Abstract 3517

Poster Board III-454


Eltrombopag (PROMACTA®; GlaxoSmithKline, Collegeville, PA, USA) is an oral, small molecule (565 Da), thrombopoietin receptor agonist that has been approved in the United States for the treatment of patients with chronic immune thrombocytopenic purpura (ITP). It is also being studied in thrombocytopenic patients with chronic liver disease, hepatitis C, myelodysplastic syndromes, and cancer. Withdrawal of treatments that stimulate platelet production may theoretically result in recurrent thrombocytopenia below pretreatment levels (below baseline). OBJECTIVE: To determine whether worsening of thrombocytopenia (ie, platelet count decrease below baseline) occurs after discontinuation of eltrombopag in patients with chronic ITP. METHODS: The lowest median platelet counts during the first 4 weeks posttherapy were compared with median baseline platelet counts. Data from 369 patients treated in 3 randomized, double-blind, placebo-controlled studies were analyzed: TRA100773A and TRA100773B were 6-week studies, and RAISE was a 6-month study. For all 3 studies, a baseline platelet count <30,000/μL was required. Platelet counts, bleeding events, and the use of ITP medication were examined in the 4 weeks following the discontinuation of eltrombopag or placebo. A transient decrease in platelet counts (ie, worsening of thrombocytopenia) was defined as a platelet count below 10,000/μL and at least 10,000/μL below each patient's baseline platelet count (Bussel N Eng J Med 2006). RESULTS: Using pooled data from the 3 studies, no decreases below baseline median platelet counts (placebo, 16,300/μL; eltrombopag, 16,000/μL) were observed compared to the lowest median platelet counts within the first 4 weeks posttherapy (placebo, 14,000/μL; eltrombopag, 17,000/μL). Across the pooled studies, a total of 10/128 (8%) of placebo-treated patients and 20/241 (8%) of eltrombopag-treated patients had a transient decrease in platelet counts in the 4 weeks following discontinuation or interruption of treatment. None of the 10 placebo-treated patients had bleeding events associated with posttreatment platelet nadirs. Three of the 20 eltrombopag-treated patients had bleeding events and/or rescue treatment associated with the platelet nadir in the 4-week posttreatment period. One patient discontinued eltrombopag after achieving platelet counts >200,000/μL following on-therapy rescue medication (corticosteroid 0.5 mg/kg/day); 9 days after discontinuing study medication, the patient had grade 1 gum bleeding and resumed daily corticosteroids at an increased dose. The second patient had grade 3 menorrhagia and was administered vincristine (patient had a history of similar symptoms). The third patient had Henoch-Schoenlein purpura, interrupted eltrombopag due to platelet counts >400,000/μL, and 7 days after holding eltrombopag had a platelet count of 2000/μL, experienced grade 1 mouth hemorrhage and grade 2 petechiae, and did not require rescue medication. The patient continued in the study for the full 6 months and following permanent discontinuation of eltrombopag, this patient did not experience a transient decrease in platelet counts or any bleeding. CONCLUSION: Across 3 placebo-controlled studies, the incidence of transient decreases in platelet counts following discontinuation or interruption of study medication was similar in patients receiving eltrombopag or placebo. Therefore, these decreases may be unrelated to study medication and may represent normal fluctuations in platelet counts in patients with chronic ITP. Transient platelet count decreases were generally not associated with bleeding events.


Cheng:GlaxoSmithKline: Research Funding. Tarantino:GlaxoSmithKline: Speakers Bureau; Lundbeck: Speakers Bureau; Baxter: Membership on an entity's Board of Directors or advisory committees. Gernsheimer:GlaxoSmithKline: Honoraria, Research Funding; Amgen: Honoraria, Research Funding. Meyer:GlaxoSmithKline: Consultancy, Honoraria. Brainsky:GlaxoSmithKline: Employment. Stone:GlaxoSmithKline: Employment.

Author notes


Asterisk with author names denotes non-ASH members.