Journal Club – Thalassemia

Cappellini MD, Viprakasit V, Georgiev P, et al. Long-term efficacy and safety of luspatercept for the treatment of anaemia in patients with transfusion-dependent β-thalassaemia (BELIEVE): final results from a phase 3 randomised trial. Lancet Haematol. 2025 Mar;12(3):e180-e189. doi: 10.1016/S2352-3026(24)00376-4. Epub 2025 Feb 10. PMID: 39947215.

Clinical Question

In transfusion-dependent β-thalassemia (TDT), does long-term luspatercept maintain durable reductions in red blood cell (RBC) transfusion burden and an acceptable safety profile?

Background

• β-thalassemia is an inherited hemoglobinopathy characterized by reduced or absent β-globin production, leading to α-chain excess, ineffective erythropoiesis, hemolysis, and chronic severe anemia.
• Patients with TDT require lifelong RBC transfusions, which correct anemia but lead to iron overload and progressive organ damage. Reducing transfusion burden is therefore a major therapeutic goal.
• Luspatercept, a TGF-β ligand trap that enhances late-stage erythroid maturation, is approved for adults with TDT based on the phase 3 BELIEVE trial, where 21% with luspatercept achieved ≥33% transfusion-burden reduction vs 5% with placebo (p<0.001), with low discontinuation and no treatment-related deaths.
• Because TDT requires lifelong management, longer-term data are essential to assess the durability of benefit and safety. This study reports approximately 3 years of follow-up from BELIEVE.

Guidelines

• Management of TDT consists of regular RBC transfusions targeting a pre-transfusion hemoglobin of 9–10.5 g/dL every 2–5 weeks, with iron chelation initiated when ferritin is ≥800–1000 ng/mL or after ~10–20 transfusions.
• These recommendations are endorsed by the Thalassemia International Federation (TIF) Guidelines, 2021, British Society for Haematology (BSH) Guidelines for Haemoglobinopathies, and the American Society of Hematology (ASH) Clinical Recommendations for Thalassemia Care.
• Hematopoietic stem cell transplantation (HSCT) and gene therapy are curative options for eligible patients, though access remains limited.

Study Design

• Design: Randomized, double-blind, placebo-controlled, phase 3 trial with crossover
• Setting: 65 sites in 15 countries
• Enrollment Timeline: May 2, 2016, and May 16, 2017
• Study Population: Adults (≥18 years) with a confirmed diagnosis of TDT, defined as requiring 6–20 RBC units with no transfusion-free period of >35 days, within 24 weeks before randomization
  • Key exclusion criteria:
    • Hemoglobin S/β-thalassemia or α-thalassemia (hemoglobin H disease) or prior exposure to luspatercept or sotatercept, a pulmonary hypertension drug with a somewhat similar MOA.Significant comorbidities such as HIV, active hepatitis B or C, poorly controlled diabetes mellitus, uncontrolled hypertension, recent deep vein thrombosis or stroke, major organ damage, or prior malignancy were also excluded.
    • Initiation of iron chelation therapy within ≤24 weeks before randomization
• Randomization: 2:1 allocation to luspatercept vs placebo
• Duration:
  • Initial blinded phase: 48–96 weeks
  • Extension: After unblinding, placebo patients could cross over to open label luspatercept
  • Follow-up: continued to 192 weeks (≈3.7 years) to assess long-term efficacy and safety
• Analysis Populations:
  • Intent-to-treat (all randomized) for primary 24-week and longer-term efficacy
  • The safety population included all individuals who received ≥1 dose of luspatercept (original and crossover)
• Endpoints:
  • Primary (original): ≥ 33% reduction in RBC transfusion burden over weeks 13–24 compared to baseline
  • Long-term (extension): durability of transfusion reductions (≥33% and ≥50% over any 12-week interval), iron-overload markers, and treatment adverse events

Populations

• 447 screened → 336 randomized (ITT): Luspatercept n=224 vs placebo n=112
• Safety population: Luspatercept n=223 vs placebo n=109
• Baseline characteristics: the two groups were well balanced (age, genotype, transfusion burden, Hb, splenectomy, iron overload); median age of 30 years (IQR 23–40) vs 30 years (24–39)
• Crossover: 92 placebo patients crossed over after unblinding for a total of 315 receiving luspatercept
• Median treatment exposure: Luspatercept: 153.6 weeks, crossover: 80.9 weeks, placebo: 74.7 weeks

Interventions

Luspatercept subcutaneously every 3 weeks, starting at 1.0 mg/kg with permitted escalation to 1.25 mg/kg if reduction in transfusion burden from baseline over ≥2 dose cycles <33% or, at the investigator’s discretion, ≥33% but ≤50% (maximum dose 120 mg) versus matched placebo on the same schedule.

• Median 33 doses in luspatercept (incl. crossover) vs 24 in placebo; 51% escalated to 1.25 mg/kg (maximum dose)
• All the patients also received best supportive care (transfusions + iron chelation)

Outcomes

• Transfusion Burden Reduction
  • Initial primary efficacy (weeks 13–24):
    • ≥33% reduction: 21% (48/224) luspatercept vs 5% (5/112) placebo, p<0.001
    • ≥50% reduction: 7% vs 2%, p=0.040
• Changes in iron parameters
  • Serum ferritin decreased –402 µg/L (Week 144) and –257 µg/L (Week 192)
  • LIC stable through Weeks 48–96, then improved; myocardial iron remained normal.
  • Iron chelation requirements decreased:
    • Deferasirox: –3.7 mg/kg (Week 96), –4.5 (Week 192)
    • Deferiprone: –6.2 (Week 96), –8.1 mg/kg (Week 192)
    • Deferoxamine: –7.2 (Week 96), –10.7 mg/kg (Week 192)
• Safety Outcomes
  • TEAEs: 97% (any), 56% (treatment-related), 33% (grade ≥3), 23% (serious)
  • Common AEs: URI, headache, back pain
  • Grade ≥3 events uncommon: anemia 3%, increased LIC 2%, bone pain 2%
  • Thromboembolic events: 13 patients (4%), all splenectomized with multiple risk factors
  • Extramedullary hematopoiesis: 5 patients (2%); two serious; two discontinuations; none in placebo/crossover
  • Deaths: 5 total (4 luspatercept, 1 placebo); none treatment-related

Commentary:

Luspatercept produces meaningful and durable reductions in transfusion burden for adults with TDT. Long-term BELIEVE data show that responses can persist for years, deeper reductions (≥50%) occur in a substantial minority, and improvements in iron parameters parallel the decreased transfusion need. The safety profile remained consistent with earlier findings, with most events being mild to moderate; rare complications such as extramedullary hematopoiesis and thromboembolic events—particularly in splenectomized patients—merit continued monitoring.

The iron outcomes reinforce the clinical value of treatment. Although LIC changes were modest early on, sustained declines in serum ferritin and reduced chelation requirements suggest a meaningful impact on transfusion-driven iron overload, the major driver of long-term morbidity in TDT.

Generalizability requires caution. BELIEVE enrolled a relatively young cohort and excluded patients with major organ damage, which is common in longstanding TDT. Thus, the trial population represents a healthier subset, and both efficacy and tolerability may differ in patients with heavier iron burden or established organ complications.

Overall, BELIEVE provides strong evidence that luspatercept offers sustained, clinically meaningful benefit in appropriately selected patients. Clinicians should consider trial selection criteria, set realistic expectations, and closely monitor high-risk individuals when applying these results to broader TDT populations.

Miriam A. Osei, MD, MPH, is a hematology fellow in the Dana Farber Cancer Institute/Mass General Brigham program. She completed medical school at Columbia Vagelos College of Physicians and Surgeons and a combined internal medicine and global health equity residency at Brigham and Women’s Hospital. Her clinical and research focus in hematology is on sickle cell disease and improving disparities.