Antibody-drug conjugates (ADC) are a class of prodrugs for targeted cancer therapy. Next-generation ADCs with impressive biophysical advances have seen a dramatic growth, as evidenced by four recent FDA approvals, yet holistic evaluation of their impact on clinical efficacy has not kept pace. Consequently, the clinical success rate of ADCs has not met expectations despite tremendous promise. ADC delivery to the site of action is a complex transport challenge. In particular, their heterogeneous tumoral distribution can severely limit cellular targeting in solid tumors. Using in vitro tumor spheroids, in vivo mouse models, and in silico PDE simulations, we show how modifying ADC physicochemical properties can improve ADC single-cell delivery and efficacy. Macromolecular properties like size and receptor-binding kinetics can influence transport parameters like ADC extracellular diffusion and internalization rate, which in turn influence ADC intratumoral distribution. For example, we show how single-domain antibodies can penetrate tissue better than a full-sized antibody, correlating to improved in vivo efficacy. Interestingly, we also show a counter-intuitive example where rapid internalization and high potency can result in lower in vivo efficacy due to poor tissue penetration. Similarly, the molecular weight, lipophilicity, and functional groups on the ADC cytotoxic payload can influence ‘bystander effects’ i.e. the payload can diffuse out of targeted cells to kill nearby untargeted cells, thereby potentially compensating for ADC distribution heterogeneity. Our simulations allow us to directly track the payload after it is released from the antibody, a feat currently unachievable experimentally. Using pharmacodynamic markers to indirectly track the payload, we show how ‘bystander payloads’ can reach nearly twice as many cell layers as the intact ADC at 72 hrs. This suggests that bystander effects can partially compensate for poor delivery, and by optimizing the payload physicochemical properties, more efficient bystander payloads can be designed. Payload bystander potential can be described by a dimensionless Damköhler number, Da (cellular uptake versus extracellular diffusion), and when Da is ~ 3, bystander payloads exhibit maximum tumor killing. Therefore, this joint experimental-simulation system can quantitatively predict in vivo ADC distribution behavior in a controlled and systematic manner, enabling improved clinical success of next-generation ADCs.

Single cell PK/PD of antibody-drug conjugates: Computational predictions and experimental validation to improve in vivo efficacy of targeted therapeutics

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.