Abstract Introduction Romiplostim was approved in 2009 in Europe for treatment of adult chronic immune thrombocytopenia (cITP) in splenectomized patients, refractory to other ITP treatments, and as second-line treatment for adult non-splenectomized patients. In 2016 the indication was extended to second-line treatment in all cITP patients. Few data exist on how romiplostim is used in routine clinical care. Methods To examine romiplostim treatment patterns we used the Nordic Country Patient Registry for romiplostim (NCPRR). NCPRR was established in 2009 and includes all adult (≥18 years) patients in Denmark, Sweden and Norway with a confirmed cITP (ITP lasting >6 months) diagnosis requiring hospital contact. The cohort is based on data from national health registries and enriched by medical record review. All patients diagnosed with cITP from April 1, 2009 to December 31, 2016 (data cut-off) are included. We describe age, comorbidity, previous treatment, platelet level, and both romiplostim dose and duration in patients who started romiplostim after date of their cITP diagnosis by line of treatment. Results Among 2895 patients diagnosed with cITP: 103 patients started romiplostim treatment before the data cut-off. Of these, 40% were aged 18- 50 years old, 30% were 35-70 years old, and 30% were ≥71 years. A total of 76% had no recorded comorbidity. Romiplostim was first-line treatment in 8 cITP patients of whom one had been splenectomized. In the month before romiplostim start, six of these patients (75%) had platelet count <50x109/L, but only one patient had experienced clinically significant bleeding (see Table). Twenty cITP patients started romiplostim as second-line treatment: duration of non-romiplostim first-line treatment had been <3 months in 50%, 3-12 months in 25%, and ≥12 months in 25%. Fourteen (70%) second-line romiplostim patients had obtained a platelet count of >150x109/L during their first-line treatment. In the month prior to romiplostim initiation, 3 (15%) patients experienced a bleeding event, while median latest platelet count was 10 x109/L (inter quartile range, IQR :10, 38). Romiplostim was third-line treatment in 44 patients: in the month prior to initiation, median latest platelet count was 15 x109/L (IQR: 10-32), and 20 (46%) patients experienced clinically significant bleeding events. Duration of non-romiplostim second-line treatment was ≤1 month in 33 patients (75%) and between 1-6 months in 20%. During their second-line treatment, 17 (39%) patients did not reach a platelet count of >50x109/L and 15 (34%) had a highest platelet count of 50 to 150x109/L. In the remaining 31 romiplostim-treated patients, it was used as fourth or later treatment line: median latest platelet count in the month preceding initiation was 14 x109/L (IQR: 5-27), with 7 (22.6%) patients experiencing clinically important bleeding during this time. Duration of the previous treatment in these patients had been ≤1 month in 16 (52%) and 1- 3 months in 12 (39%) patients. During their previous line of treatment 14 (45%) patients did not reach a platelet count of >50x109/L and an additional 9 (29%) had a highest platelet count of 50 to 150x109/L. In patients initiating romiplostim at first-line, median maximum platelet count while on therapy was 147x109/L (IQR: 109-237): this value was 299x109/L (IQR: 187,752), 295x109/L (IQR: 107,454), and 132x109/L (IQR: 52-305) for second, third, and fourth-or-later lines respectively. Median duration of romiplostim therapy was shortest at first-line (37 days, IQR: 21-180), and longest at second-line (91 days, IQR: 21, 169). Two patients on second-line, and 8 patients on third-line eltrombopag, switched to romiplostim. Conclusion Approximately 4% of cITP patients were treated with romiplostim, predominantly at third or later treatment lines: median platelet counts were seen to improve from <20 x109/L prior to romiplostim initiation to >100x109/L while on therapy across all lines. Romiplostim treatment had a relatively short duration. However, romiplostim-treated patients were characterized by a short duration on their previous non-romiplostim treatment line, and a high proportion had low platelet counts during this prior treatment. These data indicate that romiplostim is effective at increasing platelet counts in cITP patients with varying clinical history. Larger studies are needed to confirm these results and investigate drivers at different lines of therapy. Disclosures Bahmanyar: Amgen: Research Funding. Ghanima:GlaxoSmithKline and Pfizer: Other: Personal Fees; Roche, Amgen, Novartis, Bayer, BMS: Other: Personal Fees, Research Funding. Alam:Amgen: Employment, Equity Ownership. Christiansen:Amgen: Research Funding.