Liberal Versus Restrictive Red Blood Cell Transfusion Thresholds in Hematopoietic Cell Transplantation: A Randomized, Open Label, Phase III, Noninferiority Trial

Abstract

PURPOSE Evidence regarding red blood cell (RBC) transfusion practices and their impact on hematopoietic cell transplantation (HCT) outcomes are poorly understood. PATIENTS AND METHODS We performed a noninferiority randomized controlled trial in four different centers that evaluated patients with hematologic malignancies requiring HCT who were randomly assigned to either a restrictive (hemoglobin [Hb] threshold < 70 g/L) or liberal (Hb threshold < 90 g/L) RBC transfusion strategy between day 0 and day 100. The noninferiority margin corresponds to a 12% absolute difference between groups in Functional Assessment of Cancer Therapy-Bone Marrow Transplant (FACT-BMT) score relative to baseline. The primary outcome was health-related quality of life (HRQOL) measured by FACT-BMT score at day 100. Additional end points were collected: HRQOL by FACT-BMT score at baseline and at days 7, 14, 28, 60, and 100; transplantation-related mortality; length of hospital stay; intensive care unit admissions; acute graft-versus-host disease; Bearman toxicity score; sinusoidal obstruction syndrome; serious infections; WHO Bleeding Scale; transfusion requirements; and reactions to therapy. RESULTS A total of 300 patients were randomly assigned to either restrictive-strategy or liberal-strategy treatment groups between 2011 and 2016 at four Canadian adult HCT centers. After HCT, mean pre-transfusion Hb levels were 70.9 g/L in the restrictive-strategy group and 84.6 g/L in the liberal-strategy group ( P < .0001). The number of RBC units transfused was lower in the restrictive-strategy group than in the liberal-strategy group (mean, 2.73 units [standard deviation, 4.81 units] v 5.02 units [standard deviation, 6.13 units]; P = .0004). After adjusting for transfusion type and baseline FACT-BMT score, the restrictive-strategy group had a higher FACT-BMT score at day 100 (difference of 1.6 points; 95% CI, −2.5 to 5.6 points), which was noninferior compared with that of the liberal-strategy group. There were no significant differences in clinical outcomes between the transfusion strategies. CONCLUSION In patients undergoing HCT, the use of a restrictive RBC transfusion strategy threshold of 70 g/L was as effective as a threshold of 90 g/L and resulted in similar HRQOL and HCT outcomes with fewer transfusions.