Luspatercept (TGF-β superfamily ligand trap) extends transfusion-independence vs epoetin alfa in ESA‑naive LR‑MDS; OS still trending better

Luspatercept (LUSPA), a recombinant fusion protein that binds select TGF-β superfamily ligands to promote late-stage erythroid maturation, was compared with epoetin alfa (EA; an ESA) in the Phase 3, open-label COMMANDS trial (NCT03682536) in ESA-naive, transfusion-dependent lower-risk MDS (n=182 LUSPA; n=181 EA). With data cut Oct 6, 2025 (median follow-up 35.9 vs 30.8 months), median OS was NR with LUSPA vs 46.0 months with EA (HR 0.78, 95% CI 0.56–1.09) while durability clearly favored LUSPA: RBC-TI ≥12 weeks occurred in 76.4% vs 55.8%, median cumulative RBC-TI duration was 184.4 vs 95.1 weeks (HR 0.54, 95% CI 0.36–0.80), and RBC-TI ≥2.5 years was 25.3% vs 10.5%—well above what’s typically expected with frontline ESA strategies where sustained multi-year TI is uncommon, especially in higher transfusion-burden patients. Safety was consistent with prior reports (no new signals); deaths were 36.3% vs 43.0%, progression to higher-risk MDS was 3.3% vs 7.3%, and AML progression was similar (4.9% vs 5.5%).