Umbilical cord blood transplantation (UCBT) is associated with impaired immune reconstitution and high infection-related mortality. Proposed potential mechanisms involve lack of antigen-experienced cells in UCB and impaired generation of antigen specific effectors due to Th2/Tc2 skewing, imprinted by placental factors. Furthermore, transplantation of two UCB units and mixed chimerism, albeit transient, may affect immune reconstitution. We evaluated the capacity of the recovering immune system to establish antigen-specific responses by quantitative and functional assessment of cellular immunity, in adult patients undergoing UCBT. Thirty patients with a median age of 51 years with hematopoietic malignancies underwent reduced intensity conditioning (Flu/Mel/rATG) followed by two sequential UCB infusions. GvHD prophylaxis consisted of tacrolimus and sirolimus. Here, we report our preliminary results of 15 patients who completed one year of follow up. Assessments were conducted prior to and at various time intervals until 12mo post-UCBT. CD4+ and CD8+ T cells decreased until 8wks post-UCBT and gradually increased thereafter. However, CD4+ and more strikingly CD8+ cells remained below normal by 12mo (median values 0.2x103/ul and 0.06x103/ul respectively). Most of the recovery was in the CD4+CD45RA+ population, which showed a slight initial drop but subsequently increased to normal. In contrast, CD4+CD45RO+ cells after the initial drop reached a low-level plateau by 6mo and never achieved normal levels. B lymphocytes remained low until 8wk but rapidly increased thereafter, exceeding normal levels by 12mo. CD16+CD56+ NK cells and CD14+ monocytes gradually increased to reach normal values by 6 and 12mo respectively. We evaluated functional immune reconstitution in vivo by assessing CMV-specific effectors by IFN-g ELISpot and additionally, in HLA-A*0201 positive patients, by HLA-A*0201/CMV-pp65 pentamer (NLVPMVATV) and flow cytometry. 14/15 patients and 2/30 UCB products were CMV seropositive prior to transplant. At a median time of 4 weeks, CMV antigenemia was detected in 6/15 UCB recipients and 5 of them developed functional CMV-specific effectors as determined by ELISpot. Functional CMV effectors were also detected in all 8/15 patients who did not develop detectable CMV antigenemia. The one seronegative patient developed neither CMV antigenemia nor CMV effectors. In HLA-A*0201 patients CMV-specific effectors could also be detected by pentamers. Unexpectedly, identification of functional effectors by ELISpot was only associated with the numbers of naive cells (CD4+CD45RA+ (p=0.03), CD8+CD45RA+ (p=0.02)) and CD16+CD56+ NK cells (p=0.01) and was independent of GvHD or mixed chimerism. Taken together these results indicate that immune reconstitution after UCBT is characterized by delayed recovery of CD4+ and more notably CD8+ cells and predominance of naive T lymphocytes. However, reconstituted naive T cells are capable of developing into functional effectors after in vivo exposure to antigen. This raises the possibility that vaccination, in vivo expansion or adoptive transfer of ex-vivo expanded antigen-specific effectors may provide successful strategies to overcome infection-related mortality in UCB recipients.

Author notes

Disclosure: No relevant conflicts of interest to declare.