Depletion of CD45RA+ naïve T cells for haploidentical stem cell transplantations: graft generation under good manufacturing practice and establishment of a flow cytometric quality control for clinical-scale grafts

In haploidentical stem cell transplantation (SCT), achieving a balance between graft versus host disease (GvHD), graft versus leukemia effect (GvL) and bridging the vulnerable phase of aplasia against viral infections is still a challenge. Graft preparation strategies attempt to achieve this balance by removing and retaining harmful and helpful cells. At this point it is known that T cell subpopulations hold different properties concerning GvHD promotion and immunocompetence towards pathogens. CD45RA+ naïve T cells show the greatest, while CD45RO+ memory T cells show less alloreactive potential but provide immunocompetence. CD45RA depletion is a promising new approach to graft processing that potentially combines GvHD prevention, GvL promotion and transfer of immunological competence by removing potentially harmful CD45RA+ naïve T cells and retaining CD45RO+ memory cells. This work focused on manufacturing CD45RA-depleted grafts within a one- or two-step approach, as well as a feasibility assessment of the process and the establishment of a 10-color fluorescence activated cell sorting (FACS) measurement panel for clinical-scale graft generation. CD45RA depletions were conducted from granulocyte-colony stimulated factor (G-CSF) mobilized peripheral blood stem cells (PBSC) applying two different strategies, direct depletion of CD45RA+ cells (one-step approach), or depletion following preceding CD34 selection. A 10-color FACS measurement panel was established ensuring quality control and enabling preliminary data acquisition on CD45RA co-expression for cell loss estimations. Residual virus-specific T cells after depletion were measured using MHC multimers. It was observed that the depletion antibody occupied the cell binding sites, resulting in insufficient binding of the fluorescent dye for subsequent FACS measurement. Therefore, three FACS antibodies were tested and compared, and CD45RA-PE (clone:2H4) was found to be the best choice for reliable cell detection. To further characterize residual T cells, two homing markers, CD62L and CCR7, were compared, with particular attention paid to the expression of the surface markers after cooling. Both markers were complementary to each other, resulting in the decision to include an additional FACS measuring tube whenever samples are cooled or further T cell characterization is needed. With a median log depletion of -3.9 (one-step) and -3.8 (two-step) data showed equally efficient removal of CD45RA+CD3+ T cells for both approaches. Close to complete B cell removal was obtained without additional reagent use. However, also close to complete NK cell loss occurred due to high CD45RA co-expression. Stem cells recovered at a median of 52% (range: 49.7 - 67.2%) after one-step CD45RA depletion. CD45RO+ memory T cells recovery was statistically not differing between both approaches. Virus-specific T cells were detectable after depletion, suggesting that virus-specific immunocompetence is transferable. In conclusion, CD45RA depletions are equally feasible for both approaches when performed from fresh, non-cryopreserved starting products, show reliable reduction of CD45RA and B cells, but also result in co-depletion of NK cells. Stem cell recovery and NK cell losses must be considered carefully especially regarding overcoming HLA barriers, pathogen protection during aplasia, early engraftment an GvL. Therefore, a combination of CD45RA-depleted products with already established other processing methods to ensure sufficient stem and NK cells is desirable to allow high clinical flexibility