CAR-T cells manufactured after multiple myeloma triplet or quadruplet induction therapy have reduced functionality and an altered composition

Abstract Introduction: Multiple myeloma (MM) is an incurable plasma cell malignancy with increasing incidence in the United States. The approval of BCMA-targeting chimeric antigen receptor T cell (CAR-T) therapy was practice-changing for multi-drug refractory patients. Despite inducing deep remissions, CAR-T cell therapy remains non-curative, with one third of patients relapsing in just one year. Efforts to improve response duration by moving CAR-T cell therapy to second line in myeloma patients' treatment have improved progression-free survival, supporting integration into frontline therapy. However, exposure to triplet or quadruplet drug regiments during frontline therapy may be diminishing the efficacy of CAR-T. Here, for the first time, we report findings that suggest frontline anti-MM induction therapy has a negative impact on the efficacy of CAR-T cell therapy. Methods: CAR-T cells were generated fromblood samples obtained from MM patients at different timepoints throughout their treatment with informed consent and approval from an Institutional Review Board. T cells were negatively selected via Miltenyi's Human Pan T Cell Isolation Kit and manufactured into CAR-T cells as previously described. The anti-BCMA CAR T construct is composed of a Flag tag, a FHVH33 BCMA-binder, CD8 hinge and transmembrane domain, 4-1BB co-stimulatory domain and CD3ζ activation domain. Transduction efficacy was assessed by flow cytometry of CAR/Flag-PE+ T cells and manufacturing efficiency was defined as the final number of transduced CAR-T cells relative to the number of T cells transduced. Cytotoxicity was assessed by plating H929 cells with anti-BCMA CAR-T cells at varying effector:target (E:T) ratios for 48 hours followed by flow cytometry analysis of normalized MM cell viability, defined as live CD138+CD38+CD3-CD2-MM in treated conditions relative to untreated controls. Final CAR-T cell product characteristics were assessed via flow cytometry, given decreased CD4/CD8 ratios, increased exhaustion, and decreased naivety are CAR-T cell characteristics that have been reported as negative correlates with clinical response. Results: We manufactured CAR-T cells from age-matched healthy donors (n=4), MM patients who are newly diagnosed (n=5), post frontline induction therapy (n=5), or have received four or more lines of therapy (n=12). CAR-T cells from post-induction MM patients doubled during CAR-T manufacturing at the same rate as newly diagnosed or healthy donor CAR-T cells, but after heavy treatment, doubling time during manufactured increased. This resulted in a reduced manufacturing efficiency (one-way ANOVA, p=0.002). We found that post-induction CAR-T cells exhibited a 20.28% reduction in cytotoxicity against H929s at 1:64 as compared to newly diagnosed CAR-T cells (unpaired t-test p=0.03). These results remained consistent at higher E:T such as 1:16 where post-induction CAR-T cells were 14.67% less cytotoxic than newly diagnosed (unpaired t-test p=0.045). Post-induction CAR-T cells performed as poorly as those from heavily treated patients. There was no difference between CAR-T cells from newly diagnosed MM patients or age-matched healthy donors. Upon characterization, we found CAR-T cells from post-induction patients' have significantly lower CD4/CD8 ratios relative to newly diagnosed patients and similar to those from heavily treated patients (one-way ANOVA, p=0.0098). Further, we have found that post-induction CAR-T cells have increased expression of the inhibitory receptor, TIGIT, relative to newly diagnosed patients, and this marker significantly increases with additional therapy (one-way ANOVA, CD4+ p=0.002, CD8+ p=0.001). Post-induction CAR-T cells had similar expression of the inhibitory receptor PD-1 to newly diagnosed T cells that increased following heavy treatment (one-way ANOVA, CD4+ p=0.002, CD8+ p=0.034). Lastly, post-induction CAR-T cells were composed of significantly fewer naïve/stem cell memory cells compared to newly diagnosed CAR-T cells (one-way ANOVA, CD4+ p=0.016, CD8+ p=0.0099). Conclusion: Our data shows that frontline anti-MM treatment adversely effects CAR-T cell cytotoxicity. Consistent with this, we observed reduced CD4/CD8 ratios, more differentiated products, and phenotypic changes consistent with exhaustion. These results support the collection of T cells for CAR-T cell therapy prior to MM treatment initiation to improve the functionality of BCMA CAR-T cells and achieve more durable clinical responses.