Health-Related Quality of Life Outcomes for Patients with Transfusion-Dependent Beta-Thalassemia Treated with Luspatercept in the BELIEVE Trial

Introduction: Patients (pts) with transfusion-dependent (TD) β-thalassemia may require long-term red blood cell transfusions (RBCT) that can lead to iron overload and associated complications, which impact negatively on health-related quality of life (HRQoL). Administration of RBCT provides transient relief from anemia-related symptoms associated with β-thalassemia; reduction of RBCT may increase anemia-related symptoms and thereby worsen HRQoL. The phase 3 BELIEVE study (NCT02604433) showed that the first-in-class erythroid maturation agent luspatercept provided clinically meaningful reduction in RBCT burden, but the impact of luspatercept on HRQoL is not well understood. This analysis assessed the effect of luspatercept plus best supportive care (BSC, including RBCT, iron chelation therapy) vs placebo (PBO) plus BSC on HRQoL in pts with TD β-thalassemia.

Methods: Pts received luspatercept (starting dose 1.0 mg/kg with titration up to 1.25 mg/kg every 3 weeks) or PBO, subcutaneously for ≥ 48 weeks plus BSC. HRQoL was assessed using the generic 36-item Short Form Health Survey (SF-36) and the thalassemia-specific Transfusion-dependent Quality of Life questionnaire (TranQol), at screening (≤ 4 weeks prior to first study dose) and every 12 weeks up to 48 weeks of treatment. The HRQoL evaluable population included all pts who completed the HRQoL assessment at screening and ≥ 1 post-screening assessment visit. The TranQol and SF-36 were considered complete if ≥ 75% and ≥ 50% of items, respectively, were answered at a given time point. The primary analysis assessed changes from baseline between groups up to Week 48. The primary domains of interest were: TranQol total score and Physical Health (PH); and the SF-36 Physical Component Summary (PCS), Physical Functioning (PF), and General Health (GH). Other domains were considered exploratory domains. Changes from baseline were compared using ANCOVA models adjusting for baseline domain scores and geographic region. In exploratory analyses, the proportion of pts achieving a clinically meaningful improvement in domain scores were compared between pts on luspatercept achieving a clinical response (≥ 50% reduction in RBCT burden over 12 weeks; ≥ 33% reduction in RBCT burden over 12 weeks; transfusion independence [TI] any 8 weeks; or TI any 12 weeks), and PBO.

Results: 336 pts were randomized to treatment; 224 to luspatercept and 112 to PBO. The HRQoL evaluable population was 212 (94.6%) in the luspatercept arm and 104 (92.9%) in the PBO arm. HRQoL questionnaire compliance rates among pts still on treatment were > 87.5% for both questionnaires at Week 48. Baseline HRQoL scores were similar to the US general population for most SF-36 domains, although GH, Role-Emotional, and Role-Physical domain scores were impaired in the BELIEVE population. Mean scores on all primary and exploratory domains were stable over time in both treatment groups and did not differ between treatment groups at Week 24 and 48. When considering responders to luspatercept, pts receiving luspatercept and achieving a ≥ 50% reduction in RBCT burden over 12 weeks were significantly more likely than pts receiving PBO to have a clinically meaningful improvement in PCS (31.1% vs 16.5%; P = 0.024), and PF (30.0% vs 13.2%; P = 0.007) at Week 48 (Table). Statistically significant differences between luspatercept and PBO were also seen among pts achieving TI for any 8 or any 12 weeks for some SF-36 domains, but no statistical difference was seen in pts achieving a ≥ 33% reduction in RBCT burden for either SF-36 or TranQol domains although the proportion of pts with improved scores was higher with luspatercept, especially at Week 48.

Conclusions: Overall, the addition of luspatercept to BSC reduced transfusion burden while sustaining TranQol and SF-36 HRQoL scores over time through Week 48 compared with those receiving PBO. Pts with TD β-thalassemia responding to luspatercept were more likely to achieve clinically meaningful improvements in HRQoL compared with PBO.