Eltrombopag

Frequency of all adverse events (including Ophthalmic events) categorized using CTCAE toxicity grades and clinical laboratory test \[ Time Frame: Up to Week 4 Follow-up period \] Clinical laboratory assessments and frequency of all adverse events, categorized using Common Terminology Criteria for Adverse Events (CTCAE) toxicity grades will present safety and tolerability endpoints Serious Adverse Events are below. See All Adverse Events in the following section for specifics No statistical analysis was planned for this primary outcome

The number of participants (responders) with platelet count \>=50x10\^9/L after 6 weeks of Stage 1 were compared between treatments using a logistic regression model adjusted for use of primary immune thrombocytopenia (ITP) medication at Baseline (yes/no), splenectomy (yes/no), Baseline platelet count \<=15×10\^9/L (yes/no) and treatment. Complete blood count including platelet count was done at Week 1, Week 2, Week 3, Week 4, Week 5 and Week 6. Primary Analysis Data Set: a participant who withdrawals from Stage 1 or is emergently unblinded was classified as a negative response from the time of withdrawal or unblinding date and for all subsequent visits. In the event of a participant dying, information for all subsequent assessments would be considered missing. All intermittent missing data (apart from withdrawals) will be treated as missing.

overall survival rate at month 12 (year 1) between the two arms, with or without 200 mg of Eltrombopag daily after induction chemotherapy.

Platelet transfusion independence is defined as the absence of platelet transfusion for at least 8 weeks. Platelet count should be higher than the baseline count.

Complete response rate was defined as percentage of patients achieving complete response (CR). Complete response was defined as subjects meeting all the three criteria on two consecutive serial blood count measurements at least one week apart but not more than four weeks apart: * Absolute neutrophil count \> 1.0 ×10\^9/L * Platelet count \> 100 ×10\^9/L * Hemoglobin \> 100 g/L The participants who discontinued from the trial before Week 26 and those that received blood products prior to response assessment (7 days prior for platelet transfusions, 14 days for RBC transfusion and 21 days for growth factors) were treated as non-responders.

Sustained response off treatment at 52 weeks is defined as maintenance of platelet count ≥ 30 × 10\^9/L after treatment discontinuation until Week 52 in the absence of bleeding events ≥ Grade II or use of any rescue medication at all visits until Week 52. Bleeding events ≥ Grade II are assessed by the modified World Health Organization (WHO) Bleeding Scale; Bleeding is graded based on a 1-4 scale (1=minor bleeding to 4=severe bleeding). Characteristics of bleeding events ≥ Grade II include: epistaxis ≥30 minutes, large purpura, joint bleeding, melanotic stool, hematemesis, gross hematuria, abnormal vaginal bleeding, hemoptysis, Visible blood in body cavity fluid, retinal bleeding, bleeding at invasive sites, bleeding requiring transfusion, bleeding associated with moderate or severe hemodynamic instability, fatal bleeding, CNS bleeding.

Hematologic response rate: percentage of subjects who met any of the International Working Group criteria: Platelet count: If platelet transfusion independent at baseline (BL): Transfusion independent and increase from BL by 20×10\^9/L or more; If platelet transfusion dependent at BL: No platelet transfusion requirement for 8 weeks; Hemoglobin: If red blood cells (RBC) transfusion independent at BL: transfusion independent and increase from BL by 15 g/L or more; If RBC transfusion dependent at BL: A decrease of at least 4 units in RBC transfusions in the post-treatment 8-week period (1 unit = RBC derived from 200 mL blood) or no RBC transfusion requirement for 8 weeks; Neutrophil count: In the absence of granulocyte colony stimulating factor (G-CSF) taken within 21 days preceding the blood sample collection: Increase from BL by 0.5×10\^9/L or more, or (if \< 0.5×10\^9/L at BL) increase by 100% or more. Patients who discontinued from the trial before week 26 were treated as non-responders.

Sustained response off treatment (SRoT) was defined as: reaching platelet count \>= 100×10\^9/L (complete response \[CR\]) and then maintaining platelet counts around 100×10\^9/L for 2 months (no counts below 70×10\^9/L), AND then tapering off the drug until treatment discontinuation while maintaining platelet count \>= 30×10\^9/L in the absence of bleeding (no bleeding AEs) or use of any rescue therapy until month 12.

AUC tau: Area under the curve calculated to the end of the dosing interval ( tau) (mass\*time/volume) AUC last: Area under the curve calculated to the last quantifiable concentration point (Tlast) (mass\*time/volume)

Cmax is the observed maximum plasma concentration following administration (mass/volume)

Ctrough is the pre-dose plasma concentration (mass/volume).

Overall hematologic response rate by 6 months was defined as the percentage of participants with complete response (CR) or partial response (PR) any time on or before 6 months. PR was defined as any two of the following parameters at two consecutive measurements at least 7 days apart and no platelet transfusion within 7 days of platelet measurement: * Absolute neutrophil count (ANC) \>500/μL * Platelet count \>20 000/μL * Reticulocyte count \>60 000/μL CR was defined as all three parameters meet the following criteria at two consecutive measurements at least 7 days apart and no platelet transfusions within 7 days of platelet measurement and no red blood cell transfusion with 14 days of the hemoglobin measurements: * ANC \> 1 000/μL * Platelet count \>100 000/μL * Hemoglobin \>10 g/L

ORR will be calculated after 6 months of eltrombopag administration by measuring platelet, reticulocyte, neutrophil and transfusion independence. ORR includes Complete Response (CR) and Partial Response (PR) Rate.

Hematologic response rate at 6 months (at WeeK 26) after the start of study treatment was defined as the percentage of subjects who met any of the response criteria (Platelet count, Hemoglobin level, Neutrophil count). 1 ) Platelet count: an increase from baseline by 20,000/μL or more (In the absence of platelet transfusion), or no platelet transfusion requirements for 8 weeks; 2) Hemoglobin: When the baseline hemoglobin level is \<9 g/dL: Without RBC transfusion at baseline, an increase from baseline by 1.5 g/dL or more; With RBC transfusion at baseline, a decrease of at least 4 units in RBC transfusions in the post-treatment 8-week period compared to the pre-treatment 8-week period (1 unit = RBC derived from 200 mL blood); 3) Neutrophil count: an increase from baseline by 500/μL or more (in the absence of granulocyte colony-stimulating factor (G-CSF)), or (if \< 500/μL at baseline) an increase by 100 % or more. Only descriptive analysis done.

To determine if patients with chronic ITP who do not respond to 75 mg of eltrombopag daily given for at least 3 weeks but then do respond to eltrombopag given daily first for 2 weeks at doses of 100, then for 2 weeks at 125 mg and finally for 4 weeks at a dose of 150mg daily. Response will be defined as 2 consecutive platelet counts of \> 50,000 with an increase of \> 20,000 from the study baseline within the 8 week increased dose window not as a result of rescue treatment.

A participant was considered as a responder if he/she met the following response criteria: a Baseline platelet count \<20 Giga cells per liter (Gi/L) and a post-Baseline increased to \>20 Gi/L and at least 2 times the Baseline value; or a Baseline platelet count \>=20 Gi/L and a post-Baseline absolute platelet count increased to \>=50 Gi/L and at least 2 times the Baseline value. The response criteria was evaluated at each visit. Increase in platelet count observed up to 3 days after a platelet transfusion was not considered as a platelet response. The Part 1 Population was comprised of all participants enrolled into Part 1.

A participant was considered to have a CRTE at a given assessment if he/she had platelet counts \<10 Gi/L, or platelet transfusions, or \>=Grade 3 hemorrhagic adverse events. CRTEs during Weeks 5 to 12 were compared between treatments using a generalized linear mixed model. Average of weekly proportion of subjects with CRTE during Week 5 to 12 was estimated for each treatment. Intent to Treat (ITT) Population was comprised of all randomized participants during Part 2.