Allograft T Cell Content Influences Early Engraftment after Reduced-Intensity Stem Cell Transplantation (RIST).

Allogeneic hematopoietic stem cells (HSC) generally engraft rapidly and completely after myeloablative conditioning. However, with reduced-intensity conditioning (RIC), mixed chimerism and graft failure are more common. Host immune status and HSC number are factors known to affect engraftment after reduced-intensity stem cell transplantation (RIST). In addition, donor T cells within the allograft may also influencethe kinetics of donor engraftment after RIST. To evaluate this, we performed a controlled comparison of engraftment outcomes among 3 groups undergoing RIST, varying by ex vivo T cell depletion (TCD) or in vivo depletion of activated T cells with methotrexate (MTX) to prevent graft-versus-host disease (GVHD). Group I (n = 50) received T cell replete (TCR) peripheral blood stem cells (PBSC) with cyclosporine (CSA) alone for GVHD prophylaxis. Group II (n = 17) received ex vivo TCD PBSC (positive/negative selection with T cell add-back to uniform dose of 1 x 10⁵ CD3⁺ cells/kg) with CSA alone for GVHD prophylaxis. Group III (n = 31) received TCR PBSC with CSA plus MTX (5 mg/m² IV x 4 doses) for GVHD prophylaxis. The 3 groups were similarly immunosuppressed from prior therapy before RIST (median absolute lymphocyte counts 330/μL, 260/μL, and 307/μL for Groups I, II, and III, respectively), and received an identical RIC regimen (fludarabine/cyclophosphamide) plus comparable numbers of filgrastim-mobilized PBSC from HLA-matched sibling donors (median 7.9 x 10⁶, 7.6 x 10⁶, and 6.8 x 10⁶ CD34⁺ cells/kg, respectively; median 3.6 x 10⁸, 1.0 x 10⁵, and 3.2 x 10⁸ CD3⁺ cells/kg, respectively). Hematopoietic recovery was slowest in Group III, consistent with the myelosuppressive effects of MTX (Table). A greater proportion of patients in Group I achieved complete donor chimerism (≥ 95%) by day +28 than in Groups II or III (P < 0.025), and at day +100, mixed donor chimerism persisted more often in Groups II and III than in Group I patients (P < 0.01). Correspondingly, early (< day +42) occurrence of grade 3–4 acute GVHD, before initiation of planned sequential donor lymphocyte infusions (DLI) in Group II, was more frequent in Group I than in either Groups II or III (p=0.08).

Thus, the deletion of T cells by either ex vivo TCD or in vivo MTX administration measurably alters the kinetics and degree of donor T cell engraftment after RIST. These observations provide evidence that donor T cells are an independent factor affecting engraftment of allogeneic HSC after RIST by compensating for incomplete host immune ablation. These data also support the hypothesis that a graft-versus-host effect plays a significant role in engraftment after RIST. Manipulation of donor T cells through graft engineering techniques may be a useful strategy to enhance engraftment in the setting of RIST.