Introduction: CAR T-cells targeting CD19 (CAR19s) can eradicate B cell leukemias and lymphomas. The effectiveness of CAR19s is linked to their robust expansion properties but also their long-term persistence. Persistence is maintained by normal CD19+ B cells: a non-tumor dependent, self-renewing source of antigen. In this manner, CAR19s are quite unique. We have re-engineered CAR19s to secrete a wide variety of retargeting fusion proteins (FPs) by encoding expression cassettes downstream of the CAR sequence in lentiviral vectors. By hijacking CAR19s, we utilize their inherent persistence properties. By designing multispecific FP, we directly counter the clinically critical issues of tumor heterogeneity and antigen loss. Experimental Procedures: A lentiviral vector with an MCSV promoter was used to express the CAR19 construct and FPs. FPs and multispecific-FPs were designed to encode the extracellular domain of the CD19 protein, followed by one or two scFv sequences, separated from the CAR sequence by a P2A cleavage site, a design termed IMPACTtm (Integrated Modular Proteins for Adoptive Cell Therapy). The FPs therefore consist of the CD19 extracellular domain linked to one or more scFvs. The FPs redirect CAR19 T-cell cytotoxic activity to any tumor antigen of interest by coating that antigen with CD19 via the scFv. Further, multiple antigens can be coated with CD19 by encoding multiple scFv in the FP. Hijacked CAR19s therefore serve as a platform for targeting diverse antigens. Results: Here we describe one example in detail, focusing on Her2+ solid tumors. The CD19/anti-Her2 FP was highly potent in cytotoxicity assays targeting Her2+/CD19- solid tumor cell lines. The concentration of fusion protein required to reduce tumor cell number by 50% was 10 pM (0.7 ng/ml). Primary donor T-cells transduced with the CAR19 - CD19/anti-Her2 FP lentiviral vector secreted > 20 ng/ml of FP in cell culture. Cytotoxic activity of this redirected CAR19 against Her2+ SKOV3 tumor cells was demonstrated in vitro and in vivo. CD19-mediated persistence was demonstrated in serial restimulation assays. A bispecific FP containing CD19 linked to anti-Her2 and anti-EGFR scFv had specific activity against both antigens with a potency of 0.75 pM. For each antigen, the potent cytotoxicity was specifically mediated by the secreted fusion protein. Additional program examples of multispecific targeting for diverse hematologic and solid tumor types will be shown. Conclusions: The IMPACT platform addresses critical issues in cell therapy including CAR persistence, antigen escape and antigen heterogeneity, and provides important solutions for treating both hematologic and solid tumors. The potency of redirected cytotoxicity supports clinical development of CAR19/IMPACT programs, four of which are now ready for IND enabling studies.

Citation Format: Paul Rennert, Fay Dufort, Lihe Su, Lan Wu, Alyssa Birt, Christine Ambrose, Roy Lobb. Hijacking CAR19 T-cells for use in targeting diverse hematopoietic and solid tumors [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A040.