OBJECTIVES: I. Compare event free survival after total body irradiation (TBI) plus busulfan versus TBI plus cyclophosphamide followed by allogeneic peripheral blood stem cell transplantation in patients with advanced myelodysplastic syndrome (MDS) or MDS related acute myeloid leukemia. II. Determine the distribution of pharmacokinetic parameters for busulfan in those patients randomized to the busulfan treatment arm. III. Investigate the prognostic significance for event free survival of prior history of red cell transfusions, cytogenetic pattern, and of functional drug resistance at diagnosis in these patients. IV. Estimate the frequencies of cytogenetic and genetic changes during disease progression in these patients.

OUTLINE: This a randomized, multicenter study. Patients are stratified according to age (40 and under vs 41-55) and diagnosis and International Prognostic Scoring System (IPSS) risk group (myelodysplastic syndrome (MDS)/IPSS - intermediate 1 vs MDS/IPSS - intermediate 2 vs MDS/IPSS high risk vs MDS related acute myeloid leukemia). Patients are randomized to one of two treatment arms. Arm I: Patients receive busulfan IV over 2 hours every 6 hours on days -7 to -4 for a total of 16 doses. Arm II: Patients receive cyclophosphamide IV over 2 hours on days -5 and -4. Patients receive total body irradiation (TBI) twice a day on days -3 to -1; peripheral blood stem cell transplantation from genotypically HLA identical sibling on day 0; cyclosporine IV every 12 hours on days -1 to 60, and then tapering in the absence of graft versus host disease; and methotrexate IV on days 1, 3, 6, and 11. Patients are followed every 6 months for 5 years.

PROJECTED ACCRUAL: A total of 240 patients (120 per treatment arm) will be accrued for this study over 5 years.