Chronic immune thrombocytopenic purpura (ITP) is characterized by low platelet counts and mucocutaneous bleeding. In previous studies romiplostim (AMG531), a thrombopoiesis-stimulating protein, increased platelet counts in most patients with chronic ITP. This ongoing, long-term open-label, single-arm study investigated safety and efficacy in patients who completed a previous romiplostim study and had platelet counts less than or equal to 50 [corrected] x 10(9)/L. One hundred forty-two patients were treated for up to 156 weeks (mean, 69 weeks). Platelet responses (platelet count > or = 50 x 10(9)/L and double baseline) were observed in 87% of all patients and occurred on average 67% of the time in responding patients. In 77% of patients, the romiplostim dose remained within 2 microg/kg of their most frequent dose at least 90% of the time. Ninety patients (63%) received treatment by self-administration. Treatment-related serious adverse events were reported in 13 patients (9%). Bone marrow reticulin was observed in 8 patients; marrows were not routinely performed in this study, so the true incidence of this event cannot be determined. Severe bleeding events were reported in 12 patients (9%). Thrombotic events occurred in 7 patients (5%). In conclusion, romiplostim increased platelet counts in most patients for up to 156 weeks without tachyphylaxis and had an acceptable safety profile. (ClinicalTrials.gov Identifier NCT00116688).