A next generation inducible armored CAR to overcome the immunosuppressive tumor microenvironment and enhances cytotoxicity of CAR-T and TILs.

Abstract

e14517 Background: The full potential of CAR-T cells in clinic is limited by the immunosuppressive tumour microenvironment (TME), which induces expression of exhaustion markers and limits the activity of tumour infiltrating lymphocytes (TILs) including CAR-T cells. We developed a next generation inducible Armored CAR platform (aCAR) that releases an anti-PD-1 checkpoint inhibitor upon CAR-T cell activation, limiting payload release exclusively to the TME, thereby reducing the risk of systemic exposure. This differentiated strategy facilitates activation of CAR-T cells and TILs within the TME and has the potential for lower toxicity. Methods: A ROR1-targeting second-generation CAR containing a 41BB-CD3ζ intracellular domain was cloned into a lentivirus transfer vector alongside a prototypic anti-PD1 antibody. The CAR was under the control of a constitutively active promoter, and the PD1 targeting payload was controlled by an inducible promoter that was activated upon engagement of CAR with ROR1 on tumour cells. CAR-T cells were subject to in vitro co-culture assays with target ROR1+ tumour cells and their cytotoxic responses evaluated by flow cytometry. Levels and binding functionality of released payload were analyzed by ELISA and flow cytometry. In vivo xenograft models were performed in NSG mice with tumor growth assessed by bioluminescent imaging (BLI) and caliper measurement of the tumour volume. Results: Non-Armored CAR cells displayed potent and specific cytotoxic responses directed towards ROR1+ TNBC and NSCLC cells lines at different effector to target (E:T) ratios. The results were equivalent or superior to CARs generated with comparator anti-ROR1 antibodies, which may be due to the ROR1 CAR targeting a membrane proximal epitope within the ROR1 frizzled domain. The aCAR cells released measurable levels of anti-PD1 payload within 5 hours of binding to ROR1 on tumors and enhanced the cytotoxic effects at challenging 1:10 E:T ratios. Established PDL1+ TNBC xenograft models using the aCAR cells and comparing with non-Armored CAR cells displayed a qualitative abrogation in tumor growth by BLI, which was confirmed and shown to be significant by caliper measurement of the tumor volume. Continuing the experiment out to 3 months showed a significant survival advantage for the animals receiving aCAR. All other cohorts were terminated by day 70, however 20% of the aCAR cohort survived at day 95. Conclusions: Our next generation inducible aCAR platform enables the release of an immune stimulating payload only in the presence of target tumor cells, enhancing the therapeutic activity of the CAR-T cells and limiting payload exposure to the site of action. This technology provided a significant survival advantage in challenging in vivo xenograft models. This coupled with its potential safety attributes merits further clinical evaluation of this approach.