Abstract Recent studies suggest that Chimeric Antigen Receptor (CAR) binding affinity to its ligand affects CAR-T-cell functionality. Affinity engineering towards lower binding strengths might mitigate therapeutic side effects arising from intense CAR-T-cell activation as well as tumor relapse due to antigen-escape or limited persistence of CAR-T cells during sustained activation via high-affinity receptors. Here we characterize a broad range of CARs with varying affinities to the same target epitope and leverage the insights we gain to design a combined high- and low-affinity CAR product. While CAR affinity impacts in vitro functionality minimally, it strongly correlates with tumor control in vivo. Low-affinity binders cause only mild cytokine release syndrome (CRS) in humanized mouse models at the expense of anti-tumour efficiency. In mixtures with low-affinity CARs, high-affinity CARs maintain strong functionality while showing reduced signs of exhaustion and monocyte-induced cytokine production, compared to high-affinity CAR-T cells alone. In long term in vitro and in vivo settings, low-affinity CAR-T cells dominate over time, proving more resilience to chronic antigen exposure. Overall, our findings demonstrate that affinity combination represents a promising strategy to generate more effective CAR-T-cell products with an improved therapeutic index, beyond affinity engineering alone.
Warmuth et al. (2026) studied this question.