Recombinant human erythropoietin (rHuEPO) stimulates erythropoietic bone marrow cells and increases erythrocyte production. This prospective study was designed to evaluate the effects of rHuEPO on regeneration of erythropoiesis after allogeneic or autologous bone marrow transplantation (BMT). Seventeen centers participated in this randomized, double-blind, placebo-controlled multicenter trial. The randomization was performed centrally for each center and stratified according to allogeneic or autologous BMT and major ABO-blood group incompatibility. One hundred and six patients received rHuEPO after allogeneic BMT and 109 patients received placebo. After autologous BMT, 57 patients were treated with rHuEPO and 57 with placebo. Patients received either 150 IU/kg/day C127 mouse-cell-derived rHuEPO or placebo as continuous intravenous infusion. Therapy started after bone marrow infusion and lasted until independence from erythrocyte transfusions for 7 consecutive days with stable hemoglobin levels > or = 9 g/100 mL or until day 41. After allogeneic BMT, the reticulocyte counts were significantly higher with rHuEPO from day 21 to day 42 after BMT. The median time (95% confidence intervals) to erythrocyte transfusion independence was 19 days (range, 16.3 to 21.6) with rHuEPO and 27 days (range, 22.3 to > 42) with placebo (P < .003). The mean (+/- SD) numbers of erythrocyte transfusions until day 20 after BMT were 6.6 +/- 4.8 with rHuEPO and 6.0 +/- 3.8 with placebo. However, from day 21 to day 41, the rHuEPO-treated patients received 1.4 +/- 2.5 (median, 0) transfusions and the control group received 2.7 +/- 4.0 (median, 2) transfusions (P = .004). In the follow-up period from day 42 up to day 100, 2.4 +/- 5.6 transfusions were required with rHuEPO and 4.5 +/- 9.6 were required with placebo (P = .075). A multivariate analysis (ANOVA) showed that acute graft-versus-host disease (GVHD), major ABO-blood group incompatibility, age greater than 35 years, and hemorrhage significantly increased the number of transfusions. However, after day 20, rHuEPO significantly reduced the number of erythrocyte transfusions in these patient groups, as well as reducing incompatibility in the major ABO-blood group. For the whole study period, rHuEPO reduced the transfusion requirements in GVHD III and IV from 18.4 +/- 8.6 to 8.5 +/- 6.8 U (P = .05). After autologous BMT, there was no difference in the time to independence from erythrocyte transfusions and in the regeneration of reticulocytes. Marrow purging strongly increased the requirement for transfusions as well as the time to transfusion independence.

Infections during granulocytopenia are major complications of autologous bone marrow transplantation (ABMT). Since recombinant human granulocyte-macrophage colony-stimulating factor (rhuGM-CSF) has proved to accelerate bone marrow recovery after cytostatic chemotherapy, we studied its effects on hematopoietic regeneration and on infectious complications after total body irradiation (TBI) and high-dose chemotherapy followed by ABMT. Eighty-one patients with acute lymphoblastic leukemia (ALL) in complete remission (CR) or with non- Hodgkin's lymphoma (NHL) in CR or partial remission were randomized in a double-blind, placebo-controlled trial. They received either rhuGM- CSF 250 micrograms/m2 (Escherichia coli-derived) daily by continuous infusion after ABMT, or placebo. Treatment was continued until the neutrophil counts reached greater than 500/microL for 1 week. The maximum treatment duration was 30 days. Thirty-nine patients in the rhuGM-CSF group and 40 patients in the placebo group were evaluable. The median time needed to reach a neutrophil count of 500/microL was 15 days with rhuGM-CSF and 28 days with placebo (P = .0001). Bacterial infections occurred in 14 (35.9%) of the patients with rhuGM-CSF and in 25 (62.5%) of the patients given the placebo (P = .024). Nine of the 14 bacterial infections in the rhuGM-CSF group and 20 of the 25 infections in the placebo group were diagnosed within the first 10 days after ABMT. Capillary leakage and a reversible fluid retention were seen in five of the rhuGM-CSF-treated patients. Patients treated with rhuGM-CSF had lower serum protein and albumin levels than patients in the placebo group. There was no statistically relevant difference in overall survival between the two groups (P = .47). Relapse occurred in 14 (34%) patients with rhuGM-CSF and in 18 (45%) patients with placebo. We conclude that continuous infusion of rhuGM-CSF after ABMT accelerates the regeneration of granulocytes and reduces the number of bacterial infections.