In an attempt to reduce the incidence and severity of acute graft-versus-host disease (GVHD), we have decreased the number of bone marrow (BM) lymphocytes in the donor marrow graft before bone marrow transplantation (BMT) using counterflow centrifugal elutriation (CCE). In a phase I-II clinical trial, 23 patients received lymphocyte-depleted BM allografts from their HLA-identical, mixed lymphocyte culture (MLC)-nonreactive sibling donors. The patients entered in the study were deemed to be at high risk for treatment failure on the basis of age (greater than 30 years; median, 39 years) and the result of our skin explant assay predictive of acute GVHD. Patients predicted not to develop acute GVHD by this assay were excluded from this study. All patients received a standard lymphocyte dose of 0.5 x 10(6) morphologic lymphocytes per kilogram ideal body weight (IBW) in the marrow graft and were maintained on cyclosporine A (CsA) immunosuppression for 170 days after BMT. Prompt hematopoietic recovery occurred in 22 of 23 patients with a median time to an absolute neutrophil count (ANC) greater than or equal to 500/microL of 21 days. Donor cell engraftment was subsequently verified by cytogenetic and/or DNA analysis in all of 21 evaluable patients. No patient developed systemic acute GVHD. Only five (22%) developed cutaneous GVHD (clinical stage 1) that required steroid treatment, including one patient who failed to engraft. The median follow-up of the patients enrolled in this study is 14 months (range, 5 to 20 months). Actuarial survival 1 year after BMT is 83%. Thus, in two consecutive clinical trials using CCE to deplete donor BM of alloreactive lymphocytes (1.0 x 10(6) versus 0.5 x 10(6) lymphocytes/kg), we have demonstrated that the procedure does not interfere with BM engraftment and is effective in decreasing the incidence and severity of acute GVHD. Furthermore, comparison of these studies has revealed a differential dose response relationship between the number of graft lymphocytes, protection of engraftment, and induction of acute GVHD. Although there appears to be a modest relationship between lymphocyte dose and time to hematopoietic recovery, the 50% reduction in lymphocyte dose from that used in our previous trial resulted in a marked decrease in acute GVHD without compromising engraftment.