癌症中的双特异性抗体-药物偶联物：治疗窗决定因素、肿瘤-免疫相互作用及临床转化

文献信息

中文摘要

抗体-药物偶联物（ADC）已将靶向载荷递送确立为一种重要的抗癌策略，但其疗效仍受限于抗原异质性、不完全内化、耐药性和剂量限制性毒性。双特异性抗体-药物偶联物（BsADC）增加了第二抗原或表位结合特异性，旨在扩大肿瘤覆盖范围或改变受体转运。该领域在2026年跨过了一个重要的转化门槛，当时izalontamab brengitecan（iza-bren；BL-B01D1），一种EGFR×HER3 BsADC，在中国首次获得监管批准，用于既往接受过治疗的复发性或转移性鼻咽癌和食管鳞状细胞癌，而III期研究也显示其在三阴性乳腺癌中具有生存获益。这些进展将核心问题从双特异性是否能产生抗肿瘤活性转变为何时能产生具有临床意义的治疗窗。本综述通过六个相互关联的决定因素评估BsADC的开发：靶点对生物学与共表达、结合几何构型与亲合力、内化与溶酶体转运、连接子-载荷化学与药物-抗体比、瘤内分布与旁观者活性，以及生物标志物指导的给药和患者选择。我们进一步探讨载荷介导的免疫原性细胞死亡、Fc/Fcγ受体依赖性摄取、抑制性肿瘤-免疫微环境和基于检查点的联合治疗如何可能在靶点识别之外改变活性。临床项目展示了不同的结果：iza-bren为该模式提供了监管验证，而MEDI4276、zanidatamab zovodotin和REGN5093-M114的经验表明，增强的结合或内化并不能保证可持续的开发。重要的是，许多关于免疫激活的机制和联合治疗证据仍来自单特异性ADC；因此，在没有直接证据的情况下，不应将免疫获益归因于双特异性本身。未来的进展应整合人体药代动力学/药效学数据、靶点对和免疫背景生物标志物、剂量-暴露-毒性关系，以及相对于优化单特异性ADC的比较价值。

英文摘要

Antibody-drug conjugates (ADCs) have established targeted payload delivery as an important anticancer strategy, but their efficacy remains constrained by antigen heterogeneity, incomplete internalisation, resistance and dose-limiting toxicity. Bispecific antibody-drug conjugates (BsADCs) add a second antigen- or epitope-binding specificity with the aim of broadening tumour coverage or altering receptor trafficking. The field crossed a major translational threshold in 2026, when izalontamab brengitecan (iza-bren; BL-B01D1), an EGFR×HER3 BsADC, received its first regulatory approvals in China for previously treated recurrent or metastatic nasopharyngeal carcinoma and oesophageal squamous cell carcinoma, while phase III studies also showed survival benefit in triple-negative breast cancer. These developments shift the central question from whether bispecificity can produce antitumour activity to when it can generate a clinically meaningful therapeutic window. This review evaluates BsADC development through six linked determinants: target-pair biology and co-expression, binding geometry and avidity, internalisation and lysosomal trafficking, linker-payload chemistry and drug-to-antibody ratio, intratumoural distribution and bystander activity, and biomarker-guided dosing and patient selection. We further examine how payload-mediated immunogenic cell death, Fc/Fcγ receptor-dependent uptake, suppressive tumour-immune niches and checkpoint-based combinations may modify activity beyond target recognition. Clinical programmes illustrate divergent outcomes: iza-bren provides regulatory validation of the modality, whereas experience with MEDI4276, zanidatamab zovodotin and REGN5093-M114 shows that enhanced binding or internalisation does not guarantee sustainable development. Importantly, much of the mechanistic and combination evidence for immune activation still derives from monospecific ADCs; immune benefit should therefore not be attributed to bispecificity itself without direct evidence. Future progress should integrate human pharmacokinetic/pharmacodynamic data, target-pair and immune-contexture biomarkers, dose-exposure-toxicity relationships and comparative value over optimized monospecific ADCs.