Background
Phase 3, open-label, randomized controlled trial (interim analysis). 356 ESA-naive, RBC transfusion-dependent patients with lower-risk MDS (IPSS-R Very Low, Low, or Intermediate risk) requiring 2–6 RBC units per 8 weeks. Randomized 1:1, stratified by transfusion burden, endogenous EPO level, and ring sideroblast status. Conducted at 142 sites in 26 countries.
Interventions and follow up
Arm A: Luspatercept 1.0 mg/kg subcutaneously every 3 weeks, titrated up to 1.75 mg/kg (n=178)
Arm B: Epoetin alfa 450 IU/kg subcutaneously once weekly, titrated up to 1050 IU/kg (n=178)
Primary endpoint: RBC-TI ≥12 weeks with concurrent mean Hgb increase ≥1.5 g/dL during weeks 1–24 (ITT)
mFollow up: Interim analysis; 301 patients completed or discontinued the 24-week period
Arm B: Epoetin alfa 450 IU/kg subcutaneously once weekly, titrated up to 1050 IU/kg (n=178)
Primary endpoint: RBC-TI ≥12 weeks with concurrent mean Hgb increase ≥1.5 g/dL during weeks 1–24 (ITT)
mFollow up: Interim analysis; 301 patients completed or discontinued the 24-week period
Results
Primary endpoint (RBC-TI ≥12 wk + Hgb ≥1.5 g/dL): 59% (86/147) vs 31% (48/154), common risk difference 26.6% (95% CI 15.8–37.4), P<.0001
Median treatment duration: 42 weeks (luspatercept) vs 27 weeks (epoetin alfa)
Median treatment duration: 42 weeks (luspatercept) vs 27 weeks (epoetin alfa)
Adverse events
Grade 3–4 (luspatercept, ≥3%): hypertension, anemia, dyspnea, neutropenia, thrombocytopenia, pneumonia, COVID-19, syncope
Grade 3–4 (epoetin alfa, ≥3%): anemia, pneumonia, neutropenia, hypertension, iron overload, COVID-19 pneumonia
Other: Most common suspected luspatercept-related AEs were fatigue, asthenia, nausea, dyspnea, hypertension, headache (each ≤5%); one luspatercept-related death (AML, 44 days on treatment)
Grade 3–4 (epoetin alfa, ≥3%): anemia, pneumonia, neutropenia, hypertension, iron overload, COVID-19 pneumonia
Other: Most common suspected luspatercept-related AEs were fatigue, asthenia, nausea, dyspnea, hypertension, headache (each ≤5%); one luspatercept-related death (AML, 44 days on treatment)
Conclusions
In ESA-naive transfusion-dependent LR-MDS, luspatercept achieved nearly twice the rate of composite transfusion independence plus hemoglobin response compared with epoetin alfa, supporting its use as preferred first-line erythropoietic therapy in this setting.
Key Limitations
This is a prespecified interim analysis; long-term OS and AML transformation data remain immature. Open-label design introduces performance bias risk. The composite primary endpoint (TI + Hgb rise) is more demanding than simple TI — benefit in non-RS or non-SF3B1-mutated patients was numerically smaller. Epoetin alfa dosing was weight-based, which may not reflect EPO level-based clinical practice. Applicability to patients with EPO >500 U/L (excluded) is uncertain.
Clinical Context
COMMANDS expanded luspatercept's indication to ESA-naive LR-MDS (not restricted to RS-positive), gaining FDA approval in 2023. This positions luspatercept as a first-line alternative to ESAs across a broader LR-MDS population, particularly where EPO levels are ≤500 U/L. Long-term follow-up is awaited to assess OS and AML risk differences between arms.