A Multicenter, Randomized, Controlled Clinical Pilot Trial of the Feasibility of an Intensive RBC Transfusion Policy for Patients with Acute Leukemia Treated with Induction Chemotherapy or Stem Cell Transplant.

Abstract Introduction: Patients with AML or ALL undergoing induction chemotherapy or SCT are at risk for hemorrhage. Hemoglobin (Hb) concentration may affect bleeding risk in thrombocytopenic patients. We performed a single-blinded pilot RCT to evaluate the feasibility of conducting a larger RCT to determine the effect of Hb on bleeding risk in thrombocytopenic patients. Objectives of the pilot RCT were: (1) to estimate the rate and variability of bleeding during the period of thrombocytopenia; (2) to confirm feasibility of the larger trial by demonstrating that: (a) it is possible to maintain patients’ Hb within the targeted range; (b) patients assigned to the intensive threshold do not have increased donor exposure; (c) it will be feasible to enroll the required number of patients; and (d) bleeding can be systematically documented. Methods: Adults with AML or ALL undergoing induction chemotherapy or HLA-matched myeloablative allogeneic SCT were eligible. Consecutive eligible patients were randomly assigned to one of 2 treatment groups: standard RBC transfusion strategy (control group) and intensive RBC transfusion strategy (experimental group). Groups were stratified by treatment centre and diagnosis. Patients in the control group were transfused 2 RBC units when their Hb was <8g/dL to maintain their Hb between 8–10g/dL. Patients in the experimental group were transfused 2 RBC units when their Hb was <12g/dL to maintain their Hb above this threshold. Bleeding was documented daily by a blinded study assessor using a standardized form and protocol. Primary outcomes included: bleeding estimates; proportion of days with a Hb within targeted range; blood product utilization (RBC and platelets [plt]); and hospital length of stay (LOS). Secondary outcomes included the ability to document bleeding using the standardized protocol. Two-sided t-tests were used to compare blood product utilization and LOS between groups. Results: 60 patients were enrolled: 29 (control group) and 31 (experimental group). The mean age was 47.9 (range 18–77) years. 30 patients had newly diagnosed AML, 9 had relapsed AML, 5 had ALL, and 16 were undergoing SCT. Patients in the control group had Hb of 8–10g/dL on 56.4% of days. Patients in the experimental group had Hb>12g/dL 56.5% of days. The control group received fewer RBC transfusions (units/day) than the experimental group (0.167 vs 0.320, p<0.0001). The mean number of plts transfused (units/day) was not different for control and experimental groups (0.851 vs 1.063, p=0.383). The mean number of donor exposures (plt and RBC transfusions) was not different between control and experimental groups (26.9 vs 27.5, p=0.923). There was a nonsignificant trend towards a shorter LOS in the experimental group (27.9 vs 23.7, p=0.112). Bleeding data are undergoing blinded adjudication. Conclusions: It is feasible to enroll the required number of patients for a large RCT to investigate the effect of Hb on bleeding risk. Patients assigned to the intensive strategy had increased RBC utilization but no increase in total donor exposures. Funded by a grant from Canadian Blood Services.