[Cellular sorting methods in autologous (elimination of residual malignant cells) or allogeneic (T lymphocyte elimination) bone marrow grafts. Clinical pilot studies].

Ex vivo bone marrow treatment prior to autologous bone marrow transplantation for acute leukemia or solid tumors (known for their frequent medullary metastatic dissemination) has provoked a great deal of hope and enthusiasm. If, as in most cases, the feasibility of various methods (exposure to chemical products, monoclonal antibodies and complement-dependent cytolysis, immuno-magnetic procedures) has been confirmed, no study to date has shown the efficacy. Only randomized studies will result in such a proof, on the condition that all the specific technical parameters of each method have been perfectly defined. The clinical results of pilot evaluation studies for the prevention of graft-versus-host reaction through the elimination of T lymphocytes in the donor graft have been encouraging. They have shown the feasibility and efficacy of different methods of T cell depletion. One point however remains controversial: is it necessary or not to eliminate all T lymphocytes? After a depletion which is too efficient, the difficulties involved in grafting the donor's hematopoietic cells on the recipient represent a new problem, e.g., in allogenic grafts for acute leukemia (host-versus-graft). In order to resolve the problem of these take failures (10 to 20% of the published series except for Royal Free Hospital and Besançon) it appears necessary to redefine the conditioning regimen prior to allogenic bone marrow transplantation.