Complexities in comparing the impact of costimulatory domains on approved CD19 CAR functionality

Chimeric antigen receptors (CARs) are engineered to target T cells specifically to tumor cells, resulting in the engineered T cell killing the tumor cell. This technology has been developed to target a range of cancers, with the most notable successes in the treatment of B-cell malignancies where four approved therapies, all targeting CD19, are on the market. These four products differ in the costimulation domains, with axicabtagene ciloleucel (Yescarta) and brexucabtagene autoleucel (Tecartus) both utilizing the CD28 costimulation domain whilst tisagenlecleucel (Kymriah) and lisocabtagene maraleucel (Breyanzi) both utilizing the 4-1BB costimulation domain. There are clearly defined differences in how the CD28 and 4-1BB domains signal, yet it is difficult to ascertain which domain affords a superior mechanism of action given many other differences between these products, including overall CAR architecture and manufacturing methods. Additionally, while in vitro and preclinical in vivo studies have compared CARs with different costimulation domains, it remains a challenge to extrapolate differences observed in this biology across different experimental systems to the overall product performance. While there has been extensive preclinical and clinical work looking at CARs with a variety of targeting domains and architectures, this review will focus on the differences between the four marketed anti-CD19 CAR-Ts, with an additional focus on the impact of hinge and transmembrane domain on CAR activity and interaction with the target cell as well as other proteins on the surface of the T-cell.