Campath-1H-Based Reduced-Intensity Conditioning Followed by Allogeneic Blood Stem-Cell Transplantation and Preemptive CD8-Depleted Donor Lymphocyte Infusions.

Recent reports have demonstrated the feasibility of using the anti-CD52 antibody Campath-1H for in vivo T-cell depletion (TCD) in the context of a fludarabine/melphalan-based reduced-intensity conditioning regimen (Kottaridis et al. Blood 2000, 96:2419). Major disadvantage of this protocol is the severe immunosuppression which results in an increased rate of opportunistic infections and disease relapses. Donor lymphocyte infusions (DLI) are frequently used to overcome this limitation. The application of DLI, however, is associated with a profound risk of GvHD. Depletion of CD8+ lymphocytes from either the allotransplant or from DLI has proven feasible to reduce the incidence of GvHD (Nimer et al. Transplantation 1994, 57:82; Giralt et al. Blood 1995, 86:4337).

We set up a phase-I allotransplantation study which combines the Campath-1H (100mg)/fludarabine (150mg/m²)/melphalan (140mg/m²) preparative regimen with the preemptive administration of CD8-depleted DLI in patients with hematological malignancies not eligible for myeloablative conditioning. After early withdrawal of CsA immunosuppression, CD8-depleted DLI are scheduled in a dose-escalating regimen starting on day +60 after HLA-identical sibling transplantation and on day +120 after unrelated donor transplantation, respectively. For CD8-depletion of donor leukaphereses a new CliniMACS-based protocol using clinical grade CD8 Microbeads is used. Our data show that this procedure is efficient in reducing the content of CD8⁺ T-cells by at least 2.5 to 3.5 logs, while CD4⁺ cells remain largely unchanged.

In this ongoing study, 8 patients (4 myeloma, 2 CMPD, 1 ALL, 1 NHL) were included and received peripheral blood stem cells (PBSC) from HLA-identical sibling (1), HLA-matched (5) or DRB1-mismatched unrelated donors (2), according to high-resolution HLA typing. Chimerism analyses on total bone marrow-derived cells as well as on purified CD4, CD8, CD19, CD14 and CD15-positive peripheral blood cells showed durable complete donor chimerism in all but one patient (see below). After successful withdrawal of CsA, 2 patients already received CD8-depleted DLI without any acute toxicity. In the first patient (UPN 3), who was transplanted from an HLA-identical sibling, DLI were given on days 60 and 120 (1 and 5 x 10⁶ CD4⁺ T cells, respectively). The second patient (UPN 4), who was transplanted from a matched unrelated donor, received the first dose of 1 x10⁶ CD8-depleted DLI on day 120. DLI were regularly followed by a more than 2-fold increase of CD4⁺ cells during the first two weeks after application, while CD8⁺ cells remained unchanged. After the second DLI, patient UPN 3 developed a grade 2 GvHD of the skin (d+22 after DLI). Both patients were at high risk for CMV reactivation (donor: CMV negative, host: CMV positive) and repeatedly received preemptive treatment. Immediately prior to the second DLI, the CD8⁺ cell count in patient UPN 3 increased during an ongoing CMV reactivation. Line-specific chimerism analysis revealed that more than 50% of these CD8⁺ cells were of host origin. As expected, transfer of CD8-depleted DLI from CMV-negative donors was unable to entirely prevent CMV reactivation in seropositive patients. Nevertheless, our preliminary data suggest that the preemptive transfer of CD8-depleted DLI is able to accelerate immune reconstitution after reduced intensity TCD allogeneic PBSCT without rapid increase of CD8⁺ donor cells.