Unraveling the molecular basis of ADC-related interstitial lung disease: from structural determinants to pathogenic pathways and mitigation strategies

Despite being a revolutionary class of precision oncology therapeutics, antibody-drug conjugates (ADCs) face growing restrictions on clinical application due to one of their most significant and life-threatening side effects—interstitial lung disease (ILD). The incidence of ADC-related ILD (ADC-ILD) exhibits marked heterogeneity across agents (less than 1% to more than 15%), indicating that this toxicity is not a general class effect, but rather fundamentally linked to the molecular structure of individual ADC molecules. This review synthesizes the association of structural elements of ADCs (linker chemistry, properties of payload, drug-to-antibody ratio (DAR), and target selection) with the risk of ILD, based on clinical data and insights into the mechanism. We propose a multidimensional risk model, informed primarily by trastuzumab deruxtecan (T-DXd) mechanistic studies, wherein pulmonary toxicity may arise from the convergence of three structural liabilities: cleavable linkers, high DAR, and membrane-permeable payloads. From a mechanistic standpoint, evidence from preclinical models, particularly T-DXd-related studies, suggests that important events in ADC- ILD may involve target-independent processes, including Fcγ receptor-mediated sequestration of intact ADCs by alveolar macrophages, subsequent liberation of the payload into lysosomes, and bystander injury to alveolar epithelial cells; this cascade may trigger inflammatory amplification via the upregulation of secreted phosphoprotein 1 (SPP1) and contribute to fibrotic remodeling. This mechanistic paradigm supports a “safety-by-design” approach for the next-generation of ADC engineering, including Fc optimization, linker stabilization, payload permeability rationalization, and DAR optimization; this review establishes a conceptual basis for safer ADC development and better clinical risk stratification, framing ILD as a potentially predictable and structure-modulated toxicity, rather than an entirely idiosyncratic adverse event. It must be emphasized that the proposed FcγR–macrophage–SPP1 axis remains predominantly supported by murine and in vitro data, and its generalization across all ADC constructs and validation in human pulmonary tissue require further investigation.