Background

Use of G-CSF-primed bone marrow (GBM) graft can increase harvest yield, improve engraftment, and could influence graft composition promoting tolerogenic cell expansion. We analysed transplant outcomes of 61 consecutive patients receiving GBM grafts from HLA-haploidentical donors. To further explore influence of stimulation on graft composition we analysed T cell, NK cell and dendritic cell subsets in 14 GBM grafts in comparison to unstimulated bone marrow grafts.

Patients and Methods

From 5/2012 to 12/2021 61 patient received haploBMT using PT-CY platform. Donors received G-CSF (10 μcg/kg BW sc) on days -2, -1 and 0 before BM harvest. Fifty-four patients (88.5%) received non-myeloablative FluCyTBI conditioning, 7 patients (11,5%) received myeloablative conditioning (BuCy or TEC RIC).

Graft composition analysis was performed in 14 GBM grafts, while 6 bone marrow samples collected from donors without stimulation were used as controls. Unseparated cells from fresh samples were stained with appropriate fluorochrome-conjugated antibodies for dendritic cells (PDC's, MDC1's and MDC2's), lymphocytes subpopulations (CD3+ T cells, CD4+CD3+ T cells, C8+CD3+ T cells, CD3+CD4+CD127dimCD25+ Tregs, CD3+CD56+ NKT cells, CD56dimCD16bright NK cells, CD56dimCD16- NK cells, CD56brightCD16- NK cells, CD56brightCD16dim NK cells and CD56-CD16bright NK cells), iNKT cells analysis and for CD34 positive cells enumeration. The FACS Lyric was used for the acquisition of samples, data were analyzed by FACSuite ver1.2 software (BD Biosciences, San Jose, USA).

Survival curves were plotted using the Kaplan-Meier estimates. R software was used to calculate the cumulative incidence when considering the presence of competing risks. Graft composition was determined using the Mann-Whitney test, value of p < 0.05 was statistically significant.

Results

Patient median age was 44 years (20-68), 29 were female and 32 male. Twenty-one patient had AML, 20 HL, 7 ALL, 6 NHL, 4 MDS, 1 CML, 1 AA and 1 bone marow failure syndrome. Fifty-two patients (85%) were in remission, 5 (8%) lymphoma patients in PR, and 2 AML patients had minimal residual disease. In 10 (16%) patients this was the second allogeneic transplantation. Median number of infused TNC was 5.04x108/kg BW (1.76-12.93); CD34+ cells 1,65 x106/kg BW (0.28-4.47). Median follow up was 40 months (range 6-90). Engraftment was achieved in 56 (92%) patients, 2 patients (3%) had primary, 2 patients (3%) later had secondary rejection, 3 patients (5%) died in aplasia.

Median days to neutrophil recovery (ANC>500) was 23 (12-38), median days to platelet recovery (PLT>20x109/L) was 33 (15-81). In all patients MMF was discontinued at D+35. Cumulative incidence of grade II-IV aGVHD was 20% (95% CI 11-30), just 3 cases developing grade III-IV aGVHD, one after DLI for failing chimerism, one after stopping IS for failing chimerism on day 83.

Cumulative incidence of chronic GVHD was 4% (95% CI 1-11). Cumulative incidence of relapse was 29% (95% CI 18-41). Overall survival was 66% (95% CI 55-79).

G-CSF stimulation resulted in significantly higher CD34 graft count (p=0.0066) and higher CD3+CD4+ cell count (p=0.0154). Treg cell percentage was increased in reference to lymphocytes (p= 0.0075) and CD4+ T cells (p=0.0146) resulting in higher absolute values in GBM (p=0.0015). We found significantly lower percentage of NK cells in GBM group (p=0-0002). There was no difference in the CD56bright NK cell population, while percentage of CD56dim NK cells was significantly decreased (p<0.0001) and iNKT cells were significantly increased in GBM group (p=0.0381). Total number of pDC cells was significantly higher in GBM group (p=0.0191).

Conclusion

G-CSF primed bone marrow is beneficial not only in terms of TNC yield and better engraftment but also in terms of increasing CD4+ T cell count and more importantly Tregs count. It also influences NK cells, as well as dendritic cell compartments. Using GBM graft probably contributed to better GVHD control, as we showed the incidence of both aGVHD and particularly cGVHD to be lower than expected, without the loss of GvL effect (low incidence of relapse considering difficult to treat patients). While further studies in larger number of patients are warranted, this data provides additional evidence to the concept of immunomodulatory effect of GBM graft which probably should be considered, particularly in instances where GVHD control is paramount.

Vrhovac:Abbvie, Astellas, Pfizer, Pharmas: Consultancy; Abbvie, Astellas, MSD, Novartis, Pfizer, Pharmas, Servier, Teva: Speakers Bureau.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution