T-cell depleted HLA-haploidentical hematopoietic stem cell transplantation (HSCT) is a suitable option for patients in need of an allograft who lack a HLA-matched donor. Although it offers the advantage of being immediately applicable to virtually all patients, so far, graft manipulation with removal of all T lymphocyte subsets and of natural killer (NK) cells has been associated with an increased risk of life-threatening infections, as well as, in some studies, of leukemia recurrence.

We recently developed a new method of graft manipulation based on the physical removal of αβ+ T cells and CD19+ B cells, which permits to leave mature NK cells and γδ+ T cells in the graft. We, thus, started a formal study (NCT01810120) in children with acute leukemia aimed at evaluating the safety and efficacy of this approach. As of April 2013, we enrolled 45 patients (pts; 29 males, 16 females). Median age at HSCT was 10.1 years (range 0.9-17.9). Thirty-five pts had acute lymphoblastic leukemia (ALL) and 10 acute myeloid leukemia (AML); all children were transplanted in morphological complete remission (CR). Fifteen pts were transplanted in first CR, 27 in second CR and 3 in more advanced CR. All pts transplanted in CR1 had either poor cytogenetic/molecular characteristics or high levels of minimal residual disease (MRD) at the end of induction therapy. The donor was either the mother (n=25) or the father (n=20); according to the model of KIR/KIR ligand disparity, 22 pts were transplanted from an NK-alloreactive donor. The median number of CD34+ cells, NK cells, γδ+ T cells, B cells and αβ+ T cells were 14.6, 31.7, 7.8, 0.08 and 0.04x106/kg, respectively. A myeloablative regimen, containing Total Body Irradiation in 34 cases, was given to all children, who also received anti-thymocyte globulin (12 mg/kg over 3 days, from -5 to -3). Rituximab (200 mg/m2) was administered on day -1 to further prevent EBV-related lymphoproliferative disorders. No pharmacological graft-versus-host disease (GVHD) prophylaxis was employed after transplantation.

Sustained primary engraftment occurred in 44/45 pts, the remaining child being successfully re-transplanted from the other parent. The median time to reach an absolute neutrophil count >0.5x109/L and a platelet count >50x109/L was 13 days (9-18) and 11 days (8-20). No child developed gut or liver acute GVHD. Thirteen pts experienced skin-only grade I-II GVHD, this leading to a cumulative incidence (CI) of 29% (95% confidence interval, CO. IN., 18-45). Only 2 pts developed skin limited chronic GVHD. Two pts died for causes other than disease relapse (both in the first 60 days after HSCT), the CI of transplantation-related mortality (TRM) being 4% (CO. IN. 1-16). Seven pts relapsed, the CI of disease recurrence being 16% (CO. IN. 6-32). With a median follow-up of 11 months (range 2-30), the 2-year Kaplan Meier estimate of leukemia-free survival (LFS) was equal to 75% (CO. IN. 57-86); this value was 73% (CO. IN. 52-85) for pts with ALL. LFS of pts who did or did not experience skin-only acute GVHD was 83% (CO. IN. 48-96) and 72% (CO. IN. 50-86), p=NS. The probability of LFS of the 39 pts with either negative or low (<1x10-4) MRD at time of HSCT was 81% (CO. IN. 62-91) as compared to 25% (CO. IN. 1-67) of the 6 pts with positive MRD (p<0.01). The CI of cytomegalovirus (CMV) reactivation was 53% (CO. IN. 29-68). In the first 45 days after HSCT, circulating T cells mainly consisted of the γδ+ subset, while αβ+ T lymphocytes predominated thereafter. CMV reactivation promoted the expansion of Vδ1+ T cells. Mature NK cells expressing inhibitory KIRs and negative for NKG2A, including also the alloreactive cells, were detectable at 1 month in peripheral blood after HSCT, showing the persistence of transplanted cells with potential anti-leukemia activity. These percentages increased at later time points, reaching values comparable to those found in the HSCT donor.

These data indicate that a selective graft manipulation results into effective prevention of both acute and chronic GVHD, rapid recovery of neutrophil and platelet counts and low TRM. Although the median observation time is still limited, also the CI of disease recurrence is encouraging. Post-transplant add-back of limited numbers of αβ+ T cells transduced with the iCasp9 suicide gene could optimize the recovery of this subset in early weeks after HSCT, maximizing immune-mediated viral and leukemia control, without an unacceptable risk of GVHD.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.