The graft-versus-leukemia effect.

Although the graft-versus-leukemia effect was predicted from animal experiments almost 40 years ago, only recently has its role in clinical bone marrow transplantation become better defined. The graft-versus-leukemia effect was initially considered to be a minor component of the graft-versus-host reaction. However, more recent analyses of clinical bone marrow transplants and the successful treatment of relapsed leukemia with donor lymphocyte transfusions indicate that the graft-versus-leukemia effect can be very powerful and to some degree independent of graft-versus-host disease. Results of donor lymphocyte transfusions draw attention to wide variability in response according to the nature of the relapsed leukemia and also to additional therapies that may enhance the efficacy of donor lymphocyte transfusions. Although natural killer lymphocytes may play a role in the graft-versus-leukemia effect, attention is focusing increasingly on T lymphocytes as potent mediators of the effect. Both CD4+ and CD8+ cells, together with the cytokines they produce, can interact with leukemia cells to deliver a cytotoxic hit on the malignant clone. Although leukemia cells undoubtedly share common antigens with other tissues of the recipient resulting in nonspecific graft-versus-host and graft-versus-leukemia reactions, there is also the possibility that distinct antigens are presented to T cells by the leukemia cells, providing a basis for separating the graft-versus-leukemia from graft-versus-host reactions. The current clinical challenge is to devise strategies for separating the graft-versus-leukemia response from graft-versus-host disease.