Bispecific CAR T and CAR NK cells in dual-threat immunotherapy: opportunities and challenges

Antigen escape and intratumoral heterogeneity remain major barriers to durable responses in chimeric antigen receptor (CAR)-based cancer immunotherapies. While single-antigen CAR-T cell therapies have achieved notable success in hematologic malignancies, relapse driven by antigen loss, lineage plasticity, and tumor evolution remains common, and efficacy in solid tumors is limited. Similar antigen-dependent limitations are also observed in other targeted immunotherapies, including bispecific antibodies, underscoring the broader challenge of achieving durable immune control. To address these challenges, bispecific and combinatorial CAR strategies have emerged to broaden antigen coverage, enhance tumor selectivity, and reduce immune evasion. This review examines the biological mechanisms underlying antigen escape and critically evaluates dual-targeting CAR architectures, including tandem CARs, dual-CAR systems, logic-gated designs, and inhibitory CARs. We distinguish between mechanistic rationale and clinically validated benefit, highlighting the heterogeneity of outcomes across studies and the limited evidence supporting consistent superiority over optimized monospecific approaches. We summarize preclinical and clinical evidence supporting bispecific CAR-T strategies, particularly in B-cell malignancies, while discussing challenges related to toxicity, manufacturing complexity, and translational scalability. We further evaluate CAR-engineered natural killer (CAR-NK) cells as a complementary platform. Their innate cytotoxicity, favorable safety profile, and compatibility with allogeneic manufacturing offer important translational advantages, although limitations in persistence, tumor infiltration, and clinical validation remain. Finally, we position CAR-based therapies within the evolving immunotherapy landscape, emphasizing therapeutic sequencing, combination strategies, potential cross-resistance, and the need to align CAR design and platform selection with tumor-specific patterns of antigen expression and immune escape.