Although adoptive transfer of chimeric antigen receptor (CAR)-engineered T cells has achieved unprecedented response rates in patients with certain hematological malignancies, this therapeutic modality is still far from fulfilling its remarkable potential, especially in the context of solid cancers. Antigen escape variants, off-tumor destruction of healthy tissues expressing tumor-associated antigens (TAAs), poor CAR-T cell persistence, and the occurrence of functional exhaustion represent some of the most prominent hurdles that limit CAR-T cell ability to induce long-lasting remissions with a tolerable adverse effect profile. In this review, we summarize the main approaches that have been developed to face such bottlenecks, including the adapter CAR (AdCAR) system, Boolean-logic gating, epitope editing, the modulation of cell-intrinsic signaling pathways, and the incorporation of safety switches to precisely control CAR-T cell activation. We also discuss the most pressing issues pertaining to the selection of co-stimulatory domains, with a focus on strategies aimed at promoting CAR-T cell persistence and optimal antitumor functionality.

Two recent studies, by Casirati et al. and Wellhausen et al., report genetically engineering normal hematopoietic stem and progenitor cells (HSPCs) to be resistant to chimeric antigen receptor (CAR)-T cells, by changing a single amino acid in the target protein that abrogates CAR binding, without compromising protein function. This allows for selective targeting of cancer cells without harming normal hematopoietic cells.