CD8-Depleted Donor Lymphocyte Infusions Convert Mixed into Complete Donor T-Cell Chimerism after T-Cell Depleted Allogeneic Stem Cell Transplantation.

Abstract Donor lymphocyte infusions (DLI) are increasingly used to treat minimal residual disease or mixed hematopoietic donor-recipient chimerism in T-cell depleted allogeneic stem cell transplantation (SCT). In addition, several clinical trials currently investigate the prophylactic application of DLI to promote donor T-cell reconstitution after transplantation. However, DLI carry a substantial risk of inducing graft-versus-host disease (GVHD). We investigate DLI heavily depleted of CD8 T cells using a clinical grade immunomagnetic in vitro procedure in an ongoing clinical study [Meyer et al., Blood2007, 109:374]. These DLI are administered in a prophylactic setting to patients with hematological malignancies who are off immunosuppressive treatment and free of GVHD early after allogeneic SCT. The reduced-intensity conditioning regimen consists of fludarabine and melphalan and in vivo T-cell depletion (TCD) by the anti-CD52 antibody Alemtuzumab. Up to now, 24 patients have been treated with 1 to 4 increasing doses of CD8-depleted DLI starting with 1x10^6 CD4+ T cells/kg bodyweight. The median time between SCT and first DLI was 119 days (range, 60–194). Seven of 24 patients (29%) developed acute GVHD of grade 2 to 4 or extensive chronic GVHD following CD8-depleted DLI. We did a longitudinal analysis of lineage-specific T-cell chimerism in 20 patients who received CD8-depleted DLI in comparison to 17 patients who did not qualify for DLI due to spontaneously occurring acute GVHD (n=14) or unavailable donors (n=3). The patients’ characteristics in both groups were comparable with a median age of 55 (range, 35–64) years in the DLI group and of 57 (range, 29–67) years in the non-DLI group. The donor types were matched sibling (DLI: n=6; non-DLI: n=2), matched unrelated (DLI: n=8; non-DLI: n=8), and unrelated with 1 HLA-mismatch (DLI: n=6; non-DLI: n=7), respectively. Twelve of the 20 patients in the DLI group and 6 of 17 patients in the non-DLI group showed a secondary decrease of donor T-cell chimerism to a median of 52% (range, 10–90%) between 7 and 35 weeks after transplantation (median, 12 weeks). Only one of the latter spontaneously reconverted to a full donor T-cell chimera. Of the remaining 5 non-DLI patients, 4 patients subsequently relapsed with their underlying disease and one patient still had a mixed T-cell chimerism of 50% two years after transplantation. In contrast, in patients receiving CD8-depleted DLI the proportion of donor T cells significantly increased and 11 of 12 patients converted to durable full donor chimerism. Three patients of the DLI group subsequently developed disease relapse. By monitoring CD52-expression on reconstituting T cells by flow cytometry, we were able to demonstrate the impact of CD8-depleted DLI on post-transplant T-cell reconstitution: In the non-DLI group, the majority of CD4 T cells remained CD52-negative 9 months after transplantation. Simultaneously, the proportion of CD52-expressing CD4 T cells was significantly higher in the DLI group (mean: 42% versus 86%; t-test, p<0.01). A similar difference was not observed in the CD8 T-cell lineage. Our results show that the impact of adoptively transferred CD4 T cells after anti-CD52 antibody-mediated TCD can be readily demonstrated by CD52 staining of reconstituting T cells. In addition, CD8-depleted DLI can be efficiently applied to switch mixed into complete donor T-cell chimerism. On the basis of our data we propose a prospective randomized trial in patients with a secondary decrease of donor T-cell chimerism that investigates if CD8-depleted versus non-depleted DLI lead to an improved outcome with regard to disease control and GVHD.