大 B 细胞淋巴瘤中的 CD19 CAR-T 细胞治疗：临床证据、耐药、毒性与精准策略

CD19-directed chimeric antigen receptor (CAR) T-cell therapy has transformed the management of relapsed or refractory large B-cell lymphoma (LBCL), producing durable remissions in a subset of patients whose disease previously had few curative options.

Axicabtagene ciloleucel, tisagenlecleucel, and lisocabtagene maraleucel established CAR T-cell therapy in the third-line setting, and randomized studies subsequently moved axicabtagene ciloleucel and lisocabtagene maraleucel into second-line treatment for primary refractory or early relapsed disease. This review provides a clinically anchored, mechanism-focused synthesis of CAR T-cell therapy in LBCL.

We critically compare pivotal trials, long-term follow-up, patient-selection principles, and real-world evidence, emphasizing that apparent differences across products must be interpreted in light of eligibility criteria, analytic denominators, bridging therapy, manufacturing intervals, toxicity grading, and treatment crossover.

We then examine resistance and relapse as systems-level phenomena arising from antigen modulation, tumor-intrinsic evolution, impaired CAR T-cell fitness, suppressive myeloid and stromal networks, systemic inflammation, metabolic stress, and incomplete immune recovery. The biological basis and clinical implications of cytokine release syndrome, immune effector cell-associated neurotoxicity syndrome, prolonged cytopenias, infections, and late nonrelapse mortality are also reviewed.

Finally, we discuss circulating tumor DNA, metabolic imaging, single-cell and multi-omic profiling, artificial intelligence, dual-target and armored constructs, allogeneic platforms, and in vivo CAR programming as components of precision cellular therapy. The central clinical challenge is no longer whether CAR T-cell therapy can work, but how to select patients, deliver treatment rapidly, anticipate failure, and preserve long-term immune and functional health.