Meta-Analysis of Exposure-Adverse Event Relationships for Antibody-Drug Conjugates

Antibody-drug conjugates (ADCs) have become a vital class of therapeutics in oncology because of their ability to selectively deliver potent drug molecules to tumor cells. However, ADC-associated toxicities cause high failure rates in the clinic and hinder their full potential. Due to the complex structure and pharmacokinetics of ADCs, it is challenging to identify the drivers of their toxicities. Here, quantitative analysis was performed to correlate the incidence of clinical adverse events (AEs) with nine different commonly measured exposure parameters collected from study-level summary data. We considered ADC analytes for different classes of ADCs, to identify ADC analytes that are strongly associated with the AEs for ADCs. Published clinical exposure and safety data for any grade and grade ≥3 AEs from 40 publications across six ADCs and three payloads were collected and analyzed. Exposure-AE relationships were quantified using logit models, and the strength of the correlations and rank order were determined. The analysis suggests that deruxtecan ADC-related toxicities correlated most strongly with the exposure of the free payload; monomethyl auristatin E (MMAE) ADC-related toxicities correlated with the free MMAE area under the curve; and pyrrolobenzodiazepine ADC-related toxicities correlated with no specific analyte but the dose. These findings agree with the published literature and support the notion that AE profiles are often shared by ADCs that deliver the same cytotoxic payload. The exposure-AE relationships presented here, together with identification of the most informative ADC analytes, may facilitate more focused mechanistic studies on the drivers of clinical AEs and could support dosing decisions during clinical development of ADCs.