Mechanistic Modeling to Meet Challenges in ADC Development

Antibody-drug conjugates, or ADCs, are monoclonal antibodies that are linked to another drug (“payload”) via a molecular linker. This technology allows selective delivery of the payload, often a cytotoxic drug, to the target cell via receptor internalization.

While still a relatively new modality, ADCs are rapidly gaining popularity. Mylotarg was the first ADC that was approved in 2000 for the treatment of leukemia. By 2024, over 370 ADCs have entered the clinic (Colombo et al. 2024), and 15 have been approved in the US.

Most ADCs treat cancer. However, scientists are also developing novel small molecule payloads for other indications, including autoinflammatory diseases. There have also been advancements in ADC technologies in preclinical development, with bispecific ADCs, dual payload ADCs, degrader payloads, etc. (Beacon Intelligence, 2024)

Scientists used to think of ADCs as “site-directed missiles” directing their cytotoxic payload to the target cells. Yet, clinical experience has shown that ADCs behave more like “tethered mines.” Conjugation to an antibody decreases the systemic distribution of the cytotoxic payload. But it doesn’t eliminate exposure to other tissues. Therefore, maximizing the therapeutic index is still a concern during the design and dose optimization of ADCs.

Source: Barnscher (2023) https://www.clinicalleader.com/doc/the-clinical-landscape-of-adcs-in-diverse-technologies-narrow-target-0001

Successful development of ADCs starts at the design stage, continuing through clinical trial design and patient selection. Design of ADCs is especially challenging due to the many components requiring optimization, including the antibody, linker, and payload. Optimal drug properties often depend on physiological and target-specific biological parameters as well. The following table outlines the system- and/or drug-dependent or independent parameters to optimize.

As an ADC progresses through development, understanding how its drug properties and dosing regimens impact efficacy and toxicity becomes critical to program success.

Quantitative systems pharmacology modeling can inform the entire development cycle for an ADC. Modeling can address different stages of development for ADCs and Certara has two modeling solutions that answer development questions at R&D, preclinical, and clinical stages.

A platform model containing relevant mechanisms can translate preclinical findings to the clinic.

To assist ADC development at the preclinical and clinical stages, we developed a validated platform QSP model of ADCs. The model describes mechanisms generalizable to ADCs as a class. We validated the model against two clinically approved ADCs (trastuzumab-DM1 and trastuzumab-DXd). Our QSP experts developed the model using preclinical data for the two drugs as well as literature data for target-specific and physiological parameters.

The model captured both the preclinical activity of the drugs and predicted clinical efficacy. Learn more about how we developed this model:

• [Publication] Towards a platform quantitative systems pharmacology (QSP) model for preclinical to clinical translation of antibody-drug conjugates (ADCs)
• [Webinar] Exploring ADC PK using QSP

Using the platform model, we recently developed a set of 42 models for early feasibility analysis for ADCs. The models cover common ADC formats such as:

• monospecific,
• membrane receptor targeting drugs
• novel formats such as bispecific ADCs simultaneously targeting a membrane receptor and a soluble ligand.

These models can support design decisions. They can also help answer questions about how ADC properties (target, affinity, linker stability, payload potency) may impact dose-response. In conjunction with in vitro data, these models can provide an early understanding of how in vitro potency may translate to dose-response in humans.

The development of ADCs is no longer solely about more potent payloads or more stable linkers. The future lies in alignment—bringing together advanced drug design, patient-focused approaches, and cutting-edge technologies like QSP modeling.

With AI and machine learning set to augment modeling capabilities, the possibilities for precision drug development will only expand further. This will empower developers to design smarter ADCs, enter trials with greater confidence, and ultimately deliver life-changing therapies to patients faster than ever before. Adopting mechanistic modeling approaches early in the process will accelerate development and help navigate the complexities associated with ADC development.