The graft-vs.-leukemia effect. Implications for post-marrow transplant antileukemia treatment.

PURPOSE

Over the past 30 years, evidence supporting the existence of an antileukemic effect of allogeneic bone marrow has rapidly expanded. Early animal studies demonstrated this "graft-vs.-leukemia" (GVL) effect in association with graft-vs.-host disease (GVHD).

PATIENTS, METHODS, AND RESULTS: Clinical evidence for the GVL effect derived from experience in allogeneic bone marrow transplantation (BMT) includes the following: the association of GVHD, both acute and chronic, with a decreased rate of leukemic relapse, the occurrence of a higher relapse rate following identical twin transplants compared to allogeneic major histocompatibility complex-matched sibling transplants, the decrease of GVHD and increase of leukemic relapses with T-cell depletion of donor bone marrow, and that allogeneic marrow transplants without GVHD have a lower leukemic relapse rate compared to identical twin transplants and T-cell depleted transplants. In addition, some reports indicate that modifying the immunosuppressive regimens (consisting of cyclosporine A or methotrexate) that are aimed at prevention of GVHD can affect leukemic relapse rates. Although the mechanism of this GVL effect is poorly understood, there are several favored hypotheses. Cytotoxic T cells, which may mediate GVH effects, could also mediate a GVL effect. There may also be T cells with specificity for destroying leukemic blasts alone. The natural killer/lymphokine-activated killer cytotoxic cell may play a role in the GVL effect. These nonspecific cytotoxic cells can be activated with cytokines induced in the transplant setting. Finally, the GVL effect could be secondary to the in vivo induction of various cytokines that either have direct antileukemic effects or that induce leukemic differentiation. Advances in our basic understanding of the immune system and its activation, and the availability of purified immunoregulatory cytokines, have revealed new ways to induce GVL-like effects. Promising results with therapy based on use of interleukin-2 in solid tumors have been seen. It is hoped that these advances will enable the "antileukemic" components of the GVL effect to be prospectively controlled (as part of the technique of marrow transplant) and intentionally used as leukemic therapy in patients undergoing bone marrow transplantation.