(611e) Multiscale Delivery and Bystander Effect of a Trop-2 Targeted Antibody Drug Conjugate with a Hydrolysable Linker

Antibody drug conjugates (ADCs) are a rapidly growing class of targeted chemotherapy with five new drugs approved in the past two years. ADCs combine the specificity of a monoclonal antibody with the toxicity of a small molecule payload, attached via a chemical linker. Despite the recent burst in approvals, many ADCs have previously failed in the clinic, and poor drug delivery has recently been connected to low success rates. Limited tumor penetration of ADCs has been linked to poor therapeutic efficacy, but the complex design and impact of linker selection, payload potency and physicochemical properties, and dosing regimen on the distribution of the antibody and payload make development challenging. Many clinically approved ADCs have succeeded using potent payloads and highly stable linkers. However, a recently FDA-approved ADC, Trodelvy, has diverged from this typical strategy by using a hydrolysable linker with a moderately potent bystander payload (i.e. nM potency rather than pM potency) administered at higher doses than all other FDA-approved ADCs. The mechanism(s) behind this success, whether it’s higher antibody doses for improved tissue penetration, increased extracellular release for reaching adjacent ‘bystander’ cells, and/or improved intracellular payload release for more efficient killing, are currently unclear.

In this work, we employ computational (reaction/diffusion) and experimental quantitative pharmacology techniques to examine the interplay of dosing regimen, linker stability, and bystander effect on the multiscale distribution of Trodelvy given its unique drug delivery design – from the single-cell level to tissue level gradients, organ biodistribution, and systemic clearance. Near-infrared fluorescence imaging, quantitative image analysis, and single-cell flow cytometry measurements were used to characterize and quantify tissue level distribution after a single clinical dose (10 mg/kg) in an immunodeficient mouse model of TROP-2 expressing gastric cancer. After establishing that a single dose of Trodelvy targets the majority of cells, we examined the clinical dosing regimen of 10 mg/kg on days 1 and 8 of a 21-day cycle. Multiple doses further improved ADC distribution, likely because the remaining antibody from the first dose combines with the second dose, demonstrating effective tissue penetration from the high clinical doses of Trodelvy.

Another strategy to achieve more homogenous payload delivery uses payloads that exhibit bystander effect, which is determined by the payloads’ physicochemical properties. Bystander payloads can diffuse through cell membranes, typically from directly targeted cells to adjacent untargeted cells after intracellular linker cleavage. However, the hydrolysable linker on Trodelvy can additionally release payload in the tumor interstitium to target nearby cells. The effect of the released payload was investigated in vitro and in vivo using an immunofluorescent pharmacodynamic (PD) stain (γH2AX) to mark double stranded DNA breaks. In vitro studies revealed rapid appearance of PD signal when using Trodelvy compared to an equivalent ADC with a stable, enzyme cleavable linker. This indicates a more rapid and/or efficient intracellular linker cleavage. Furthermore, we observed the bystander effect in vivo using fluorescence imaging after ADC dosing. The comparison between Trodelvy and an equivalent ADC with an enzyme cleavable linker suggests that lower linker stability can increase efficiency of total payload release and the bystander effect, which could be facilitated by extracellular payload release.

In summary, we determined that the dose and schedule of Trodelvy result in efficient tumor penetration, and lower linker stability results in fast payload release to increase ADC efficacy. The results clarify how the unique payload release mechanism and lower potency payload (allowing higher doses) enable improved cancer cell killing. This quantitative approach to study multi-scale delivery can be used to inform the design of next-generation ADCs and prodrugs for other targets.