An Animal Model That Predicts Unit Engraftment in Double Unit Cord Blood Transplant Recipients.

Patients receiving double unit cord blood transplant (DCBT) have demonstrated improved engraftment and survival compared to single unit historical controls despite the fact that only one unit is responsible for sustained donor engraftment. However, the mechanism of this advantage and unit predominance is not understood. Therefore, we investigated DCBT in NOD/SCID/IL2Rγc^−/− (NOG) mice using mononuclear cells (MNC) from each unit of clinical double unit allografts with correlation of the murine engraftment with which unit engrafted in the patients. Our hypothesis was that the animal model would correlate with clinical unit engraftment. Nine patients with high-risk hematologic malignancies received myeloablative (n=5) or non-myeloablative (n=4) conditioning according to age and diagnosis, and 4–6/6 HLA-matched double unit allografts. Patients consented to ≤ 5% of each unit being used for research purposes. A minimum of 1 million MNC per mouse were transplanted by tail vein injection into sublethally irradiated NOG mice: each unit alone and then in combination in the same proportion as transplanted into the patients according to the relative TNC/kg of each unit. Experiments were established in duplicate or triplicate according to cell number available. Mice were sacrificed at 4–6 weeks and engraftment was determined by FACS analysis for human CD45+. In both human and murine DCBT recipients the contribution of each unit to engraftment was measured by quantitative PCR for informative short tandem repeat (qSTR) regions that distinguished each donor. Consistent with prior clinical studies all patients demonstrated donor engraftment by STR evaluation of day +21 bone marrow (BM) with one unit predominating (designated the “winner”). The average percent human engraftment in the murine experiments is shown in the Table (W = winning and L = losing unit as seen in the patients, DCBT is W+L). Notably, in 9/9 (100%) experiments, while each unit engrafted when transplanted alone, a single unit predominated when given as a double unit transplant which correlated with clinical unit predominance. Interestingly, there was no improvement in engraftment of the winner with the double unit co-transplant as compared to the winner transplanted alone.

Therefore, for the first time we have established a reliable in vivo model of DCBT using clinical specimens that predicts clinical engraftment. To date, the data suggests that unit predominance is due to an inherent advantage of the winning unit (exact mechanism to be determined) and that the presence of the losing unit does not augment the engraftment of the winner. This experimental model should permit further investigation of double unit biology.