Background:The β‐thalassemias are a group of inherited disorders characterized by absent or reduced production of the β‐globin chains of hemoglobin (Hb), leading to ineffective erythropoiesis, chronic anemia, and multiple morbidities, and often require lifelong RBC transfusions. Increasing HbF is a therapeutic approach to rescuing the defect imposed by reduced/absent β‐globin in β‐thalassemias. Luspatercept is a first‐in‐class erythroid maturation agent that binds to select TGF‐β superfamily ligands and enhances late‐stage erythropoiesis by a mechanism that is still not fully understood. The phase 3, randomized, double‐blind, placebo‐controlled BELIEVE study evaluated the efficacy and safety of luspatercept in adult β‐thalassemia patients requiring regular RBC transfusions. In the BELIEVE study, treatment with luspatercept resulted in significant reductions in RBC transfusion burden compared with placebo.Aims:To explore the effects of luspatercept treatment on HbF levels in patients with RBC transfusion‐dependent (TD) β‐thalassemia, and to explore the relationship of HbF changes with transfusion burden reduction.Methods:In the BELIEVE study, 336 adults with TD β‐thalassemia were randomized to receive either luspatercept or placebo, administered subcutaneously every 21 days for 48 weeks, plus best supportive care. Hb variants were analyzed by HPLC and HbF was reported as a percentage of total Hb from whole blood samples collected at baseline and throughout the treatment period. P values were derived by 2‐tailed t‐test (parametric method) and Wilcoxon test (non‐parametric method). Differences in HbF changes during treatment were compared between responders and non‐responders (as defined by transfusion burden decrease of ≥33% over any 12 weeks), patients with baseline HbF≤1% and HbF>1%, and luspatercept‐ and placebo‐treated patients.Results:HbF levels increased by a mean 1.2‐fold in luspatercept‐treated patients beginning at Dose2 Day1, increasing to 2.5‐fold by the end of the evaluation period (Dose16 Day1). Mean change in HbF levels remained relatively unchanged in placebo‐treated patients through Dose16 Day1, ranging from 0.9‐ to 1.2‐fold of baseline levels.Among luspatercept‐treated patients starting at Dose5 until Dose16 Day1, HbF fold increase was greater in responders vs non‐responders (Dose5 Day1: 2.1 vs 1.8 mean fold increase, P < 0.024, Dose16 Day1: 2.7 vs 2.1 mean fold increase, P = 0.012). To determine whether changes in HbF were secondary to RBC transfusion reduction or a direct effect of luspatercept, a subgroup analysis was performed in patients with normal HbF (HbF≤1%) at baseline. In luspatercept‐treated patients with normal HbF at baseline, HbF increased by a mean 1.4‐fold in responders and 1.5‐fold in non‐responders, beginning at Dose2 Day1, and continued to increase to 3.8‐fold in responders and 2.9‐fold in non‐responders by Dose16 Day1, with no statistically significant difference between HbF changes in responders and non‐responders (Figure). HbF levels were not modulated in placebo‐treated patients. In luspatercept‐treated patients with elevated (>1%) levels of HbF at baseline, HbF levels increased more in responders compared with non‐responders (2.8 vs 2.0 mean fold increase, P = 0.009).Summary/Conclusion:Luspatercept treatment was associated with increased HbF in patients with RBC TD β‐thalassemia. This luspatercept treatment effect on increasing HbF levels was observed in both responders and non‐responders. Luspatercept‐mediated increases in HbF were observed early and maintained throughout the treatment period.image