Immunotherapy with allodepleted donor T-cells generated using a CD25 immunotoxin improves immune reconstitution after haplo-SCT, but leukaemic relapse remains a problem. To develop a rational approach to refining allodepletion, we characterised the phenotype of proliferating alloreactive T cells flow cytometrically. CFSE-labelled T cells were co-cultured with HLA-mismatched dendritic cells and serial FACS analyses performed to determine expression of CD25, CD69, CD71, HLA-DR, OX40, ICOS, CD95, CD45RA, CCR7, IFNγ, TNFα, and IL–2. By staining each sample with CD3, CFSE and CD25 we were able to determine the phenotype of proliferating CD25 negative alloreactive T cells. In 5 donors, after a 3 day culture, CD25 was expressed in a mean of 83% (range 67–89%) of the proliferating (CFSE-dim) alloreactive T cells, confirming CD25 as an excellent target for allodepletion. 70% (39–81%) of the proliferating alloreactive CD25 –ve population expressed CD71, and 62% (50–74%) expressed CD45RA, identifying these as markers to target alloreactive T-cells that would persist after CD25 based allodepletion. To optimize our CD25 based allodepletion, we compared allodepletion using a CD25 immunotoxin or immunomagnetic beads in 6 HLA mismatched donor-recipient pairs. There was no significant difference between residual alloreactivity to host in primary and secondary MLRs and IFNγ ELISPOT assays. We then compared CD25 vs CD25/71 immunomagnetic depletion in 8 HLA-mismatched donor-recipient pairs. Rested allodepleted cells were restimulated with host or 3rd party stimulators in 2° proliferation or IFNγ ELISPOT assays. The median residual reactivity to host in IFNγ ELISPOT assays was significantly lower after combined CD25/71 than CD25 allodepletion (14.1% vs 54.6%, p < 0.05). Likewise, in 2° MLRs, CD25/71 depletion resulted in significantly lower residual proliferative response to host than CD25 depletion (median 4.8% of the response of unmanipulated PBMC vs 9.9%, p < 0.01). Third party responses after CD25/71 allodepletion were equivalent to unmanipulated PBMCs from the same donors in both assays. CD25/71 allodepletion also gave lower residual responses to host in both assays than combined CD25/45RA allodepletion. In IFNγ ELISPOT assays, anti-viral responses to CMV and EBV were preserved after combined CD25/71 allodepletion (CMV: unmanipulated 151 spots vs.CD25/71 123 spots, EBV unmanipulated 100 spots vs. CD25/71 142 spots).Similarly, the frequency of CD8 +CMV and EBV–specific T-cells assessed by pentamer analysis was not significantly reduced compared to unmanipulated PBMCs. We conclude that CD25/71 allodepletion will selectively delete 94% of the proliferating alloreactive T cells and enhances the depletion of alloreactivity compared with CD25-based methods. This strategy may facilitate immunotherapy with larger doses of allodepleted donor T-cells after haplo-SCT, enhancing graft versus leukaemia and antiviral effects.

Author notes

Disclosure:Research Funding: Medical Research Council.