Kaplan-Meier estimates for the median months until death from any cause within the intent-to-treat population. Patients surviving at the end of the follow-up period were censored at the date of last contact. If a patient withdrew consent to follow-up or was lost to follow-up, the patient was censored as of the last date of contact.

Intent to treat population. Includes participants who died and participants who were censored.

Kaplan-Meier estimates for the median months until death from any cause within the intent-to-treat population. Patients surviving at the end of the follow-up period were censored at the date of last contact. If a patient withdrew consent to follow-up or was lost to follow-up, the patient was censored as of the last date of contact.

Subgroups that were analyzed are age, gender, French-American-British (FAB) classification, World Health Organization (WHO) classification and International Prognostic Scoring System (IPSS) classification.

Intent to treat population. Includes participants who died and those who were censored.

The time to transformation to AML or death from any cause (whichever occurred first) was defined as the number of days from the date of randomization until the date of documented AML transformation or death from any cause. Patients who did not transform to AML or die were censored at the date of last follow-up.

Intent to treat population. Participants who either transformed to AML or died are 120 for azacitidine and 132 for conventional care. Remaining participants were censored.

The time to transformation to AML was defined as the number of days from the date of randomization until the date of documented AML transformation, defined as a bone marrow blast count ≥ 30% independent of baseline bone marrow count. Patients who did not transform to AML were censored at the date of last follow-up or date of death.

Intent to treat population. Participants who were transformed to AML are 78 for azacitidine and 71 for conventional care. Remaining participants were censored based on the last bone marrow assessment.

Summary of dependence and independence from red blood cell (RBC) transfusion at baseline and during treatment, for patients who were dependent at baseline. A patient was considered transfusion independent at baseline if the patient had no transfusions during the 56 days prior to randomization. During study, a patient was considered transfusion independent during the on-treatment period if the patient had no transfusions during any 56 consecutive days or more. Otherwise, the patient was considered transfusion dependent.

Intent to treat population

Summary of dependence and independence from red blood cell (RBC) transfusion at baseline and during treatment, for patients who were independent at baseline. A patient was considered transfusion independent at baseline if the patient had no transfusions during the 56 days prior to randomization. During study, a patient was considered transfusion independent during the on-treatment period if the patient had no transfusions during any 56 consecutive days or more. Otherwise, the patient was considered transfusion dependent.

Intent to treat population

Summary of dependence and independence from platelet transfusion at baseline and during treatment for patients who were dependent at baseline. A patient was considered transfusion independent at baseline if the patient had no transfusions during the 56 days prior to randomization. During study, a patient was considered transfusion independent during the on-treatment period if the patient had no transfusions during any 56 consecutive days or more. Otherwise, the patient was considered transfusion dependent.

Intent to treat population

Summary of dependence and independence from platelet transfusion at baseline and during treatment for patients who were independent at baseline. A patient was considered transfusion independent at baseline if the patient had no transfusions during the 56 days prior to randomization. During study, a patient was considered transfusion independent during the on-treatment period if the patient had no transfusions during any 56 consecutive days or more. Otherwise, the patient was considered transfusion dependent.

Intent to treat population

Investigator determined responses followed IWG criteria for

• complete remission(CR): repeat bone marrow show <5% myeloblasts, and peripheral blood evaluations lasting >=2 months of hemoglobin(>110 g/L), neutrophils(>=1.5x10^9/L), platelets(>=100x10^9/L), blasts (0%) and no dysplasia
• partial remission(PR) is the same as CR for peripheral blood: bone marrow shows blasts decrease by >=50% or a less advanced FAB classification from pretreatment
• stable disease(SD) is a failure to achieve at least a partial remission, but with no evidence of progression for at least 2 months.

Intent to treat population.

IWG 2000 Criteria: Pretreatment=hemoglobin <100g/L or RBC transfusion-dependent, platelet count <100x10^9/L or platelet transfusion dependent, absolute neutrophil count <1.5x10^9/L.

Erythroid response: Major->20g/L increase or transfusion independent. Minor- 10-20g/L increase or >=50% decrease in transfusion requirements.

Platelet response: Major-absolute increase of >=30x10^9/L or platelet transfusion independence. Minor->=50% increase.

Neutrophil response: Major->=100% increase or an absolute increase of >0.5x10^9/L. Minor->=100% increase and absolute increase of <0.5x10^9/L.

Intent to treat population.

The time to disease progression, relapse after complete or partial remission (CR, PR), or death from any cause was defined as the time from the date of randomization until the first date of documented disease progression, relapse after CR or PR, or death from any cause.

Intent to treat population. The number of participants with disease progression, relapse after remission or death from any cause is 84 for azacitidine and 79 for conventional care. Remaining participants were censored.

The duration of improvement was defined as the time from the date of hematologic improvement until the date of first documented progression or relapse after hematologic improvement or death from any cause.

Intent to treat population. Participants showing hematologic improvement were 48 in azacitidine and 31 in conventional care.

The on-treatment adverse event rate of infection requiring IV antibiotics, antifungals, or antivirals per patient-years. The on-treatment period was considered the period from the date of randomization to the last treatment study visit.

Intent to treat population

Patient counts for a variety of subsets of adverse experiences for the core study period (day 1 to 42 months). The individual options for Conventional Care Regimens (Best Supportive Care Only, Low-Dose Cytarabine, and Standard Chemotherapy) are presented as separate treatments.

Safety population excludes 4 Azacitidine patients, 3 Best Supportive Care Only patients, 5 Low-dose Cytarabine patients, and 6 Standard Chemotherapy patients who were randomized/assigned to those regimens but did not receive treatment.

A sensitivity analysis of time to transformation to AML during the entire study was performed based on the last bone marrow assessment. Patients were censored based on the last bone marrow assessment.

Intent to treat population. Participants who were transformed to AML are 78 for azacitidine and 71 for conventional care. Remaining participants were censored based on the last bone marrow assessment.

Count of participants who died during the study

Intent to treat population