Preclinical advances and mechanistic insights of CAR-T therapy for acute myeloid leukemia: from target iteration to microenvironment regulation.

[INTRODUCTION] Relapsed/refractory acute myeloid leukaemia (AML) carries a dismal prognosis, primarily due to profound biological heterogeneity and the scarcity of effective targeted therapies. Chimeric antigen receptor T (CAR-T) cell therapy has emerged as a transformative investigational strategy for AML by genetically engineering T cells to specifically target tumour antigens; however, its clinical translation is severely impeded by target antigen scarcity and the immunosuppressive tumour microenvironment (TME).

[DISCUSSION] This review systematically summarizes the preclinical advances and mechanistic underpinnings of CAR-T therapy for AML, focusing on two core bottlenecks. First, it elaborates on the screening of candidate antigens (e.g. CD33, CD123) and cutting-edge target optimization strategies, including nanobody-based CARs, logic-gated systems, adapter-CAR technology and combinatorial approaches to mitigate antigen escape. Second, it dissects the inhibitory effects of the AML TME on CAR-T function and reviews corresponding intervention tactics, such as immune checkpoint blockade, cytokine arming and gene editing. Integrating key updates from the 2024 ASH Annual Meeting, the review also highlights emerging preclinical technologies, including off-the-shelf CAR-T, CAR-NK cells and γδ T cell therapy. Importantly, it acknowledges the prevalent preclinical-to-clinical translation gap, where promising lab efficacy has not yet translated into consistent clinical success.

[CONCLUSIONS] Multidimensional technological innovation and the synergistic optimization of combination therapies are critical to overcoming AML-specific barriers. These advances hold the potential to unlock the precise clinical application of CAR-T therapy, ultimately improving survival outcomes for patients with relapsed/refractory AML.