Background: Anemia is common in myelofibrosis (MF) and current treatments are unsatisfactory. Anemia is not improved, and often initially worsened by ruxolitinib (rux), the only approved treatment for MF. Sotatercept is a first-in-class, activin receptor IIA (ActRIIA) “ligand trap” that sequesters ligands that bind to the ActRIIA, some of which inhibit terminal erythropoiesis. Aims: To evaluate the efficacy and safety of sotatercept in anemic subjects with MF, alone and with rux. Methods: This ongoing, single-institution, investigator-initiated, phase 2, open-label trial (ClinicalTrials.gov Identifier: NCT01712308) enrolls adults with primary MF or MF arising from polycythemia vera or essential thrombocythemia with anemia with or without RBC transfusion-dependence (per International Working Group for Myelofibrosis Research and Treatment criteria). Sotatercept is given subcutaneously every 3 weeks. Subjects on the sotatercept-only arm receive 0.75 or 1 mg/kg. Subjects on a stable dose (for ≥8 weeks) of rux receive 0.75 mg/kg of sotatercept and must have been on rux for ≥6 months. Responses include increases in hemoglobin (Hgb) of ≥1.5 g/dL (for ≥12 weeks) from baseline in anemia-only subjects (without transfusions) and RBC transfusion independence in RBC transfusion-dependent subjects. Subjects had to be on treatment for ≥84 days to be evaluable. Results: Of 48 subjects enrolled, 6 never began sotatercept and 1 received 0.3 mg/kg/dose; these subjects are not considered further. Of the remainder, 26 subjects received sotatercept only (12 received 0.75 mg/kg and 14, 1 mg/kg) and 15, sotatercept and rux. Baseline variables are displayed in the Table. Seven of 20 (35%) evaluable subjects in the sotatercept-only cohort responded; 4 of 10 anemia-only subjects and 3 of 10 RBC-transfusion-dependent subjects. Two other subjects who achieved ≥1.5 g/dL Hgb increases from baseline discontinued early for other reasons. Of 13 evaluable subjects receiving sotatercept and rux, 3 (23%) responded (all anemia-only subjects). An additional subject had a 1.5 g/dL Hgb increase from baseline that was not sustained for ≥84 days due to the need for a rux dose increase from 10 to 15 mg twice daily. The median numbers of cycles of sotatercept were 5 (range, 1–53) and 10 (range, 2–26), and median times on study were 4 (range, 1–39+) and 7 (range, 2–31+) months, respectively, in the sotatercept-only and sotatercept plus rux cohorts. Median time to response in the monotherapy cohort was 21 (range, 7–22) days and median duration of response was 12 (range, 4–39+) months. In the combination cohort, the 3 responses began at 7, 14 and 140 days, and are ongoing at 17+, 18+ and 30+ months. Several subjects in both cohorts required sotatercept to be held for Hgb ≥11.5 g/dL and resumed when Hgb <11 g/dL. Eight subjects remain on study, 4 in each cohort, including 2 responders in the sotatercept-only cohort and 3 in the combination cohort. Reasons for discontinuation (n = 33) include no or loss of benefit (11), censoring because of a transplant (6), MF progression (7), patient choice (3), inability to comply with study requirements (2), hypertension (1), leukemic transformation (1), loss of health insurance (1) and unrelated medical problems (1). Sotatercept was well tolerated. Most adverse events (AEs) were grades 1 or 2. Grade 3 myalgias/arthralgias occurred in 2 subjects (5%). Grade 3 hypertension occurred in 6 subjects (15%) and grade 2 in 2 (5%). These were not associated with high Hgb values. No grade 4 AEs occurred.Summary/Conclusion: Sotatercept improves anemia of MF, both when given alone and in subjects receiving rux.