Graft-Vs-Graft Immune Interaction Is the Likely Mechanism of Absolute Unit Dominance in Double Unit Cord Blood (DCB) Transplantation Using Patient DCB Grafts: An in Vivo but Not in Vitro Phenomenon

DCB transplantation (DCBT) appears to enhance engraftment despite sustained hematopoiesis usually being derived from a single unit. Therefore, to further investigate DCB biology we established in vitro and in vivo DCB models using samples from patient DCB grafts.

METHODS: Patients with high-risk hematologic malignancies were transplanted with DCB grafts and consented to ≤5% of each unit to be used for research purposes. We used mononuclear cells (MNC) or MicroBead selected CD34+ cells (85–90% enrichment) to assess short-term (2 week) progenitor colony-forming units (CFU) and long-term (5 week) stem cell (cobblestone-area-forming-cells, CAFC) potential of each unit alone and in DCB combination. In CFU assays cells were cultured immediately with G-CSF, KL, IL-3 and erythropoietin, or after overnight incubation in serum-free media with Flt3L, KL, and TPO. CAFC assays were performed using MS5 stroma. In addition, sublethally irradiated NOD/SCID IL2R-γ null (NOG) mice were transplanted with MNC or CD34+ via IV injection in the same ratio as transplanted into the patients. In all DCB assays the contribution of each unit was measured by qPCR for informative short tandem repeat (STR) regions and correlated with the engrafting (winning) unit in the patient.

RESULTS: All DCBT recipients studied (n=35) engrafted with a single unit except one patient who had sustained co-engraftment of both units. In CFU assays with MNC (n=5) or CD34+ (n=5) CFU were seen in winner cultured alone, loser alone, and with DCB. CFU number was not superior with winner alone and there was no augmentation of CFU with DCB. CFU DCB assays were a mix of both units with in vitro unit proportions concordant with the number of colonies formed by each unit alone (mean % winner in DCB ±SEM: 53%±12) with no correlation with the clinical winner. Further, overnight incubation did not significantly alter CFU numbers or unit proportions in DCB co-culture (54%±10). In CAFC assays with MNC (n=4) or CD34+ (n=5) CAFC were also present in all 3 conditions, not significantly superior in the winner, and there was no CAFC augmentation with DCB. CAFC in DCB assays were mixed in 6/9 and a single unit in 3/9 using either CD34+ or MNC with no correlation with the winner in the patient. Further, unlike the CFU assays, there was no concordance between CAFC number when each unit cultured alone and the in vitro unit dominance with DCB co-culture. The clinical winner was not predicted by the calculated number of infused CFU or CAFC to the patient. Murine experiments (n=27) included 19 with and 8 without DCB; using MNC in 20 and CD34+ in 7 (including 2/7 with CD34+ DCB plus CD34− add-backs). All mice engrafted with either unit alone or DCB at 4–7 weeks. Engraftment was not superior with the clinical winner alone (n=27) or in DCB (n=19).

Table: NOG Mice Engraftment by STR in DCBT of MNC or CD34+

CONCLUSION: in vitro CFU and CAFC DCB assays were predominantly mixed with no correlation with the clinical winner. However, as seen in human DCBT, murine MNC DCBT mimicked patient engraftment (11/12) with absolute single unit dominance (11/12). Notably, single unit dominance and clinical correlation was lost with CD34+ selection. Further, absolute unit dominance was restored by the add-back of the CD34− fraction. This is strong support that absolute unit dominance in vivo is linked to a graft-vs-graft immune interaction. Further experiments using this murine model are underway which promise to further elucidate DCB biology.