Purpose: The purpose of this report was to review the Seattle experience in bone marrow transplantation (BMT) for acute myeloid leukemia (AML) during untreated first relapse.
Patients and methods: Through 1990, 126 patients were transplanted during untreated first relapse of AML. Several preparative regimens were used, two of which involved more than 20 patients. Regimen 1 (29 patients) consisted of cyclophosphamide (CY) 120 mg/kg and 15.75 Gy of fractionated total-body irradiation (TBI) with methotrexate (MTX) given intermittently during a 102-day period to prevent graft-versus-host disease (GVHD). Regimen 2 (22 patients) consisted of the same CY and TBI treatment and a combination of MTX and cyclosporine (CSP) for GVHD prophylaxis. The remaining 75 patients were treated with 17 other transplant regimens. Outcome was compared for patients who were treated with regimen 1, regimen 2, and any other regimen.
Results: The 5-year probabilities of relapse-free survival (RFS), relapse, and nonrelapse mortality for 126 patients were .23, .57, and .44, respectively. With regimen 1, relapse (.26) was significantly less than for regimen 2 (.70; P = .004) or any other regimen (.76; P = .004). Regimen 1 patients developed more acute GVHD (.67) than regimen 2 patients (.26; P = .02) or patients on other regimens (.41; P = .02), and had increased nonrelapse mortality. Nevertheless, regimen 1 patients had a significantly higher 3-year RFS (.38) than those treated with regimen 2 (.18; P = .04) or any other regimen (.20; P = .05).
Conclusions: For patients who received 120 mg/kg CY and 15.75 Gy TBI, relapse incidence was less and survival was better after GVHD prophylaxis with MTX alone than after a combination of MTX and CSP, despite a significantly higher incidence of acute GVHD. The results of treatment with regimen 1 justify future studies of the optimal timing of allogeneic BMT in the treatment of patients with AML.