Cord Blood (CB) Unit Mononuclear Cell (MNC) Dose: Effect on Transplantation Outcome and Relevance to Processing Method and CBU Selection.

Among the CB graft characteristics that affect post-transplant engraftment and survival TNC dose has become a critical variable in the selection of CB units. However, the cryopreserved TNC depends on the method of CB processing. The AXP^TM system, used by the NYBC-NCBP since August 2006, automatically reduces the volume and separates the components of a CB unit in a closed system, resulting in >95% recovery of MNCs, CFU and CD34+ cells and excellent viability, consistently, while the average recovery of TNC is 80%. MNCs represent 47% (±0.07) of the pre-processing TNC. Thus, when processing has no effect on type of cells recovered, a CB unit with a TNC of 900x10^6 (cell number criterion used for inclusion in the National Cord Blood Inventory) would have an average of 425x10^6 MNCs. As a result of the differential recoveries, however, of all CBUs with a pre-processing TNC >1100x10^6, 58% will have a post-processing TNC count ≥ 900x10^6; in contrast, 81.5% will have MNC ≥ 425x10^6 (N=11012 units processed with the AXP system during the period: 2006–2008). To evaluate whether the MNC dose has an impact on transplant outcome, and compare it to that of the TNC dose, we evaluated the effect on time to engraftment (time to ANC ≥ 500) and transplant related mortality (TRM) in all patients with leukemia or myelodysplasia that received single unit CB grafts from our Bank following myeloablative cytoreduction, during the period 1993–2006. A total of 1044 patients (mean age: 13.9 years, median: 9.2 years, range: 0.2–64 years, 22% >20 years) were included, 87% had follow-up data. CBUs had 0 (5%), 1 (33%) or ≥ 2 (61%) HLA mismatches with the recipient; the median cryopreservation TNC/kg was 4.1x10^7. Pre-freezing complete counts were obtained by automated hematology analyzers (H*1, Technicon, Bayer Corporation or Sysmex XE-2100, Roche Diagnostics). Until July 2006, CBUs were processed manually [PNAS 94(22);1995]. For this study, MNC included nucleated RBC. In a multivariate Cox regression analysis TNC/kg, HLA mismatch level, GvHD prophylaxis, transplant center and year were independently predictive of time to ANC ≥ 500. When MNC/kg was added to the analysis with TNC/kg, only the MNC dose was a significant predictor (RR:1.5, p:0.03), and the TNC dose was no longer predictive (p:0.7). Low MNC/kg (<1x10^7) was associated with delayed engraftment; time to ANC ≥ 500 shortened (showed by increasing RR of engraftment) with higher doses of MNC. Similar variables: HLA mismatch level, TNC/kg, race, transplant center and year affected TRM independently, and again, the significant effect of MNC/kg on TRM, when included in the multivariate analysis, removed the significance of TNC dose. As shown, low MNC/kg (<1x10^7) was associated with increased TRM (RR:1.8, p<0.001).

MNC dose, therefore, emerges as a better predictor of time to engraftment and TRM than the TNC dose, and thus, as a better measure of graft quality to be used in selecting CBUs for transplantation. In this context, because AXP processing retains more of the MNCs originally present in the CBU than other systems, this processing maybe used as a criterion of CBU quality when both the TNC and MNC are used as indices of CBU cellular contents.