ABSTRACT
Key defining attributes of an antibody-drug conjugate (ADC) include the choice of the targeting antibody, linker, payload, and the drug-to-antibody ratio (DAR). Historically, most ADC platforms have used the same DAR for all targets, regardless of target characteristics. However, recent studies and modeling suggest that the optimal DAR can depend on target expression level and intratumoral heterogeneity, target internalization and trafficking, and characteristics of the linker and payload. An ADC platform that enables DAR optimization could improve the success rate of clinical candidates. Here we report a systematic exploration of DAR across a wide range, by combining THIOMAB protein engineering technology with Dolasynthen, an auristatin-based platform with monomeric and trimeric variants. This approach enabled the generation of homogeneous, site-specific ADCs spanning a discrete range of DARs 2, 4, 6, 12, and 18 by conjugation of trastuzumab IgG1 THIOMAB constructs with 1, 2, or 3 engineered cysteines to monomeric or trimeric Dolasynthen. All ADCs had physicochemical properties that translated to excellent in vivo pharmacology. Following a single dose of ADCs in a HER2 xenograft model with moderate antigen expression, our data demonstrated comparable pharmacokinetics for the conjugates across all DARs and dose-dependent efficacy of all test articles. These results demonstrate that the Dolasynthen platform enables the generation of ADCs with a broad range of DAR values and with comparable physiochemical, pharmacologic, and pharmacokinetics profiles; thus, the Dolasynthen platform enables the empirical determination of the optimal DAR for a clinical candidate for a given target.
