Innovative Bone Marrow CAR-T Cell Manufacturing for AML: Enhancing Viability and Tumor Migration

Introduction

Manufacture CAR T lymphocytes is challenging for patients suffering acute myeloid leukemia (AML) due to several limitations, such as high number of circulating blasts, lymphopenia and dysfunctional T cells. Accordingly, current methods fail to produce CAR Ts for the majority these patients. When production is achieved, response rates range 30-49%, lower than those reported in B ALL or multiple myeloma. Thus, immunotherapies such as CAR Ts represent an unmet medical need for AML.

We have developed a new CAR T manufacturing method that solves this production challenge for AML, which we refer to Immunocoaching T Cells (ICTs).

Methods

49 cryopreserved BM samples with > 60% of blast cells from AML patients were included in the study. For CAR ICT cells manufacturing, AML whole BM samples were cultured with a bispecific antibody CD3-CD123 in combination with IL2, IL15, IL21, PD1, TIM-3, TIGIT, CTLA4 IDO-1, A2A and JQ1 for 7 days, in the presence of blasts. After that, ICTs or Healthy Donor (HD) T cells were transduced with a D28 & 4-1BB CD123 lentivirus (MOI=5) in combination with BX795. Before intravenously injection to the PDX mice at day 10, a double magnetic bead selection removes 1^st remaining AML cells and 2^nd positive selection of T cells. Minimal Residual Disease testing is used as a quality control to ensure no tumor cells are present in the final CAR ICT therapy product. The CAR T HD were produced following standard methods: HD PBMNCs were separated by ficoll, isolating CD3+ T cells with magnetic beads, cultured in IL-2 + IL-15 before transduction. Flow cytometry was used to monitor blast cell death, T cell expansion and activation, and subsequent transduction and phenotype.

To test the activity and migration of these cells, a PDX model using primary AML cells in immunocompromised NSG mice was used to evaluated the activity of these cells. To generate the model, 5x10^6 blasts were intravenously infused via the tail vein of the animals. For activity studies, 2.5 to 5x10^6 CAR+ T cells were infused two weeks post-infusion, when > 20% BM infiltration was achieved. Two weeks later, the animals were sacrificed, and a flow cytometry analysis was conducted to assess CAR T cell activity and various markers in the BM, peripheral blood, and spleen. For studies on CAR T cell migration to tissues and viability, a similar process was followed, but 5x10^6 CAR-T cells were infused at four weeks, with > 90% BM tumor infiltration, and 1 week later after the animals were sacrificed for studies. Additionally, cytokine production was studied in vitro.

Results

During the manufacturing process, an almost complete elimination of AML tumor cells was observed in 44 out of 49 samples (90%). The 7-day fold expansion of ICTs was 33 ± 49 (5-174). These ICTs cells exhibited a larger size and higher cytokine production compared to conventional methods (CD3-CD28 or IL-2 stimulation). In this regard, CAR ICT produced from four different AML samples showed substantially higher concentration at 4 days in the PDX mice compared to CAR T HD in BM (p-value 0.0001, median 0.01327 vs 9.950e-005%), spleen (p-value 0.0001, median 0.5944 vs 0.001603%) and blood (p-value 0.0001, median 0.2682 vs 0.004599%) (n= 12 vs 6 mice, 4 vs 2 donors). In addition to the high viability observed, CAR ICT also exhibit high killing activity vs control in all three organs, at least comparable to CAR T HD cells produced by the standard method (% AML cells BM, p value 0.5902, HD 28.31, ICT 37.49; Spleen, p value 0.2319, HD 1, ICT 9; Blood, p value 0.9398, HD 0.45, ICT 1.04; n>19 mice/group; n>6 donors/group). Notably, despite these enhancements, the animals did not experience a decline in well-being during the study period.

Conclusions

The ICT method for CAR T CD123 production solves the challenges of CAR T production for AML. Achieving high antitumor activity in vivo, persistence, and viability within tumor tissues, accompanied by increased cytokine production and greater in vivo expansion of the CAR+ population. Our hypothesis is that these enhanced T cells create anautocrine loop that promotes increased in vitro viability and survival.