Background:

In the phase 3 COMMANDS trial (NCT03682536) evaluating pts with ESA-naive transfusion-dependent (TD) LR-MDS, luspatercept was superior to epoetin alfa (EA) in achieving red blood cell-transfusion independence (RBC-TI) ≥12 weeks (wks) and demonstrated a more durable clinical benefit. Higher Hb, lower transfusion burden (TB), and lower serum erythropoietin (sEPO) levels suggest a less advanced disease state and may reflect an earlier stage of LR-MDS. Here, we report the clinical benefit and outcomes of treatment with luspatercept and EA in pts with earlier vs later stage disease characteristics.

Methods:

Eligible pts (≥18 years; ESA-naive; RBC-TD; IPSS-R very low/low/intermediate-risk MDS) were randomized 1:1 to luspatercept (1.0 mg/kg; titrated up to 1.75 mg/kg) SC Q3W or EA (450 IU/kg; titrated up to 1050 IU/kg) SC QW for ≥24 wks. RBC-TI ≥12 wks (Wk 1-end of treatment [EOT]), duration of RBC-TI ≥12 wks, and erythroid hematologic improvement (HI-E; Wk 1-EOT) were analyzed. To evaluate how disease severity affects outcomes, efficacy was categorized by Hb (≥8, <8 g/dL), TB (=2, ≥4 U), and sEPO (≤100, >200 U/L) at baseline (BL). The data cutoff was Feb 7, 2025.

Results:

Median follow-up was 30.6 months in the luspatercept arm (n=182) and 28.8 months in the EA arm (n=181). At BL, 39.6% of pts in the luspatercept arm and 40.3% of pts in the EA arm had Hb ≥8 g/dL; 44.5% and 45.3% had TB =2 U; and 56.6% and 58.6% had sEPO ≤100 U/L.

More pts with Hb ≥8 g/dL achieved RBC-TI ≥12 wks than Hb <8 g/dL with luspatercept (87.5% vs 69.1%) and a similar trend was seen with EA (74.0% vs 43.5%). Pts with Hb ≥8 g/dL had a longer median duration (95% CI) of RBC-TI ≥12 wks than Hb <8 g/dL with luspatercept (150.0 wks [72.0-not estimable (NE)] vs 108.3 wks [53.7-132.6]) but not EA (75.6 wks [41.9-101.1] vs 86.7 wks [37.3-186.1]); the HR (95% CI) for luspatercept vs EA was 0.60 (0.36-1.00 [Hb ≥8 g/dL]) and 1.04 (0.64-1.71 [Hb <8 g/dL]). Similarly, more pts with Hb ≥8 g/dL achieved a HI-E response than Hb <8 g/dL with luspatercept (83.3% vs 78.2%) but not EA (57.5% vs 57.4%).

More pts with TB =2 U achieved RBC-TI ≥12 wks than TB ≥4 U with luspatercept (86.4% vs 60.9%) and a similar trend was seen with EA (73.2% vs 34.3%). Pts with TB =2 U also had a longer median duration (95% CI) of RBC-TI ≥12 wks than TB ≥4 U with luspatercept (120.9 wks [81.0-NE] vs 110.0 wks [43.4-184.4]) and a similar trend was seen with EA (95.1 wks [61.9-186.1] vs 62.9 wks [26.9-NE]); the HR (95% CI) for luspatercept vs EA was 0.80 (0.49-1.30 [TB =2 U]) and 0.90 (0.46-1.76 [TB ≥4 U]). Likewise, more pts with TB =2 U achieved a HI-E response than TB ≥4 U with luspatercept (87.7% vs 73.4%) and a similar trend was seen with EA (64.6% vs 57.1%).

More pts with sEPO ≤100 U/L achieved RBC-TI ≥12 wks than sEPO >200 U/L with luspatercept (87.4% vs 54.1%) and a similar trend was seen with EA (71.7% vs 18.9%). Pts with sEPO ≤100 U/L also had a longer median duration (95% CI) of RBC-TI ≥ 12 wks than pts sEPO >200 U/L with luspatercept (143.3 wks [120.1-235.9] vs 48.3 wks [26.3-132.6]) and a similar trend was seen with EA (95.1 wks [69.7-186.1] vs 24.6 wks [14.9-NE]); the HR for luspatercept vs EA was 0.75 (0.48-1.17 [sEPO ≤100 U/L]) and 0.85 (0.30-2.39 [sEPO >200 U/L]). Likewise, more pts with sEPO ≤100 U/L achieved a HI-E response than sEPO >200 U/L with luspatercept (90.3% vs 62.2%) and a similar trend was seen with EA (73.6% vs 29.7%).

Regardless of the disease severity as indicated by these markers, luspatercept consistently performed better than EA. When comparing luspatercept vs EA (OR [95% CI]), luspatercept had >2-fold odds of achieving RBC-TI ≥12 wks (Hb ≥8 g/dL: 2.46 [1.03-5.89]; Hb <8 g/dL: 2.90 [1.67-5.05]; TB =2 U: 2.33 [1.05-5.20]; TB ≥4 U: 2.99 [1.48-6.05]; sEPO ≤100 U/L: 2.73 [1.33-5.61]; sEPO >200 U/L: 5.04 [1.77-14.36]) or HI-E (Hb ≥8 g/dL: 3.69 [1.70-8.00]; Hb <8 g/dL: 2.66 [1.47-4.80]; TB =2 U: 3.88 [1.74-8.66]; TB ≥4 U: 2.07 [1.00-4.30]; sEPO ≤100 U/L: 3.34 [1.53-7.30]; sEPO >200 U/L: 3.88 [1.47-10.23]).

Conclusions:

Pts with less severe disease characteristics, reflective of an earlier disease stage, derive greater clinical benefit from treatment than those with more advanced characteristics. Pts on luspatercept vs EA demonstrated higher response rates and longer duration of response regardless of disease stage, further supporting its role as a preferred first-line therapy in LR-MDS.

This content is only available as a PDF.
2025
Sign in via your Institution