CAR 2.0: The Next Generation of Synthetic Receptor–Based Cellular Therapy for Cancer

The Advantage of Costimulation

The major design improvement leading to success of CAR T-cells in patients was the addition of a costimulatory domain to the CAR backbone, providing a second signal in addition to that provided by the T-cell receptor signaling domain, leading to enhanced T-cell engagement and persistence in preclinical models and clinical trials. ^, The two most common costimulatory domains included in the CAR construct are CD28 and 4-1BB (CD137) although others, such as OX40 , have been tested in patients. Inherent differences between CD28 and 4-1BB in relation to T-cell activation via the endogenous T-cell receptor appear to have functional implications when implemented in the CAR design. CD28 is uniformly expressed on T-cells, is activated by binding to CD80 or CD86 on antigen-presenting cells, and leads to rapid activation of the T-cell with a skewing toward effector differentiation. 4-1BB, on the other hand, is expressed only on activated T-cells and, when engaged by its ligand (4-1BBL), leads to further stimulation of the T-cell but skewed toward a memory T-cell phenotype. Consistent with this natural biology, preclinical work has shown that synthetic 4-1BB-containing CARs are less prone to T-cell exhaustion and lead to superior T-cell persistence in animal models. ^, Indeed, T-cell durability has been observed in many clinical trials using 4-1BB-based CARs and, although not directly compared, seems more prolonged than persistence of CD28-based CARs. Third-generation CARs, incorporating two costimulatory domains, introduce an additional layer of complexity and have shown superiority to second-generation CD28-based CARs in small series although results have been inconsistent. Preclinical work incorporating ICOS in addition to 4-1BB has demonstrated superiority of these third-generation CARs in some studies but have not been tested in patients. An alternative design was introduced, which uses the natural ligand for 4-1BB, 4-1BBL introduced into the T-cells with a CD28-based second generation CAR, and generates the complementary ligand activation by cis-signaling, leading to improved function and durability in preclinical models. This approach, termed “armored CAR,” is currently tested in a clinical trial ( NCT03085173 ), with preliminary results demonstrating persistence in several patients. Finally, targeted mutation-selective immune-receptor tyrosine-based activation motifs (ITAMs) in the CD3 zeta chain of CD28 containing CARs have also demonstrated improved persistence in preclinical studies, suggesting that the alterations to the CAR design can modify in vivo behavior of CAR T-cells.