHO classification, neutrophil count, platelet count, blast count between patients receiving less and more than 3 AZA cycles. 50% vs 28% of patients did not achieve 3 AZA cycles if hb is inferior or superior to 9g/dl respectively (p=0.08). As expected, median OS was significantly lower in patients with less than 3 AZA cycles in MDS/RAEB-t patients (1.7 vs 13.4 months, p<0.0001) and in AML (3.8 vs 17.6 months, p<0.0001).

Conclusion: We identified hb level less than 9g/dl as a prognostic factor of unachievement of 3 AZA cycles in MDS/RAEB-t and AML. Like in AML with hyperleukocytosis, we could discuss to treat preferentially these patients with intensive chemotherapy instead of AZA in first line. Interestingly, IPSS and IPSS-R did not predict unachievement of 3 AZA cycles and were not good tools to predict short outcome.