RBC transfusion (tfx) independence was defined as the absence of any RBC transfusion during any consecutive "rolling" 56 days within the treatment period. Participants who did not receive any RBC transfusion during a consecutive rolling 56 days (i.e., day 1 to day 56, day 2 to day 57) were considered as a 56-day RBC transfusion independent responder.

The intent-to-treat (ITT) population included all participants who were randomized, regardless of whether they received treatment or not.

Duration of RBC transfusion independence was analyzed only for participants who achieved RBC transfusion independence of ≥ 56 days on treatment. Duration of RBC transfusion independence was defined as the time from the date transfusion independence is first observed (day 1 of a ≥ 56 days period without a transfusion) until the date the participants had a subsequently documented RBC transfusion. In the event a participant had more than one ≥56 days rolling periods which met the RBC independence criteria, the duration with the longest rolling period was used in the analysis. Participants who maintained RBC TI through the end of the treatment period were censored at the date of treatment discontinuation, death, or 1 day before the start of the subsequent MDS treatment (if any), whichever occurred first, or the particiapnts latest available assessment date in the database if the treatment was still on-going.

Intent to Treat population. Participants who achieved RBC transfusion independence of ≥ 56 days on treatment.

Time to RBC transfusion independence of ≥ 56 days was defined as the time between randomization and the date onset of transfusion independence was first observed (ie, Day 1 of 56 without any RBC transfusions).

Responders in the intent to treat population who achieved a 56-day TI response

A participant was considered as a RBC transfusion reduction responder if the participant had at least 4 units reduction in transfusion units over any consecutive 56 days period compared to the baseline transfusion units in 56 days.

Intent to Treat population; includes participants who achieved RBC transfusion reduction of at least 4 units for at least 8 weeks.

RBC transfusion independence was defined as the absence of any red blood cell (RBC) transfusion during any consecutive "rolling" 84 days within the treatment period. Participants who did not receive any RBC transfusion during a consecutive rolling 84 days (i.e., day 1 to day 84, day 2 to day 85) were considered as a 84-day RBC transfusion independent responder.

The ITT population includes all participants who were randomized, regardless of whether they received treatment or not.

Duration of RBC transfusion independence was analyzed only for participants who achieved RBC transfusion independence of ≥ 84 days on treatment. Duration of RBC transfusion independence was defined as the time from the date transfusion independence is first observed (day 1 of a ≥ 84 days period without a transfusion) until the date the participants had a subsequently documented RBC transfusion. In case a participant had more than one ≥84 days rolling periods which met the RBC independence criteria, the duration with the longest rolling period was used in the analysis.

Intent to Treat population. Participants who achieved RBC transfusion independence for at least 84 days on treatment.

Time to RBC transfusion independence of ≥ 84 days was defined as the time between randomization and the date onset of transfusion independence was first observed (ie, Day 1 of 84 without any RBC transfusions).

Participants who achieved a 84-day TI response. Responders in the intent to treat population.

Erythroid HI-E improvement was defined as a hemoglobin increase of ≥ 1.5 g/dL; or a reduction in units of RBC transfusions by an absolute number of at least 4 RBC transfusions/8 weeks compared with the pretreatment transfusion number in the previous 8 weeks. Only RBC transfusions given for a hemoglobin of ≤ 9.0 g/dL on treatment were counted in the RBC transfusion response evaluation.

The ITT population included all participants who were randomized, regardless of whether they received treatment or not.

HI-P response was defined according to IWG 2006 criteria (Cheson, 2006) and as: 1. Absolute increase of ≥ 30 X 10^9/L for participants^ starting with > 20 X 10^9/L platelets; 2. Increase from < 20 X 10^9/L to > 20 X 10^9/L and by at least 100%. HI-P must have lasted at least 8 weeks.

The ITT population included all participants who were randomized, regardless of whether they received treatment or not.

Platelet transfusion independence was defined as the absence of any platelet transfusion during any consecutive "rolling" 56 days during the treatment period, (ie, Day 1 to 56, Day 2 to 57, Days 3 to 58, etc.). Participants were considered platelet transfusion dependent at baseline if they had received ≥ 2 platelet transfusions during the 56 days immediately preceding randomization and had no consecutive 28-day period during which no platelet transfusions were administered.

ITT population; includes participants who were baseline platelet transfusion dependent.

Time to platelet transfusion independence was defined as the time between randomization and the first documented date of onset of transfusion independence (ie, Day 1 of 56 without any platelet transfusions).

Intent to Treat population; includes participants who achieved platelet transfusion independence for at least 56 days.

Overall survival was defined as the time from randomization to death from any cause and was calculated using randomization date and date of death, or date of last follow-up for censored participants. All subjects were followed until drop out (withdrawal of consent from further data collection or lost to follow-up), death, or study closure. Participants who dropped out or were alive at study closure (or at the time of the interim analysis) had their OS times censored at the time of last contact, as appropriate.

Intent to Treat population

Hematologic response was defined as: • A complete response (CR): <5% myeloblasts, and normal maturation of all cell lines; Peripheral blood (PB) shows: hemoglobin >10 g/dL, neutrophils ≥1.0x10^9/L, platelets ≥100x10^9/dL, blasts (0%) • Partial Response (PR): same as CR bone marrow (BM) shows blasts decreased by ≥ 50% over pre-treatment but still > 5%; Cellularity and morphology not relevant • Marrow CR: BM: ≤ 5% myeloblasts and decrease by ≥ 50% over pre-treatment PB • Stable disease (SD): failure to achieve at least PR, but no evidence of progression for > 8 wks • Failure: death during treatment or disease progression • Disease Progression for those with: - Less than 5% blasts: ≥ 50% increase in blasts to > 5% blasts - 5%-10% blasts:≥ 50% increase to > 10% blasts - 10%-20% blasts:≥ 50% increase to > 20% blasts - 20%-30% blasts ≥ 50% increase to > 30% blasts Any of the following: - ≥ 50% decrease from maximum remission/response in granulocytes or platelets

Population includes participants with a CR, PR and mCR who had baseline bone marrow blasts > 5%. ITT population.

Participants with a documented diagnosis of AML arising from previous MDS documented diagnosis.

The intent-to-treat (ITT) population included all participants who were randomized, regardless of whether they received treatment or not.

Time to AML progression was defined as the time from the date of randomization until the date the subject has documented progression to AML. For participants who had progression to AML documented in MLL central lab report, the earliest sample collection date with the diagnosis of "s-AML arising from previous MDS" was used as the date to AML progression.

The intent-to-treat (ITT) population who progressed to AML.

Clinically significant bleeding event was defined as: any intracranial or retroperitoneal bleed; bleeding requiring transfusions of > 2 units of blood/blood products; bleeding associated with a decrease in hemoglobin of > 2 g/dL; or bleeding from any site requiring transfusions of > 2 units of blood.

The intent-to-treat (ITT) population included all participants who were randomized, regardless of whether they received treatment or not.

A TEAE was defined as an adverse event that begins or worsens in intensity of frequency on or after the first dose of study drug through 28 days after last dose of study drug. A serious adverse event (SAE) is any: • Death; • Life-threatening event; • Any inpatient hospitalization or prolongation of existing hospitalization; • Persistent or significant disability or incapacity; • Congenital anomaly or birth defect; • Any other important medical event The investigator determined the relationship of an AE to study drug based on the timing of the AE relative to drug administration and whether or not other drugs, therapeutic interventions, or underlying conditions could provide a sufficient explanation for the event. The severity of an AE was evaluated by the investigator according to National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) (Version 4.0) where Grade 1 = Mild, Grade 2 = Moderate, Grade 3 = Severe, Grade 4 = Life-threatening and Grade 5 = Death.

The safety population includes all randomized participants who received at least one dose of study drug.

The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being (PWG), social/family (SWB), emotional well being (EWB) and Functional Well-Being (FWB) and an additional 20-item anemia questionnaire that measures fatigue associated items and 7 non-fatigue items. The scales are formatted on 1 to 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general health related quality of life (HRQoL), the FACT-An measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various therapeutic areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 indicates the poorest QOL and 100 denotes the highest QOL.

The Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being (PWG), social/family (SWB), emotional well being (EWB) and Functional Well-Being (FWB) and an additional 20-item anemia questionnaire that measures fatigue associated items and 7 non-fatigue items. The scales are formatted on 1 to 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general health related quality of life (HRQoL), the FACT-An measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various therapeutic areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 indicates the poorest QOL and 100 denotes the highest QOL.

The Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being (PWG), social/family (SWB), emotional well being (EWB) and Functional Well-Being (FWB) and an additional 20-item anemia questionnaire that measures fatigue associated items and 7 non-fatigue items. The scales are formatted on 1 to 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general health related quality of life (HRQoL), the FACT-An measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various therapeutic areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 indicates the poorest QOL and 100 denotes the highest QOL.

The Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being (PWG), social/family (SWB), emotional well being (EWB) and Functional Well-Being (FWB) and an additional 20-item anemia questionnaire that measures fatigue associated items and 7 non-fatigue items. The scales are formatted on 1 to 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general health related quality of life (HRQoL), the FACT-An measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various therapeutic areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 indicates the poorest QOL and 100 denotes the highest QOL.

The FACT-Anemia evaluable population was defined as all ITT participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being (PWG), social/family (SWB), emotional well being (EWB) and Functional Well-Being (FWB) and an additional 20-item anemia questionnaire that measures fatigue associated items and 7 non-fatigue items. The scales are formatted on 1 to 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general health related quality of life (HRQoL), the FACT-An measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various therapeutic areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 indicates the poorest QOL and 100 denotes the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being (PWG), social/family (SWB), emotional well being (EWB) and Functional Well-Being (FWB) and an additional 20-item anemia questionnaire that measures fatigue associated items and 7 non-fatigue items. The scales are formatted on 1 to 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general health related quality of life (HRQoL), the FACT-An measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various therapeutic areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 indicates the poorest QOL and 100 denotes the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-G and FACT-An score are summed to form the FACT-An total score. The FACT-An Trial Outcome Index (TOI) consists of the summation of a "summary scale" and includes the Physical Well-being, (PWB; 7 items; score range, 0-28), the Functional Well-being (7 items; score range, 0-28) and the Anemia subscale consisting of 20 items on the same five-point scale, with 13 of them measuring fatigue related symptoms (FS) and seven measuring non-FS. The FACT-An TOI has been demonstrated to be a sensitive indicator of clinical outcomes in a number of diseases including MDS. The Fact-TOI score ranges from 0 to 136. Higher scores on all scales of the Fact-An and subscales on the FACT-TOI reflect better quality of life or fewer symptoms.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-An is a 47-item, cancer-specific questionnaire consisting of a core 27-item general questionnaire (i.e., the Functional Assessment of Cancer Therapy-General [FACT-G]) The FACT-G measures the 4 domains on a 5-point scale ranging from 0 (not at all) to 4 (very much). The 4 domains are: • Physical Well-being (PWB; 7 items; score range, 0-28), • Social/Family Well-being (SWB; 7 items; score range, 0-28), • Emotional Well-being (EWB; 6 items; score range, 0-24), and • Functional Well-being (7 items; score range, 0-28). The FACT-G is a summation composed of a "summary scale" including the PWB, SWB, EWB and FWB. The FACT-G score range is from 0 to 108. For all summary scales including FACT-G, a higher score indicates better HRQoL or lower level of symptoms.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The FACT-G and the anemia subscale (AnS) are summed to form the FACT-An total score and the total score ranges from 0 to 188. The FACT-G measures the 4 domains on a 5-point scale ranging from 0 (not at all) to 4 (very much). The 4 domains are: • Physical Well-being (PWB; 7 items; score range, 0-28), • Social/Family Well-being (SWB; 7 items; score range, 0-28), • Emotional Well-being (EWB; 6 items; score range, 0-24), and • Functional Well-being (7 items; score range, 0-28). The AnS consists of 20 items on the same 5-point scale, with 13 of them measuring fatigue-related symptoms (FS) and 7 measuring non-FS. The AnS and FS scores can range from 0-80 and 0-52, respectively. For all domains and summary subscales, a higher score indicates better HRQoL or lower level of symptoms.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

A clinically meaningful improvement or deterioration was defined by domain specific thresholds of change from baseline. The FACT-An questionnaire is a 47-item, cancer specific questionnaire consisting of a core 27 items measuring 4 general domains physical well being, social/family, emotional well being and Functional Well-Being and an additional 20-item anemia questionnaire that measures fatigue and 7 non-fatigue items. The scales are formatted on 4 pages for self-administration using a 5-point Likert rating scale (0 = Not at all; 1 = A little bit; 2 = Somewhat; 3 = Quite a Bit and 4 = Very much). Also, general HRQoL measures the impact of fatigue and other anemia-related symptoms on patient functioning and is used to assess the effect of treatments in various areas, including MDS. The instrument and the fatigue and non-fatigue subscales are scored by summing points from all questions, then converting this sum to a 100 point scale; 0 = the poorest QOL and 100 = the highest QOL.

Health Related Quality of Life (HR-QoL) evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score at C1D1 and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The distribution (frequency and percentage) of the observed responses (i.e., "Not at all (0)," "A little bit (1)," "Somewhat (2)," "Quite a bit (3)," "Very much (4)," and missing) to Item GP-5 ("I am bothered by side effects of treatment" in the past seven days) of the FACT-An at each scheduled visit were summarized for each treatment group. The denominator for the percentage calculation per treatment group was based on the number of the FACT-An evaluable population at baseline. The distribution of change in responses (improved [i.e., change score from 1 to 4], no change [0], worsened by one level [-1], worsened by ≥2 levels [-2 to -4], and missing) from baseline at each post-baseline scheduled visit were summarized by treatment group.

The HRQoL evaluable population was defined as participants with a non-missing FACT-An Trial Outcome Index (TOI) score baseline visit and at least one post-baseline assessment visit.

The EQ-5D-3L is a generic, self-administered questionnaire that consists of 5 dimensions: mobility, self-care, pain, usual activities, and anxiety/depression. Each dimension has 3 levels of severity corresponding to no problems, some problems, and extreme problems. It also includes a Visual Analog Scale that recorded the respondent's self-rated health on a vertical, 0-100 scale, where 100 = Best imaginable health state and 0 = Worst imaginable health state. Distribution of the observed responses (i.e., no problems, moderate problems, severe problems, and missing) of the 5 dimensions at each visit was summarized per arm. The denominator for the percentage calculation per group was based on the number of the EQ-5D-3L evaluable population at baseline. The distribution of change in responses (i.e., improved [by ≥1 level], no change, worsened [by ≥1 level], and missing) from baseline are reported.

The HRQoL evaluable population was defined as participants with a EQ-5D-3L health utility at baseline and at least one post-baseline assessment visit.

The EQ-5D-3L is a generic, self-administered questionnaire that consists of 5 dimensions: mobility, self-care, pain, usual activities, and anxiety/depression. Each dimension has 3 levels of severity corresponding to no problems, some problems, and extreme problems. It also includes a Visual Analog Scale that recorded the respondent's self-rated health on a vertical, 0-100 scale, where 100 = Best imaginable health state and 0 = Worst imaginable health state. Distribution of the observed responses (i.e., no problems, moderate problems, severe problems, and missing) of the 5 dimensions at each visit was summarized per arm. The denominator for the percentage calculation per group was based on the number of the EQ-5D-3L evaluable population at baseline. The distribution of change in responses (i.e., improved [by ≥1 level], no change, worsened [by ≥1 level], and missing) from baseline are reported.

The HRQoL evaluable population was defined as participants with a EQ-5D-3L health utility at baseline and at least one post-baseline assessment visit.

TThe EQ-5D-3L is a generic, self-administered questionnaire that consists of 5 dimensions: mobility, self-care, pain, usual activities, and anxiety/depression. Each dimension has 3 levels of severity corresponding to no problems, some problems, and extreme problems. It also includes a Visual Analog Scale that recorded the respondent's self-rated health on a vertical, 0-100 scale, where 100 = Best imaginable health state and 0 = Worst imaginable health state. Distribution of the observed responses (i.e., no problems, moderate problems, severe problems, and missing) of the 5 dimensions at each visit was summarized per arm. The denominator for the percentage calculation per group was based on the number of the EQ-5D-3L evaluable population at baseline. The distribution of change in responses (i.e., improved [by ≥1 level], no change, worsened [by ≥1 level], and missing) from baseline are reported.

The HRQoL evaluable population was defined as participants with a EQ-5D-3L health utility at baseline and at least one post-baseline assessment visit.

The EQ-5D-3L is a generic, self-administered questionnaire that consists of 5 dimensions: mobility, self-care, pain, usual activities, and anxiety/depression. Each dimension has 3 levels of severity corresponding to no problems, some problems, and extreme problems. It also includes a Visual Analog Scale that recorded the respondent's self-rated health on a vertical, 0-100 scale, where 100 = Best imaginable health state and 0 = Worst imaginable health state. Distribution of the observed responses (i.e., no problems, moderate problems, severe problems, and missing) of the 5 dimensions at each visit was summarized per arm. The denominator for the percentage calculation per group was based on the number of the EQ-5D-3L evaluable population at baseline. The distribution of change in responses (i.e., improved [by ≥1 level], no change, worsened [by ≥1 level], and missing) from baseline are reported.

The HRQoL evaluable population was defined as participants with a EQ-5D-3L health utility at baseline and at least one post-baseline assessment visit.

The EQ-5D-3L is a generic, self-administered questionnaire that consists of 5 dimensions: mobility, self-care, pain, usual activities, and anxiety/depression. Each dimension has 3 levels of severity corresponding to no problems, some problems, and extreme problems. It also includes a Visual Analog Scale that recorded the respondent's self-rated health on a vertical, 0-100 scale, where 100 = Best imaginable health state and 0 = Worst imaginable health state. Distribution of the observed responses (i.e., no problems, moderate problems, severe problems, and missing) of the 5 dimensions at each visit was summarized per arm. The denominator for the percentage calculation per group was based on the number of the EQ-5D-3L evaluable population at baseline. The distribution of change in responses (i.e., improved [by ≥1 level], no change, worsened [by ≥1 level], and missing) from baseline are reported.

The HRQoL evaluable population was defined as participants with a EQ-5D-3L health utility at baseline and at least one post-baseline assessment visit.

The number of reasons for hospitalizations and hospital admissions during the treatment period were monitored and include those associated with: AEs, protocol-driven procedures, transfusions, non-protocol procedures, elective procedures or those associated with social, practical or technical reasons in the absence of AEs. HRU was defined as any consumption of healthcare resources directly or indirectly related to the treatment of the patient.

The safety population includes all randomized participants who received at least one dose of study drug.

The total number of days hospitalized due to any reason during the treatment period was monitored. HRU was defined as any consumption of healthcare resources directly or indirectly related to the treatment of the patient.

The safety population includes all randomized participants who received at least one dose of study drug.

The number of days hospitalized per total patient years. HRU was defined as any consumption of healthcare resources directly or indirectly related to the treatment of the patient.

The safety population includes all randomized participants who received at least one dose of study drug.