Outcome of allogeneic peripheral blood stem cell transplantation by donor graft CD3+/Tregs ratio: a single-center experience.Biol Blood Marrow Transplant. 2013 Mar; 19(3):495-9.BB
The therapeutic efficacy of allogeneic peripheral blood stem cell transplantation (PBSCT) for hematological malignancies relies largely on the graft-versus-leukemia (GVL) effects exerted by the donor CD3 cells, but there is a risk of onset of uncontrolled graft-versus-host disease (GVHD). Regulatory T cells (Tregs) (CD4+CD25(high) Foxp3+) are believed to maintain tolerance and to inhibit acute GVHD (aGVHD) after allogeneic PBSCT. Nevertheless, when looking at post-allotransplantation patient outcomes, although the impact of aGVHD on survival is amply documented, so far there is no evidence that the donor graft CD3/Tregs ratio may affect overall survival (OS), nonrelapse mortality (NRM), disease-free survival (DFS), and relapse rates. Our aim was to study the possible impact of the gCD3/Tregs ratio on survival after myeloablative allogeneic PBSCT. We analyzed 74 consecutive patients diagnosed with acute myeloid leukemia (n = 62), acute lymphoblastic leukemia (n = 10), and chronic myeloid leukemia (n = 2) who underwent transplantation with unmanipulated PBSCs from a human leukocyte antigen-identical related donor (n = 48) or a human leukocyte antigen-identical unrelated donor (n = 26). Patients were subdivided into a high gCD3/Tregs ratio (≥36) group (HR group, n = 30) and a low gCD3/Tregs ratio (<36) group (LR group, n = 44). The OS, DFS, NRM, and relapse rates at 3 years were 53%, 51%, 29%, and 34%, respectively. Comparing the LR and HR groups, a statistically significant difference was demonstrated for the 3-year OS, DFS, and NRM rates (65% vs 31%, P = .0001; 67 versus 26%, P = .0001; 5% versus 71%, P < .0001, respectively) but not for relapse (30% vs 25%, P = ns). By multivariate analysis, LR significantly predicted better OS (P = .019), DFS (P = .003), and NRM (P = .05), whereas there was no statistically significant association between LR and relapse (P = .155). Overall, our data may suggest that LR preserves GVL effects but is also protective against aGVHD in allotransplantation patients.