Automatic generation of alloreactivity-reduced donor lymphocytes and hematopoietic stem cells from the same mobilized apheresis product

Abstract

Introduction: In vitro or in vivo depletion of alloreactive T cells can facilitate haplo-identical hematopoietic stem cell transplantation (HSCT). Very satisfactory transplant outcomes were thus reported for TCRαβ/CD19-depleted hematopoietic stem/progenitor cell (HSPC) grafts. The current semi-automatic manufacturing process on the CliniMACS Plus, although robust, still requires a significant amount of manual labor to be completed. Towards advancing and further facilitating large scale cell processing, a new TCRαβ/CD19 depletion module combined with the previously described CD45RA depletion module (to serve as allo-reactivity attenuated donor lymphocyte infusion) was established on the CliniMACS Prodigy.

Methods: We evaluated six apheresis products from G-CSF-mobilized volunteer donors which were split automatically by the Prodigy, one portion each depleted of CD45RA⁺ or of TCRαβ⁺ and CD19⁺ cells. We investigated critical quality attributes for both products. Products were assessed for recovery of HSPCs and mature subsets, as well as depletion efficiency of targeted cells using flow cytometry. Effects of apheresis and product age post 48 h storage at 2-6 °C as well as freeze-thawing on product viability and recovery of WBC and HPSCs were assessed by flow cytometry.

Results: Ten sequential automatic processes were completed with minimal hands-on time beyond tubing set installation. Depletion efficiency of CD45RA⁺ resp. TCRαβ⁺ and CD19⁺ cells was equivalent to previous reports, achieving mean depletions of 4 log of targeted cells for both products. HSPC products retained TCRγδ⁺ and NK cells. 48 h storage of apheresis product was associated with the expected modest loss of HSPCs, but depletions remained efficient. Depleted products were stable until at least 72 h after apheresis with stem cell viabilities > 90%. Freeze-thawing resulted in loss of NK cells; post-thaw recovery of viable CD45⁺ and HSPCs was > 70% and in line with expectation.

Conclusion: The closed, GMP-compatible process generates two separate medicinal products from the same mobilized apheresis product. The CD45RA-depleted products contained functional memory T cells, whereas the TCRαβ/CD19-depleted products included HSPCs, TCRγδ⁺ and NK cells. Both products are predicted to be effectively depleted of GVH-reactivity while providing immunological surveillance, in support of haplo-identical HSCT.

Keywords: CD45RA-depleted DLI; CliniMACS Prodigy; GMP-compatible process; Haploidentical HSCT; TCRαβ/CD19-depleted graft.

Fig. 1

Overview of the study approach. A CliniMACS Prodigy TS 320 setup for the automated LP-TCRαβ-19-45RA Depletion procedure. A brief description of the components and preparation steps of the tubing set is added in Additional file 1: Figure S1. Starting material (mobilized apheresis products, mob LP) is automatically split for CD45RA and for TCRαβ/CD19 depletion which are performed sequentially and which results in two separate cell products. Target Cell Bag 1 (TCB1) contains a cell product depleted of naïve T cells and other CD45RA+ leukocyte subsets. Target Cell Bag 2 (TCB2) includes TCRγδ+ cells, HSPCs, NK cells, among others. Copyright © 2023 Miltenyi Biotec B.V. & Co. KG. All rights reserved. B Diagram showing the different use cases of the depletion software. C Workflow diagram of the manufacturing procedure with timeline and quality controls. White, dashed and gray shading mark hands-on, automatic-supervised and autonomous preparation steps, respectively. D Graphical description and schedule of experimental setup. All processes consisted of sequential CD45RA and TCRαβ/CD19 depletion. Ten depletion procedures from six mob LP (red) were performed. Four thereof were split runs to compare immediate processing vs. processing after 48 h storage at 2–6 °C to analyze the effect of apheresis product aging (mob LP, patterned red) on manufacturing performance and product quality (TCB1, patterned green and TCB2, patterned blue, respectively). Additionally, the effect of aging on the depleted products (TCB1, light green and TCB2, light blue, respectively) was assessed after storage at 2–6 °C for 48 h. Normal scale (5/10 runs NS) and large scale (5/10 runs LS) depletion reagents were used to test both, for depletion performance and final product quality. Final products from four split runs were cryopreserved to test how freeze–thaw procedure affects the final product quality at the end of the of their shelf life (TCB1, light green with snowflake and TCB2, light blue with snowflake, respectively, as well as TCB1, patterned green with snowflake and TCB2, patterned blue with snowflake, resepctively)

Fig. 2

Flow cytometric panels for characterization of CD45RA-depleted DLI (TCB1) and TCRαβ/CD19-depleted HSPCs (TCB2) in comparison to apheresis product (mob LP). A The gating strategy using the express mode “TCRab_CD45RA_Depletion_h_02” determines automatically the TCRαβ⁺ and CD45RA⁺ frequencies in the different fractions before and after depletion. Additionally, this panel includes the identification of the CD34⁺ cell population among CD45⁺ leukocytes, however the stem cell enumeration was determined using the BD’s CE-certified SCE kit following stringent ISHAGE gating (panel not shown, [26]). The arrows indicate, left to right, TCRαβ over TCRγδ gated on T cells, CD45RA over CD45RO gated on T cells, and CD45RA over CD45RO gated on TCRαβ T cells. The same analysis is shown, from top to bottom, for apheresis product (Mob LP), CD45RA-depleted DLI and TCRαβ/CD19-depleted HSPCs. CD45RA-depletion depletes almost all TCRαβ T cells, but the inverse is not true for TCRαβ-depletion: the lower panel demonstrates a high frequency of CD45RA⁺ T cells (lower panel, middle) despite almost complete depletion of TCRαβ T cells (lower panel, left). B The gating strategy using the express mode “Immune_Cell_Composition_human” facilitates automatic determination of cell concentration, viability of leukocytes, cell composition and frequency of B cells. The arrows identify B cells. Both CD45RA-depletion (middle panel) and TCRαβ/CD19-depletion (lower panel) almost completely deplete B cells from DLI (middle) or HSPC product (lower)

Fig. 3

Characterization of CD45RA-depleted DLIs. Apheresis product was either processed immediately after apheresis (fresh mob LP, red) or after storage at 2–6 °C for 48 h (aged mob LP, patterned red). CD45RA-depleted DLI product (fresh TCB1, green) was analyzed immediately after processing of fresh or aged mob LP or after storage at 2–6 °C for 48 h (aged TCB1, light green). A Relative contribution of individual mature leukocyte subpopulations in starting material and target cell fractions among CD45⁺ leukocytes at the different time-points. The percentage of each subtype was normalized to 100% CD45⁺ leukocytes. B Percentage of viable CD45⁺7-AAD⁻ leukocytes in all fractions at different time-points. Depletion efficiency of naïve CD45RA⁺ T cells and B cells (universally CD45RA⁺) was compared for products (C) derived from mobilized apheresis product (mob LP), processed immediately after apheresis (fresh mob LP) or after storage at 2–6 °C for 48 h (aged mob LP) and D derived from processes using NS or LS depletion reagents, respectively. Log depletion for each individual run is depicted with dots and squares representing products generated from fresh and aged mob LP or from NS and LS reagents, respectively. Red line indicates mean values achieving mean log depletions of around 4 log in all tested settings. Box plots and whiskers represent median with interquartile range and min and max values of all 10 depletion runs, respectively. Mean yield of CD34⁺ stem cells, NK cells, TCRγδ⁺ cells and CD3⁺ T cells among total CD45⁺ cells in depleted products (E) derived from fresh or aged mob LP and F from processes using NS or LS depletion reagents showing a high loss of cells of interest. G The CD4/CD8 ratio among total T cells was 6.5-fold increased in all depleted products. H Residual alloreactive CD45RA⁺ T cells in CD45RA-depleted products were calculated per 1 × 10⁶ CD3⁺ T cells which are the main active component of a DLI product. All data were tested for statistical significance using Student’s t-test. Shown is mean ± SEM, n = 4–10 individual runs; ns (not significant) indicates a p > 0.05, *: p < 0.05, ***: p < 0.001. P-values are reported above the corresponding groups

Fig. 4

Characterization of TCRαβ/CD19-depleted HSPC grafts. Apheresis product was either processed immediately after apheresis (fresh mob LP, red) or after storage at 2–6 °C for 48 h (aged mob LP, patterned red). The TCRαβ/CD19-depleted product (fresh TCB2, blue) was determined immediately after processing of fresh or aged mob LP or after storage at 2–6 °C for 48 h (aged TCB2, light blue). A Relative contribution of individual mature leukocyte subpopulations in starting material and target cell fractions among CD45⁺ leukocytes at the different time-points. The percentage of each subtype was normalized to 100% CD45⁺ leukocytes. B Percentage of viable CD45⁺7-AAD⁻ leukocytes in all fractions at different time-points. The efficiency of depletion of TCRαβ⁺ T cells and CD19⁺ B cells was compared for products C derived from mobilized apheresis product (mob LP) processed immediately after apheresis (fresh mob LP) or after storage at 2–6 °C for 48 h (aged mob LP) and D derived from processes using NS or LS depletion reagents, respectively. Log depletion for each individual run is depicted with dots and squares representing products generated from fresh and aged mob LP or from NS and LS reagents, respectively. Red line indicates mean values achieving mean depletion efficiency of around 4 log in all tested settings. Box plots and whiskers represent median with interquartile range and min and max values of all 10 depletion runs, respectively. Mean yield of CD34⁺ stem cells, NK cells and TCRγδ⁺ cells among total CD45⁺ cells in depleted products (E) derived from fresh or aged mob LP and F from processes using NS or LS depletion reagents showing a good recovery of cells of interest. G A CFU-C assay was performed via a 14-day culture followed by counting total number of colonies of mob LP and TCB2 at different time-points. Each individual run is plotted, as is mean ± SEM. H The residual alloreactive TCRαβ⁺ T cell number in HSPC graft was calculated per 4 × 10⁶ CD34⁺ stem cells representing the transplant dosage of 4 × 10⁶ CD34⁺ stem cells per kg as the main active component. All data were tested for statistical significance of differences using Student’s t-test. Shown are individual runs and mean ± SEM, n = 4–10 individual runs; ns (not significant) indicates a p > 0.05, *: p < 0.05, **: p < 0.01, ***: p < 0.001. P-values are reported above the corresponding groups

Fig. 5

Post-thaw characterization of aged products cryopreserved at the end of their shelf life (72 h post-apheresis). The effect of a freeze–thaw procedure after storage at -180 °C for at least 1 month was analyzed on A aged CD45RA-depleted DLI product and B aged TCRαβ/CD19-depleted HSPC graft regarding the parameters CD45 viability, viable CD45 recovery, CD3 recovery or CD45 viability, viable CD45 recovery, CD34 viability and viable CD34 recovery, respectively. Individual runs depicted as dots as well as mean ± SEM are shown; n = 4 of each independent cryopreserved products of aged TCB1 and aged TCB2 from 4 freshly processed mobilized apheresis products per depletion process. Recovery values were calculated compared to the corresponding pre-cryopreserved fresh product