Effects of Cord Blood Cell Subset Populations in the Development of the Dominant Cord Blood Unit in Non-Myeloablative Sequential Double Cord Blood Transplantation (DCBT).

Umbilical cord blood has become an established alternative hematopoietic stem cell source for patients without suitable donors. However, single cord blood transplantation in adults has been associated with high mortality rates due to graft failure, delayed engraftment, and poor immune reconstitution. In order to improve engraftment, sequential cord blood transplantation has been employed. One consistent observation with this strategy has been the eventual dominance of one cord graft over the other. However, the factors that determine the fate of one cord over the other have not been well elucidated. Our previous work demonstrated that the dominant cord tends to have an initial higher CD34⁺ cell dose. This study attempted to identify other cell populations in the cord units that may determine which unit becomes the ultimate source of hematopoiesis. Forty patients with hematologic malignancies underwent non-myeloablative conditioning with fludarabine, melphalan, and ATG followed by two sequential cord blood infusions. All cord units were at least a 4/6 HLA match with the recipient and each other. GVHD prophylaxis consisted of cyclosporine & mycophenolate mofetil or FK-506 & rapamycin. Chimerism analysis of peripheral blood leukocytes was performed by PCR of short tandem repeat loci. Flow cytometry for T-cells (CD4 and CD8), B-cells (CD19), and NK cells (CD3⁻CD56⁺CD16⁻ & CD3⁻CD56⁺CD16⁺) were done using standard immunofluorescence methods. A total of 10 patients were available for this analysis. The Wilcoxon signed rank test was used to evaluate the significance of the difference in cell populations between the transplanted cords on graft outcome. This analysis demonstrated that 80% (8 of 10) of the dominant cords had higher NK cells when compared to the rejected unit with average total NK cells of 4.15 x 10⁶ vs 2.30 x 10⁶ per kg of recipient, respectively. Total NK cells included both CD16⁻ and CD16⁺ cells, but the difference was much higher for CD16⁺ NK cells, which represent a more mature NK phenotype. B-cells, CD4⁺ T-cells, and CD8⁺ T-cells were similar between the two cords. The mean total cell/kg of recipient for each subpopulation between cord units are summarized in Table 1. A trend towards a difference in NK cells (p=0.1) was observed; the small number of samples, however, limited the interpretation of the analysis. Future studies involving additional DCBT recipients are ongoing to better define the influence of NK cells as well as other cell populations including mesenchymal stem cells and T-reg cells on cord graft outcome.