P-004: Pharmacological inhibition and depletion strategies for SLAMF7 CAR-T cells in multiple myeloma

Background Recent and soon-to-expect approvals have significantly increased the number of patients treated with CAR-T cells and clinical responses can be very impressive. However, side effects including cytokine storm, neurotoxicity or on-target/off-tumor toxicity can be challenging to manage. Further, malignant transformation of piggyBac modified CAR-T cells was recently reported in a phase I clinical trial as a potentially life threatening event. Reliable strategies to control or even deplete CAR-T cells are thus urgently needed. Methods We generated SLAMF7 CAR-T cells as previously described and exposed the cells to clinically relevant doses of the following drugs: dexamethasone, fludarabine, mafosfamide, dasatinib, cetuximab, and belantamab-mafodotin. After defined time periods, we evaluated absolute numbers of CAR-T cells and their effector functions against myeloma cell lines (cytotoxicity, cytokine secretion, proliferation). Statistical analysis was performed using one-way ANOVA, followed by Dunnett’s multiple comparisons test if appropriate. The study was approved by the local ethics committee. Results Dexamethasone, used in the setting of severe CRS or ICANS, minimally decreased T cell numbers and cytokine release, but did not interfere with lytic capacity of the CAR-T cells. Fludarabine and the pre-activated cyclophosphamide analog mafosfamide, both drugs with potent lymphodepleting properties, significantly depleted numbers of CAR-T cells, reduced antigen-specific proliferation and led to an overall decrease in target cell elimination, even if specific lysis and cytokine release of the few surviving cells was preserved. Of note, fludarabine and mafosfamide similarly depleted unmanipulated T cells. Dasatinib, previously identified as a potent inhibitor of intracellular CAR signaling, had no impact on absolute T cell numbers, however lysis of myeloma cells, cytokine secretion and antigen-specific proliferation were entirely blocked. Evaluation of selective elimination of CAR-T cells with cetuximab, targeting the EGFR transduction marker of our CAR, showed modest CAR-T cell depletion in an ADCC assay. This was potentially related to competing natural killer cell elimination by the SLAMF7 CAR-T cells. To circumvent the problem of impaired ADCC, T cells were genetically engineered to express BCMA to enable simultaneous depletion of T cells and myeloma cells. Exposure to the BCMA-directed antibody-drug conjugate belantamab-mafodotin resulted in potent elimination of BCMA-positive T cells while unmanipulated cells were largely spared. Conclusions In aggregate, our data show that the different drugs can have justification in different clinical settings of CAR-T cell toxicity. While dasatinib and dexamethasone mainly act on T cell function, fludarabine and mafosfamide effectively deplete T cells in an unselective manner. Targeted elimination of CAR-T cells with antibodies can be augmented further by the use of antibody-drug-conjugates.