Peripheral Blood Grafts Vs Bone Marrow Grafts in Young Acute Myeloid Leukemia Patients with HLA-Matched Sibling Donors Undergoing Myeloablative Conditioning

In allogeneic hematopoietic cell transplantation (allo-HSCT), graft source has been changed from bone marrow grafts (Bone marrow stem cell, BMSC) to peripheral blood grafts (Peripheral blood stem cell, PBSC) mobilized by granulocyte colony-stimulating factor. PBSC transplantation is known to have numerous advantages over BMSC, including faster engraftment, improved survival outcomes, and for the donor, the elimination of invasive marrow harvesting procedure under general anesthesia. However, previous studies regarding graft source included various hematologic malignancies and lacked consistency in conditioning intensity and disease status at allo-HSCT. Still, there are limited data on the comparison of graft sources specifically in patients with acute myeloid leukemia (AML) undergoing myeloablative conditioning. In this background, we conducted an analysis to investigate the impact of graft source on transplant outcomes in AML patients who underwent HLA A, B, C, DR-matched sibling donor (MSD) HSCT with myeloablative conditioning.

This study was a retrospective, single-center study for patients with AML who received first allo-HSCT from 2002 to 2022. We included the patients aged < 60 at the time of allo-HSCT, in a state of complete remission (CR) or CR with incomplete hematologic recovery, from a MSD, and underwent myeloablative conditioning. We classified the patients into three groups: the BMSC graft group (BMSC group), the PBSC graft group without antithymocyte globulin (ATG) (PBSC/ATG- group), and the PBSC graft group with ATG (PBSC/ATG+ group). Patients who received both PBSC and BMSC grafts were excluded from the analysis. We examined the differences in overall survival (OS), relapse-free survival (RFS) among these three groups.

A total of 480 patients were included in this analysis, with 234 patients in the BMSC group, 66 patients in the PBSC/ATG+ group, and 180 patients in the PBSC/ATG- group. BMSC transplantation was mostly performed before 2010, whereas PBSC transplantation started in 2007, and allo-HSCT with ATG administration were conducted after 2014. In the PBSC/ATG+ group, the median ATG dose was 2.5 mg/kg (Interquartile range, IQR: 1.6-2.5). The BMSC group tended to be younger at the time of HSCT (median 38 years [31-46]) compared to the other groups (median age of PBSC/ATG+ group: 44 years [37-51], PBSC/ATG- group: 46 years [39-51], p<0.01). Of total patients in the three groups, there was only 1 case of neutrophil engraftment failure (in the PBSC/ATG- group) and 5 cases of platelet engraftment failure (3 in the BMSC group and the other 2 in the PBSC/ATG- group). BMSC patients demonstrated superior OS compared to PBSC patients, with the most pronounced difference was observed over PBSC/ATG- group (Figure A). Similar trends were presented in relapse-free survival (RFS) as well. After adjusting for other covariates through multivariate Cox modeling, the difference caused by the graft source remained significant, particularly in terms of RFS (Figure B). The hazard ratios for RFS of PBSC comparing to BMSC were as follows; PBSC/ATG+: 1.64 [1.04-2.58], PBSC/ATG-: 1.95 [1.45-2.63]. In conclusion, Despite the clear differences in allo-HSCT era between groups, even BMSC patients underwent allo-HSCT at an earlier time, patients receiving BMSC grafts showed superior survival compared to those receiving PBSC grafts in young AML patients undergoing MSD myeloablative allo-HSCT. Parallel evaluation of the transplant outcomes between BMSC and PBSC would be crucial for assessing the advantages and disadvantages of PBSC and BMSC grafts in this setting.