Allogeneic bone marrow transplantation is an effective first line and salvage therapy in-patients with AML, CML or MDS. In the past, allogeneic transplantation has been limited to younger patients due to the increased risk of regimen-related toxicity and graft-versus-host disease (GVHD). In order to expand the use of alloBMT to older patients, researchers at MD Anderson Cancer Center have pioneered the use of nonmyeloablative preparative therapy for allogeneic transplantation. Nonmyeloablative therapy utilizes low-dose chemotherapy in order to decrease regimen-related toxicity. The low-dose chemotherapy induces host immunosuppression, spares toxicity, and allows engraftment of donor stem cells. The goal of this therapy is to exploit the anti-tumor effects of donor immune cells.

One shortcoming of nonmyeloablative transplantation is an increased risk of disease relapse, especially in patients with high-risk disease. This includes patients with relapse or refractory AML, advanced MDS and advanced CML. Mylotarg is a novel immunotoxin directed against the CD33 antigen found on most AML, CML and MDS clones. Mylotarg has been shown to have significant anti-leukemia activity with little toxicity. The goal of this Phase I/II protocol is to evaluate the safety and efficacy of adding escalating doses of Mylotarg to Fludarabine and Melphalan in patients with CML, MDS or acute myelogenous leukemia undergoing related, mismatch-related or matched unrelated donor allogeneic transplantation. The hypothesis is that Mylotarg will provide potent anti-leukemic effects without additional toxicity. Mylotarg, Melphalan, and Fludarabine have non-overlapping toxicities. A more potent anti-leukemic response may increase the complete remission rates and induce a state of minimal residual disease (MRD). Therefore, the GVL effect will have a better chance for success.