ABSTRACT
Clinical studies using autologous CAR T cells have achieved spectacular remissions in refractory CD19+ B cell leukaemia, however some of the patient treatments with CAR T cells failed. Beside the heterogeneity of leukaemia, the distribution and senescence of the autologous cells from heavily pretreated patients might be further reasons for this. We performed six consecutive large-scale manufacturing processes for CD20 CAR T cells from healthy donor leukapheresis using the automated CliniMACS Prodigy® platform. Starting with a CD4/CD8-positive selection, a high purity of a median of 97% T cells with a median 65-fold cell expansion was achieved. Interestingly, the transduction rate was significantly higher for CD4+ compared to CD8+ T cells and reached in a median of 23%. CD20 CAR T cells showed a good specific IFN-γ secretion after cocultivation with CD20+ target cells which correlated with good cytotoxic activity. Most importantly, 3 out of 5 CAR T cell products showed an increase in telomere length during the manufacturing process, while telomere length remained consistent in one and decreased in another process. In conclusion, this shows for the first time that beside heterogeneity among healthy donors, CAR T cell products also differ regarding cell senescence, even for cells manufactured in a standardised automated process.
ABSTRACT
Multiple clinical studies have demonstrated that adaptive immunotherapy using redirected T cells against advanced cancer has led to promising results with improved patient survival. The continuously increasing interest in those advanced gene therapy medicinal products (GTMPs) leads to a manufacturing challenge regarding automation, process robustness, and cell storage. Therefore, this study addresses the proof of principle in clinical-scale selection, stimulation, transduction, and expansion of T cells using the automated closed CliniMACS® Prodigy system. Naïve and central memory T cells from apheresis products were first immunomagnetically enriched using anti-CD62L magnetic beads and further processed freshly (n = 3) or split for cryopreservation and processed after thawing (n = 1). Starting with 0.5 × 108 purified CD3+ T cells, three mock runs and one run including transduction with green fluorescent protein (GFP)-containing vector resulted in a median final cell product of 16 × 108 T cells (32-fold expansion) up to harvesting after 2 weeks. Expression of CD62L was downregulated on T cells after thawing, which led to the decision to purify CD62L+CD3+ T cells freshly with cryopreservation thereafter. Most important in the split product, a very similar expansion curve was reached comparing the overall freshly CD62L selected cells with those after thawing, which could be demonstrated in the T cell subpopulations as well by showing a nearly identical conversion of the CD4/CD8 ratio. In the GFP run, the transduction efficacy was 83%. In-process control also demonstrated sufficient glucose levels during automated feeding and medium removal. The robustness of the process and the constant quality of the final product in a closed and automated system give rise to improve harmonized manufacturing protocols for engineered T cells in future gene therapy studies.
