Early host CD8 T-cell recovery and sensitized anti-donor interleukin-2–producing and cytotoxic T-cell responses associated with marrow graft rejection following nonmyeloablative allogeneic bone marrow transplantation

Objective. We developed a nonmyeloablative conditioning regimen for allogeneic bone marrow transplantation (BMT) followed by donor lymphocyte infusions (DLI) for treatment of chemotherapy refractory malignancies. Although the majority of patients who receive this regimen achieve lasting mixed or full allogeneic chimerism, approximately 30% show initial mixed chimerism followed by loss of the donor graft. These patients recover host hematopoiesis without significant cytopenias. To assess the role of immunologic rejection in graft loss, we compared T-cell recovery and in vitro alloresponses in six patients who lost their marrow graft to that in 16 concurrent patients with sustained donor chimerism. Patients and Methods. Conditioning included pretransplant cyclophosphamide (150–200 mg/kg), thymic irradiation (700 cGy), and pre- and post-transplant equine antithymocyte globulin (ATG; ATGAM). HLA-identical related donor BMT was followed by DLI at approximately day 35 in patients without graft-vs-host disease. Results. The group with transient chimerism showed significantly increased circulating host T-cell (median 416 cells/mm3 vs 10 cells/mm3, p<0.05) and CD8 T-cell numbers (354 cells/mm3 vs 71 cells/mm3, p<0.05) compared to the group with stable mixed or full donor chimerism within the first 100 days post-BMT. All DLI recipients who lost chimerism following DLI had greater than 80% recipient T cells at the time of DLI, whereas those with persistent chimerism had <60% host T cells. Graft rejection was associated with the development of a sensitized anti-donor bulk cytotoxic T-lymphocyte (CTL) response in 4 of 6 evaluated patients, compared to only 1 of 10 evaluated patients with sustained chimerism (p<0.05). Additionally, 3 of 5 evaluated transient chimeras showed high anti-donor CTL precursor frequencies in limiting dilution assays, and 3 of 4 evaluated transient chimeras showed high anti-donor interleukin-2 (IL-2)–producing T-helper (TH) cell frequencies. High anti-donor TH or cytotoxic T-lymphocyte precursors were not detected in sustained chimeras. Conclusion. These data indicate that loss of chimerism in patients receiving this nonmyeloablative regimen is due to immune-mediated rejection. This rejection appears to bemediated by recovering recipient cytolytic CD8+ cells as well as IL-2–producing recipient TH cells. These data are the first to demonstrate sensitization of recipient anti-donor IL-2–producing cells in association with human marrow allograft rejection.