Limiting antigen escape through optimized design and alternative antigen targeting of chimeric antigen receptor T cells

Limiting antigen escape through optimized design and alternative antigen targeting of chimeric antigen receptor T cells

Purpose
T cells engineered to express a chimeric antigen receptor (CAR) against the pan B-cell marker CD19 have shown impressive efficacy as treatment for a variety of B cell malignancies. These impressive results have led to the approval of two CD19-directed CAR T cell therapies, Kymriah and Yescarta, for relapsed/refractory acute lymphoblastic leukemia (ALL) and/or diffuse large B-cell lymphoma (DLBCL), respectively. CARs are fusion proteins coupling the specificity of an antibody with the intracellular signaling components of a T cell. With this approach, the T-cell effector functions are redirected against a surface tumor antigen independent of major histocompatibility complex (MHC). Despite initial success and curative potential of CAR T-cell therapy several reports describing disease relapse with antigen-negative tumors are now emerging. This phenomenon occurs independently of tumor type and targeted antigen, thus indicating a general problem for mono-targeting CAR T-cell therapy. These observations highlight the clinical need for therapies that circumvent antigen escape. We propose targeting novel antigens alone or in combination as way to overcome antigen escape.

Results
We designed CAR T cells targeting CD79b and CD37, expressed on a variety of B-cell malignancies, in a mono-specific or tandem format with the CD19 single-chain variable fragment (scFv). We found CAR-specific activation, cytokine production, and in vitro cytotoxicity against a panel of cell lines with various expressions of tumor antigen. When tested in vivo we found potent cytotoxic efficacy of both CD79b- and CD37-based CAR T cells against cell lines and patient-derived xenograft models. Additionally, CD79bbased CAR T cells cleared CD19-negative tumors in vivo. Moreover, utilizing a truncated version of the natural ligand, a proliferation-inducing ligand (APRIL) as the CAR-binding moiety, we generated CAR T cells simultaneously targeting B-cell maturation antigen (BCMA) and transmembrane activator and CAML interactor (TACI). We found superior efficacy of APRIL-based CAR T cells when arranged in a trimeric - compared to monomeric form. In vivo the trimeric form of APRIL-based CAR T cells (TriPRIL-CAR) could eliminate BCMA-negative tumors.

Conclusion
With the acknowledged problem of antigen escape in relation to monospecific CAR T cells, designing CAR T cells targeting novel and multiple antigens are of clinical relevance. Therefore, we designed a panel of novel CAR T cells with the aim of mitigating antigen escape and found these to be promising alternatives to standard mono-targeting CD19 and BCMA CAR T cells for treating, and potentially preventing, CD19 or BCMA antigen escape in different disease settings.