Luspatercept

What is Luspatercept?

Luspatercept (Reblozyl, luspatercept-aamt) is a recombinant fusion protein consisting of a modified extracellular domain of human activin receptor type IIB (ActRIIB) linked to the Fc region of human IgG1. Originally developed by Acceleron Pharma as ACE-536, the molecule was specifically engineered with amino-acid substitutions in the ActRIIB ligand-binding domain that bias its affinity toward GDF11, activin B, and select related TGF-β superfamily ligands while reducing affinity for activin A — a deliberate design choice intended to target the ligands that suppress late-stage erythroid maturation without provoking the broader off-target effects (notably the BMP9/10-mediated vascular signal that halted the myostatin-focused ACE-031 program). Luspatercept is FDA-approved as Reblozyl, marketed by Bristol-Myers Squibb (which absorbed Celgene, Acceleron's commercial partner) and Merck (which acquired Acceleron in November 2021). Approved indications are: transfusion-dependent anemia in adults with β-thalassemia (FDA, November 2019, on the basis of the Phase 3 BELIEVE trial); anemia in lower-risk myelodysplastic syndromes (MDS) with ring sideroblasts after erythropoiesis-stimulating agent failure (FDA, April 2020, on the basis of the Phase 3 MEDALIST trial); and, since August 2023, first-line treatment of anemia in ESA-naive transfusion-dependent lower-risk MDS (on the basis of the Phase 3 COMMANDS trial).

What Luspatercept Is Investigated For

Luspatercept is one of the cleanest case studies in the activin-receptor-trap class: a deliberately retuned ActRIIB-Fc designed to capture only the ligands relevant to erythroid maturation, with an FDA-approval trail to match. The pivotal evidence is three Phase 3 trials. BELIEVE (Cappellini et al., NEJM 2020) randomized 336 adults with transfusion-dependent β-thalassemia to luspatercept 1.0 mg/kg subcutaneously every three weeks vs placebo; 21.4% of luspatercept patients vs 4.5% of placebo patients achieved at least a 33% reduction in transfusion burden over weeks 13–24 — establishing a durable transfusion-sparing effect in a disease with no other approved disease-modifying therapy at the time. MEDALIST (Fenaux et al., NEJM 2020) randomized 229 patients with lower-risk MDS with ring sideroblasts who had failed or were intolerant of ESAs to luspatercept vs placebo; 38% of luspatercept patients vs 13% of placebo achieved transfusion independence for at least 8 weeks. COMMANDS (Platzbecker et al., Lancet 2023) compared luspatercept head-to-head against epoetin alfa as first-line therapy in 363 ESA-naive transfusion-dependent lower-risk MDS patients; the interim analysis showed 58.5% of luspatercept patients vs 31.2% of epoetin alfa patients achieved the composite primary endpoint of red-cell transfusion independence ≥12 weeks plus mean hemoglobin increase ≥1.5 g/dL, leading to first-line FDA approval in August 2023. The mechanistic rationale traces to Suragani and colleagues' 2014 Nature Medicine paper, which introduced the ACE-536 molecule and showed that it relieves a Smad2/3-mediated block on terminal erythroid maturation — distinct from the EPO-driven proliferative pathway, which is the conceptual foundation for why luspatercept works after ESA failure (and combined with ESAs in some real-world protocols). The honest caveats: hypertension is the most clinically significant adverse event (single-digit-percent serious events across trials); thromboembolic events occurred at higher rates in the luspatercept arm of BELIEVE in patients who had undergone splenectomy and were not on antiplatelet/anticoagulant therapy; and durability of response — and whether the transfusion-independence benefit translates to reduced iron overload and survival benefit — remains under continued post-marketing surveillance with longer-term BELIEVE follow-up (Cappellini et al., Lancet Haematol 2025) and the SOTERIA-equivalent registries for MDS.

History & Discovery

Luspatercept emerged from Acceleron Pharma's TGF-β superfamily receptor program in Cambridge, Massachusetts. The company's founding thesis — that soluble extracellular-domain Fc fusions of the activin-receptor family could be selectively retuned to capture disease-relevant ligand subsets — produced a series of related molecules: sotatercept (ACE-011, ActRIIA-Fc, originally targeted at bone biology and ultimately approved in pulmonary arterial hypertension); ACE-031 (wild-type ActRIIB-Fc, the muscle-focused asset halted in 2011 after BMP9/10-attributable vascular adverse events); and luspatercept (ACE-536, modified ActRIIB-Fc). The ACE-536 design choice — re-engineering the ActRIIB extracellular domain to bias selectivity toward GDF11 and activin B and away from activin A and the BMP9/10 vascular ligands — was deliberate, and was published in Suragani and colleagues' 2014 Nature Medicine paper as the foundational mechanistic description of the molecule. That paper demonstrated that ACE-536 corrects anemia in mouse models of MDS and β-thalassemia by relieving a Smad2/3-mediated maturation block at the late erythroblast stage — a mechanism conceptually distinct from EPO's effect on early erythroid progenitor proliferation. The companion Suragani 2014 Blood paper extended these findings specifically into the murine β-thalassemia model with detailed characterization of ineffective erythropoiesis, iron metabolism, and splenomegaly endpoints. Clinical development was conducted through Acceleron's commercial partnership with Celgene (later acquired by Bristol-Myers Squibb in 2019). The pivotal BELIEVE trial (NCT02604433) randomized 336 adults with transfusion-dependent β-thalassemia to luspatercept 1.0 mg/kg subcutaneously every three weeks vs placebo; the primary endpoint of ≥33% reduction in transfusion burden during weeks 13–24 was met (21.4% vs 4.5%), and the FDA approved Reblozyl on November 8, 2019 for transfusion-dependent anemia in adults with β-thalassemia. The MEDALIST trial (NCT02631070) randomized 229 patients with lower-risk MDS with ring sideroblasts who had failed or were intolerant of ESAs to luspatercept vs placebo; the primary endpoint of transfusion independence ≥8 weeks was met (38% vs 13%), and the FDA expanded the Reblozyl label on April 3, 2020 to include this population. The COMMANDS trial (NCT03682536) compared luspatercept head-to-head against epoetin alfa as first-line therapy in 363 ESA-naive transfusion-dependent lower-risk MDS patients; the interim analysis published by Platzbecker and colleagues in Lancet 2023 showed 58.5% vs 31.2% achieved the composite primary endpoint, and the FDA expanded the Reblozyl label on August 28, 2023 to include first-line use in this population. Merck's $11.5 billion acquisition of Acceleron in November 2021 was driven primarily by sotatercept (ACE-011) for pulmonary arterial hypertension, which Merck would subsequently bring to FDA approval as Winrevair in March 2024. Reblozyl (luspatercept) remained a Bristol-Myers Squibb asset following the BMS–Celgene transaction. The two related molecules — same conceptual scaffold (modified extracellular activin-receptor domain plus IgG1 Fc), different receptor (ActRIIB vs ActRIIA), different selectivity profile — illustrate how engineered ligand-affinity tuning produced clinically distinct therapeutics from a single molecular family. The BELIEVE long-term follow-up published by Cappellini and colleagues in Lancet Haematology 2025 extends the β-thalassemia efficacy and safety database to multi-year exposure, and the SOTERIA-style long-term registries for the MDS indications are ongoing.

How It Works

Red blood cell production happens in two stages. Erythropoietin (EPO) drives the first stage — making more early progenitor cells. The second stage — those progenitors maturing into actual circulating red cells — is controlled by a separate set of signals (GDF11 and related proteins) that act as a brake. In diseases like β-thalassemia and certain myelodysplastic syndromes, that brake is stuck in the 'on' position, so the bone marrow makes plenty of progenitors but few finished red cells. Luspatercept is an engineered protein that floats in the bloodstream and soaks up those brake signals, letting the late-stage maturation finish. That is why it works after EPO drugs have stopped helping — they push the gas; luspatercept releases the brake.

Luspatercept is a recombinant homodimeric fusion protein composed of a modified extracellular ligand-binding domain of human activin receptor type IIB (ActRIIB) fused to the hinge-CH2-CH3 region of human IgG1 Fc. The ActRIIB domain carries deliberate amino-acid substitutions (relative to wild-type ActRIIB used in the predecessor ACE-031) that retune its ligand-binding profile: luspatercept retains high affinity for GDF11 and activin B and lower affinity for activin A and the BMP9/BMP10 ligands implicated in vascular endothelial signaling. The Fc moiety supports homodimer formation (which contributes to high-avidity ligand binding) and provides FcRn-mediated extended plasma half-life that supports the every-three-weeks dosing schedule. In circulation, luspatercept binds and sequesters its target TGF-β superfamily ligands, preventing them from engaging membrane-bound ActRII receptors on erythroid precursors. The relevant downstream pathway is canonical Smad2/3 signaling: GDF11 and activin B engage type II/type I receptor complexes (ActRII paired with ALK4/ALK5), phosphorylating Smad2 and Smad3, which form complexes with Smad4 and translocate to the nucleus. In ineffective erythropoiesis, this Smad2/3 axis is hyperactive and produces a maturation block at the late basophilic-to-orthochromatic erythroblast transition. By sequestering the upstream ligands, luspatercept reduces Smad2/3 phosphorylation, releases the maturation block, and allows late erythroblasts to progress to reticulocytes and circulating red cells. This effect is complementary to, and mechanistically distinct from, EPO's action on early erythroid progenitor proliferation — which is the conceptual foundation for why luspatercept works in patients refractory to ESA therapy and why combining the two pathways produces synergistic responses in the COMMANDS trial. The Suragani et al. 2014 Nature Medicine paper (the foundational preclinical description of ACE-536 / luspatercept) demonstrated that the molecule corrects anemia in mouse models of β-thalassemia and MDS by selectively driving late-stage erythroid maturation, with minimal effect on the early progenitor pool. The companion Suragani et al. 2014 Blood paper extended these findings specifically into the murine β-thalassemia model, showing improvement in red-cell indices, reduction in ineffective erythropoiesis, and amelioration of secondary disease complications (iron overload, splenomegaly). Off-target consequences relative to broader-spectrum activin-receptor traps are smaller but not zero. The hypertension signal observed across all three pivotal trials is generally attributed to systemic effects of activin/BMP-pathway modulation on vascular tone and endothelial biology, though the precise molecular mediator has not been definitively established. The thromboembolic signal in splenectomized β-thalassemia patients reflects the additive effect of luspatercept-induced erythroid expansion in a population with a pre-existing hypercoagulable state, more than a direct procoagulant action of the molecule itself.

Evidence Snapshot

Research Gaps & Open Questions

What the current literature has not yet settled about Luspatercept:

• 01Whether luspatercept's transfusion-independence benefit translates to reduced cardiac iron load, fewer thromboembolic events long-term, and improved survival in β-thalassemia — the BELIEVE long-term follow-up addresses this in part but the question is open.
• 02Whether earlier first-line use in MDS (per COMMANDS) changes the natural history of the disease — including time to AML transformation — versus simply shifting the time of transfusion-burden management; a question that requires multi-year follow-up of the COMMANDS cohort.
• 03Optimal sequencing and combination with erythropoiesis-stimulating agents, hypomethylating agents, and other MDS therapies — the trials studied luspatercept as monotherapy; real-world combination practice is evolving without head-to-head data.
• 04Application to non-thalassemia, non-MDS forms of ineffective erythropoiesis — myelofibrosis, certain congenital dyserythropoietic anemias, sickle cell disease — where the same Smad2/3 maturation-block biology may be relevant but where pivotal trial data do not yet exist.
• 05Predictive biomarkers for response — baseline erythropoietin level has been examined in real-world cohorts (Chang et al., Leukemia Research Reports 2025) but no validated biomarker yet selects patients who will respond.
• 06Long-term vascular safety — the hypertension signal across all three pivotal trials warrants continued post-marketing surveillance for cardiovascular events with multi-year exposure.
• 07Pediatric efficacy and safety — the approved label is adult-only; pediatric β-thalassemia trials are needed to extend the indication.

Forms & Administration

Subcutaneous injection of reconstituted luspatercept-aamt powder, administered every three weeks by a healthcare professional. Adult β-thalassemia starting dose is 1.0 mg/kg, with titration up to a maximum of 1.25 mg/kg based on transfusion-burden response. Adult MDS starting dose is 1.0 mg/kg, with titration up to a maximum of 1.75 mg/kg based on hemoglobin response. Reblozyl is supplied as 25 mg or 75 mg single-dose vials of lyophilized powder, reconstituted with water for injection and administered into the upper arm, abdomen, or thigh. Luspatercept is a hospital/clinic-administered biologic — it is not self-administered, not available through compounding pharmacies, and is dispensed only via specialty distribution channels covered under the prescribing physician's healthcare system.

Common Questions

Who Luspatercept Is NOT For

Drug & Supplement Interactions

Formal drug-interaction studies on luspatercept are limited because the molecule is a biologic with no cytochrome-P450 metabolism and minimal direct pharmacokinetic interactions. Mechanism-based and clinically relevant interactions are as follows. In the COMMANDS trial, luspatercept was used as monotherapy in ESA-naive patients; in real-world MDS practice, sequential or combined use with erythropoiesis-stimulating agents (epoetin alfa, darbepoetin alfa) is sometimes employed when patients have partial responses to either agent alone — the mechanisms are complementary (EPO drives early progenitor proliferation, luspatercept relieves the late-maturation block) but combination data are observational rather than randomized. Antiplatelet and anticoagulant therapy is specifically relevant in splenectomized β-thalassemia patients, where the BELIEVE thromboembolic signal supports concomitant antithrombotic therapy. Adding luspatercept to a patient already on chronic anticoagulation does not require dose adjustment but warrants vigilance for hypertension-related bleeding considerations. Iron-chelation therapy (deferasirox, deferiprone, deferoxamine) is standard in transfusion-dependent β-thalassemia and was used concurrently in BELIEVE without dose modification of luspatercept; reduced transfusion burden on luspatercept can affect long-term iron-chelation requirements. Live vaccines have not been formally studied in luspatercept-treated patients; standard biologic-class precautions apply. Concurrent use with other agents that modulate TGF-β superfamily signaling (sotatercept, bimagrumab, apitegromab, follistatin preparations) has not been studied and would produce overlapping pathway blockade with unpredictable cumulative effects. There is no clinical scenario in which combining these agents would be appropriate outside a formal trial.

Safety Profile

Legal Status

Regulatory status changes over time. Verify current local rules with a qualified professional.

Myths & Misconceptions