Changes in Red Blood Cell Transfusion Burden with Luspatercept Versus Epoetin Alfa in Patients with Lower-Risk Myelodysplastic Syndromes in the Phase 3, Open-Label, Randomized, Controlled COMMANDS Trial

Introduction: Many patients (pts) with lower-risk myelodysplastic syndromes (LR-MDS) require regular red blood cell transfusions (RBCTs), which carry a treatment burden on health and quality of life. To manage chronic anemia, pts may also receive erythropoiesis-stimulating agents (ESAs), though the efficacy of ESAs is often limited and transient. Luspatercept is a first-in-class erythroid maturation agent approved in the United States to treat anemia in ESA-naive adult pts with LR-MDS who may require regular RBCTs. The first interim analysis of the phase 3 COMMANDS trial (data cutoff, August 31, 2022) showed significantly improved RBCT independence (RBC-TI) rates, longer response durability, and longer durations of RBC-TI ≥12 weeks (wk) for luspatercept versus epoetin alfa. Here, we report changes in RBCT burden over time from the COMMANDS trial (data cutoff, September 22, 2023).

Methods: COMMANDS is an ongoing, phase 3, open-label, randomized, controlled trial (NCT03682536) involving adult (≥18 years of age) pts with LR-MDS who are ESA-naive, RBCT-dependent, and have endogenous serum erythropoietin <500 U/L. Baseline RBCT burden was defined as the number of RBCT units/8 wk, categorized as high (HTB; ≥4 units/8 wk) or low (LTB; <4 units/8 wk). Changes from baseline in RBCT burden, measured by cumulative mean number of RBC units transfused and by RBCT visits in the overall population, stratified by baseline RBCT burden, were analyzed. Expected cumulative number of post-baseline RBC units transfused and RBCT-related visits were analyzed using the Nelson-Aalen nonparametric estimator with robust variance estimate (with 95% confidence intervals [CIs]).

Results: 363 pts were randomized to either luspatercept (n=182) or epoetin alfa (n=181). Median age was 74 years, 55% were male, and median baseline Hb level was 7.8 g/dL. Median baseline RBCT burden was 3.0 units/8 wk in luspatercept and epoetin alfa arms. Most pts (63%) had LTB at baseline (luspatercept, 65%; epoetin alfa, 61%).

Mean cumulative RBC units transfused (95% CI) were 3.8 (3.0–4.8) for pts treated with luspatercept and 5.7 (4.7–6.8) for pts treated with epoetin alfa by wk 24. The gradual increase in mean cumulative RBC units transfused by wk 96 was higher in the epoetin-alfa arm of treatment, resulting in 8.7 (7.0–10.8) in pts treated with luspatercept versus 13.6 (11.2–16.5) in those treated with epoetin alfa. Mean cumulative RBCT visits for pts receiving luspatercept and epoetin alfa, respectively, were 2.4 (1.9–2.9) and 3.6 (3.0–4.4) by wk 24, and 5.7 (4.5–7.1) and 8.6 (7.1–10.4) by wk 96.

When analyzed by baseline RBCT burden (ie, HTB and LTB), luspatercept treatment resulted in a lower cumulative mean of RBC units transfused over time. Among baseline LTB pts, mean cumulative RBC units transfused by wk 24 were 2.3 (1.6–3.3) in the luspatercept arm (n=118) and 3.2 (2.4–4.2) in the epoetin-alfa arm (n=111), increasing to 6.0 (4.5–8.0) and 8.6 (6.5–11.2) by wk 96, respectively. A larger difference in mean cumulative RBC units transfused was found among baseline HTB pts, with 6.6 (5.1–8.5) in pts treated with luspatercept (n=64) and 9.7 (8.0–11.8) in those treated with epoetin alfa (n=70) by wk 24, respectively, and 14.3 (10.6–19.4) and 23.4 (18.1–30.3) by wk 96.

Among baseline LTB pts, mean cumulative RBCT visits for luspatercept versus epoetin alfa arms (n=118 and 111) were 1.3 (1.0–1.9) and 2.2 (1.6–3.0) by wk 24, and 3.9 (2.9–5.2) and 5.6 (4.3–7.3) by wk 96, respectively. For baseline HTB pts, a greater difference was found in mean cumulative RBCT visits for luspatercept versus epoetin alfa (n=64 and 70), respectively, with 4.2 (3.3–5.4) and 5.9 (4.8–7.3) by wk 24, and 9.2 (6.6–12.9) and 14.4 (11.0–18.7) by wk 96.

Conclusions: The findings from the COMMANDS trial demonstrated that luspatercept treatment led to a decrease in RBCT burden compared with epoetin alfa in ESA-naive pts with LR-MDS, evident by the number of RBC units transfused and RBCT visits over a 1.5-year period. When analyzed by RBCT burden category, transfusion burden was also substantially lower for luspatercept compared with epoetin alfa.