Graft-versus-host disease following transplantation of 'one log' versus 'two log' T-lymphocyte-depleted bone marrow from HLA-identical donors

Abstract

Prevention of acute graft-versus-host disease (GVHD) after allogeneic bone marrow transplantation, requires the depletion of mature T-lymphocytes from bone marrow grafts. The optimal degree of T-cell reduction is still an open question. We compared two procedures of T-cell separation in 18 consecutive recipients of genotypically HLA-matched bone marrow, who also received cyclosporin A for 6 months. The first method (A) was based on a discontinuous albumin gradient fractionation and resulted in an average T-lymphocyte content of 50 X 10(5)/kg body weight (n = 9 patients); the second method (B) was based on E-rosette sedimentation and reduced the contamination to 15 X 10(4) grafted T-lymphocytes/kg body weight on the average (n = 9 patients). Thus, approximately 90 and 99% of the original T-lymphocytes were removed from the marrow grafts respectively. Of the seven patients of the first group who were at risk of GVHD (excluding two cases of early death), five developed a minimal-to-moderately severe acute GVHD and in two cases chronic GVHD ensued. Lethal GVHD was not seen. Of group B, all recipients engrafted and none developed GVHD (0/9). The difference in the frequency of GVHD between the two groups was highly significant (P less than 0.0025). These data confirm our preclinical studies. They demonstrate that a one-log T-lymphocyte reduction of the marrow inoculum, when combined with cyclosporin A prophylaxis after major histocompatibility complex (MHC)-matched transplantation, is still associated with a considerable incidence of GVHD, whereas a two-log reduction of T-lymphocytes may provide full protection against acute GVHD.