Luspatercept Misses Primary RBC-TI End Point But Shows Clinically Meaningful Improvement for Anemia in Myelofibrosis

Although luspatercept-aamt (Reblozyl) did not meet the prespecified threshold for statistical significance on the red blood cell transfusion-independence (RBC-TI) primary end point in the phase 3 INDEPENDENCE trial (NCT04717414), the agent did display clinically meaningful improvements in RBC-TI, transfusion burden (TB), and hemoglobin (Hb) levels vs placebo in the treatment of patients with myelofibrosis-associated anemia who required RBC transfusions while receiving concomitant JAK inhibitor therapy.¹

Data presented at the 2026 EHA Congress showed that RBC-TI for at least 12 consecutive weeks within the first 24 weeks—the study’s primary end point—was achieved by 23.1% of patients treated with luspatercept (n = 208) vs 13.3% with placebo (n = 105; risk difference, 9.74%; 95% CI, −0.27% to 18.17%; P = .0674). The prespecified significance threshold at this primary analysis was P < .044, reflecting alpha spending from an earlier interim analysis.

Outcomes for the primary end point favored luspatercept in most prespecified patient subgroups, other than those with a baseline serum EPO level of at least 500 U/L.

Another subgroup analysis looked at outcomes by region. In Europe, the Middle East, Africa, and Latin America, RBC-TI for at least 12 consecutive weeks within the first 24 weeks was achieved by 22.0% of patients treated with luspatercept (n = 127) vs 4.7% of patients treated with placebo (n = 64; risk difference, 17.36%; 95% CI, 5.52%-26.49%). These respective rates were 28.6% for luspatercept (n = 14) and 0% for placebo (n = 3) in the North American population (risk difference, 28.57%; 95% CI, –39.27% to 59.26%).

However, in the Asia-Pacific population, RBC-TI for at least 12 consecutive weeks within the first 24 weeks was achieved by 23.9% of patients treated with luspatercept (n = 67) vs 28.9% of patients treated with placebo (n = 38; risk difference, –5.07%; 95% CI, –24.03% to 12.14%).

“The high RBC-TI for 12 weeks with placebo in the Asia-Pacific regions, together with differences shown in patient characteristics and variations in routine clinical practice [for China vs the rest of the world], contributed to the narrow miss of statistical significance of the primary end point in the intention-to-treat population,” lead study author Francesco Passamonti, MD, said in a presentation of the data. “The primary results from INDEPENDENCE establish luspaterept as a clinically meaningful approach for myelofibrosis-associated anemia, a setting where effective therapies remain a critical unmet need.”

Although not currently approved by global health authorities for the treatment of anemia in patients with myelofibrosis, luspatercept is listed under other recommendations for the treatment of myelofibrosis-related anemia in the National Comprehensive Cancer Network Guidelines.²

How was the INDEPENDENCE trial designed?

INDEPENDENCE was a phase 3, double-blind, randomized trial that enrolled patients with myeloproliferative neoplasm–associated myelofibrosis, including primary myelofibrosis, post–polycythemia vera myelofibrosis, and post–essential thrombocythemia myelofibrosis, who were receiving 4 to 12 RBC units per 12 weeks and had been on a stable JAK inhibitor dose for at least 16 weeks.¹ Patients also required peripheral blood blast cells of 5% or fewer.

Patients were randomly assigned 2:1 to receive subcutaneous luspatercept plus best supportive care (BSC) or placebo plus BSC on day 1 of every 21-day treatment cycle. Concomitant JAK inhibitor therapy was maintained throughout. A 24-week blinded core treatment period was followed by an extension period of 25 or more weeks.

Along with the primary end point of RBC-TI for at least 12 consecutive weeks within the first 24 weeks, notable secondary end points comprised RBC-TI for at least 16 consecutive weeks within the first 24 weeks; the rate of patients with a reduction of TB by at least 50% and at least 4 RBC units from baseline over any consecutive 12-week period; the rate of patients with RBC-TI for at least 12 consecutive weeks and a mean Hb increase of at least 1 g/dL from baseline.

Baseline characteristics were generally balanced. Median age was 72.0 years in the luspatercept arm and 73.0 years in the placebo arm. Primary myelofibrosis was the most common MPN-associated diagnosis (62.5% and 70.5%, respectively). Approximately 91.6% of luspatercept-treated patients had DIPSS Intermediate-2 or High risk scores.

What other regional differences drove the primary end point outcome?

Mainland China patients differed substantially from the rest of the world (ROW) population at baseline, with data showing a median Hb level of 5.8 g/dL for China (n = 45) vs 7.8 g/dL for ROW (n = 268). Additionally, 80.0% of patients from China had an ECOG performance status score of 1 vs 50.4% for ROW, and 84.4% had serum EPO levels of at least 500 U/L vs 48.1% in ROW. e.

A post-hoc sensitivity analysis re-assessed the primary end point at Day 169 using a blinded investigator response assessment. A placebo-arm patient previously classified as an RBC-TI for at least 12 weeks responder was reclassified as a non-responder after investigators determined the observed transfusion independence reflected a regional blood shortage rather than a treatment effect. Under this analysis, RBC-TI for at least 12 in the first 24 weeks was achieved by 23.1% of patients treated with luspatercept vs 12.4% with placebo (nominal P = .0398).

Other efficacy end point data showed RBC-TI for at least 16 consecutive weeks within the first 24 weeks was achieved by 19.2% of patients treated with luspatercept vs 11.4% with placebo (risk difference, 6.5%; 95% CI, −1.3% to 14.3%). Reductions in transfusion burden of at least 50% with at least 4 RBC units from baseline over any consecutive 12-week period were achieved by 40.9% vs 22.9% of patients, respectively (risk difference, 17.3%; 95% CI, 6.9%-27.7%). Combined RBC-TI of at least 12 weeks with a mean Hb increase of at least 1 g/dL was achieved by 15.9% vs 5.7% of patients, respectively (risk difference, 9.3%; 95% CI, 2.7%-15.8%).

What did the safety analysis reveal?

Treatment-emergent adverse effects (TEAEs) of any grade occurred in 96.1% of luspatercept-treated patients (n = 207) and 92.4% of placebo-treated patients (n = 105). Treatment-related grade 3 or higher TEAEs were reported in 9.2% and 1.9% of patients, respectively. The most commonly reported TEAEs (>15% in either arm) included thrombocytopenia (27.1% vs 17.1%), hypertension (19.3% vs 13.3%), peripheral edema (18.4% vs 5.7%), and diarrhea (15.0% vs 15.2%).

Among AEs of interest (EOIs), hematologic acute myeloid leukemia (AML)–like events were observed in 6 (2.9%) luspatercept-treated patients vs none with placebo. Investigators noted that AML progression occurred predominantly in patients with a long-standing myelofibrosis history and high-risk baseline mutational profiles; 5 of the 6 patients harbored an ASXL1 mutation at baseline, and 1 had missing baseline mutation data. The incidence was described as consistent with the natural rate of AML progression in patients with advanced myelofibrosis receiving JAK inhibitor therapy. No clear treatment-related trend in causes of death was observed.

Investigators concluded the safety profile of luspatercept was manageable and consistent with prior studies and the expected profile in patients with advanced myelofibrosis.

References

1. Passamonti F, Kiladjian JJ, Mascarenhas J, et al. Efficacy and safety of luspatercept in patients with myelofibrosis on Janus kinase inhibitors who require red blood cell transfusions: primary analysis of the phase 3 INDEPENDENCE trial. Presented at: 2026 EHA Congress; June 11-14, 2026; Stockholm, Sweden. Abstract S215.
2. NCCN Clinical Practice Guidelines in Oncology Myeloproliferative Neoplasms Version 1.2026. NCCN. January 22, 2026. Accessed June 13, 2026. https://www.nccn.org/professionals/physician_gls/pdf/mpn.pdf