Introduction:Romiplostim (RMP) is a thrombopoietin-mimetic protein that enhances platelet production. RMP has already been approved in numerous countries for treatment of immune thrombocytopenia. We previously reported the efficacy and safety of RMP through a 1-year treatment period in patients with aplastic anemia (AA) that was refractory to immunosuppressive therapy (IST). In this study, we report the interim exploratory analysis of the influence of RMP after a 2-year treatment period (105 weeks) on the hematopoietic stem cells (HSCs) and progenitor cells, and on mesenchymal stromal cells, a specialized HSC niche, in patients with AA who were refractory to IST.Methods:This study is a multicenter, randomized, open-label, parallel, comparative, dose-finding study in South Korea (NCT02094417). Eligible patients with AA, refractory to prior IST, were randomized into 4 dose groups that received 1, 3, 6, or 10 μg/kg of RMP for the initial 8-week treatment. After the initial treatment, the dose was adjusted from 1 to 20 μg/kg to maintain their platelet response. After week 53, if the blood cell count improved and was maintained at a predefined range, the dose was tapered and the patient was considered to be in a “tapering-dose” status; if the dose was decreased to 0 µg/kg at this point, the patient was considered to be off the drug. The range was defined as follows: platelet count, > 50x109/L; hemoglobin concentration, > 10.0 g/dL; and neutrophil count, > 1.0x109/L. The blood cell response was defined based on the absolute cell count. To examine the association between cellular changes in the bone marrow in response to RMP and the clinical course of the patients with AA, the primitive HSC and progenitor cell subsets were analyzed by flow cytometry and the colony formation assay at the baseline and every 6 months after initiation of RMP. Further, changes in the mesenchymal stromal cells within the bone marrow were also analyzed by the CFU-F assay.Results:A total of 35 subjects were randomized; of these, 18 received RMP beyond 1 year, and 10 are still receiving it over 2 years.At week 105, the platelet response was observed in all 10 subjects, erythroid response in 9 subjects, neutrophil response in 5 subjects, and tri-lineage response in 5 subjects.Of the ongoing 10 subjects, 7 have achieved “tapering-dose” status, and 2 have achieved “off-the-drug” status. In addition, The RMP was suspended in 1 subject due to excessive increase in platelet-count (>400x109/L). This subject has also been exhibiting the tri-lineage response.At week 53, the median counts of all colony assays were higher in subjects with any blood cell response than in those without the response. When we compared the serial colony assays in the subjects with any blood cell response at week 53 up to week 105, the colony counts of BFU-E, CFU-E, CFU-M, and CFU-GM gradually increased until week 77 from the baseline, and then declined at week 105. Further, the colony counts of CFU-F also exhibited a similar pattern. The colony counts of CFU-G and CFU-MK increased until week 53, and then decreased. However, the CD34+CD38- cells steadily increased throughout up to week 105 from the baseline.Conclusion:These data indicate that RMP influences the stimulation of primitive HSC (CD34+CD38- cells) and bone marrow microenvironment (HSC niche), and enhances the differentiation of primitive HSCs to late progenitor cells resulting in hematologic response in the patients with AA.Although further follow up is ongoing the up to the end of 3-year study period, but this interim analysis may strongly implies that the RMP might be beneficial for patients with AA, refractory to IST. DisclosuresNo relevant conflicts of interest to declare.