INTRODUCTION. While total CD34 counts in PBSC graft products have correlated with overall likelihood of hematopoietic recovery after allogeneic hematopoietic stem cell transplantation (HSCT), analyses of the HSPC composition and its relationship to relevant post-transplant clinical outcomes are lacking. In fact, the biological basis for different dynamics of hematopoietic/immune recovery, the risk of infection, and graft-versus-host disease (GvHD) is not fully understood. We have performed the first analysis of HSPC graft composition in 6 αβ T-cell/CD19 B-cell depleted haploidentical (αβhaplo) HSCT. Additionally, we correlated the HSPC graft composition with the distribution of the same HSPC subsets in serial post-HSCT bone marrow aspirates performed at Days 30, 60, and 90, with the peripheral blood counts [neutrophils, monocytes, platelet (PLT)] and with the immune recovery (CD3+, CD3+CD4+, CD3+CD8+, αβT, γδT, NK cells) at the same time points. The patients were divided in two groups: 3 patients had a robust and sustained hematopoietic recovery (Group 1) while 3 patients experienced mild cytopenia after Day 60 (Group 2). All patients were transplanted for acute leukemia and received a myeloablative TBI-based conditioning regimen. See Table 1 for details about patients and graft composition.

METHODS. All patients were enrolled in the Stanford IRB approved BMT Protocols 179/351/361 and had peripheral blood (PB) and bone marrow (BM) evaluated at Day 30, 60 and 90 post HSCT for the primitive CD34+ Lin- HSPC subsets: HSC (CD38-CD45RA-CD90+), MPP (CD38+CD45RA+), CMP (CD38+CD45RA-CD123+), GMP (CD38+CD45RA+CD123+), MEP (CD38+CD45RA-CD123+) and CLP (CD38+CD127+). Aliquots of αβhaplo-HSCT grafts were cryopreserved for later analyses. Mononuclear cells were isolated from PBSC, PB and BM by Ficoll-Hypaque (Sigma-Aldrich) density gradient centrifugation. FACS analyses were performed on either fresh or frozen cells on Becton Dickinson (BD) Aria II flow cytometer. At least 5x105 events were acquired and analyzed using FlowJo software (BD).

RESULTS. Despite consistent levels of αβT-cell depletion and CD34 enrichment, the frequency of the HSPC subsets varied between the grafts. Notably, while CMP and GMP were very consistent across the 6 grafts, the frequency of HSC, CLP, MEP and MPP showed a 2-fold range of variation (Fig1A). No significant correlation was observed between the HSC frequency in the graft and the hematopoietic/immune recovery. However, the frequency of HSC in the BM at Day 30 is statistically correlated (P=0.027) with the PLT at Day 90 (Fig1B). In these preliminary results, the different distribution of CMP, GMP, MEP and MPP did not impact on the hematopoietic/immune recovery. However, there was a significant correlation (P=0.02) between CLP and γδ T cells reconstitution at Day 90 in Group 1 patients (Fig1C). Additionally, the neutrophils, monocytes and phagocytic cells recovery at Day 90 is statistically correlated with the GMP frequency in the BM at Day 30 (P=0.017; P=0.018; P=0.0132, respectively) (Fig1D). Interestingly, the same strong correlation is observed between the CMP in the BM at Day 60 and the recovery of neutrophils and phagocytic cells (P=0.016, P=0.019) at Day 90 (Fig1E), but the CMP at day 30 are already predictive of a robust engraftment in the Group 1 patients (data not shown).

CONCLUSIONS. In αβhaplo-HSCT, previously identified factors influencing hematopoietic recovery have been mainly limited to the enumeration of bulk CD34 counts and of mature effector cells, such as αβ/γδ T and NK cells. On the other hand, the presence of GvHD and thymic injury have been correlated to the kinetics of immune reconstitution. We hypothesized that the HSPC composition of the graft would impact lymphohematopoietic recovery in αβhaplo-HSCT recipients. Although preliminary, our data indicate that even with a consistent method of graft manipulation, the HSPC graft composition is heterogeneous. Variations in HSPC subsets frequency and number can contribute to significant differences in lymphohematopoietic recovery and, therefore, clinical outcome. The evaluation of a larger number of patients with longer follow up after HSCT are required. Comparative studies with unmanipulated T-cell replete and cord blood grafts are ongoing. Such analyses will be instrumental not only for prediction of clinical outcome, but also for optimization of novel graft engineering strategies.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.